Can I avoid oxy. & nit. peaks while quantitation-6890 RG

[avrc]

Hi friends,

I am using an RGA 6890. I analyse for H2, C1 to c5 in my gas samples and I am using a calib mix to calibrate my GC. We are using bladders to collect gas samples. But almost every time I analyse a sample, due to the unwanted inclusion of oxygen and nitrogen(which gets in during the time of sample collection by the plant people), my hydrogen value gets lowered. I have to ask for another sample and the problem vanishes only when the air contamination is less than 1%.

I cannot use a sampling bomb due to practical difficulties!

The unwanted air takes up some volume in the sampling loop which otherwise will be occupied by hydrogen(my sample contains around 90% of hydrogen), so just leaving out oxygen & nitrogen peaks will not solve my problem.

I hope I am making myself clear on the above point!

Waiting for your suggestions!

Thankyou all.

[chromatographer1]

Why can't you just calculate Hydrogen by difference? if it is generally 90% or higher?

Measure the other C1-C5 peaks and then calculate the hydrogen as the difference from 100%.

It has been done before !

[carrolr]

Implement equivalent of sample "bomb" in bladder by flushing bladder with concentric inlet / outlet tubes. Use reducing "T" to join concentric tubes. Have used this arrangement but not for hydrogen -- would assume you could safely vent the excess sample during flush of bladder.

No direct experience with this sample matrix, but have used drier for sample into thermal conductivity detector. Is a moisture / oxygen scavenger possible at the inlet to your GC sample loop? These play havoc with the hot wire sensors in the detector.

[avrc]

Thanks chromatographer1 and carrol for your suggestions.

I am taking the calibrated value for hydrogen and keeping it as it is and normalise the other components so that the total comes to hundred. I am doing this because i am mainly concerned with my hydrogen concentration only. So, when air gets into the sample it reduces the area counts for my hydrogen and so naturally my hydrogen conc. comes down.
What i am asking for is a way to avoid this with present type of sampling.

For the suggestion by carrol; I dont understand what u are trying to say! Do u want me to use a bladder with opening in both sides with tubes connected in each side? We dont use bladders of that kind; we just use football bladders with a pinch-cork to close the mouth.

Thanks.

[carrolr]

I was just suggesting a way of excess flushing the bladder when there is just one entrance hole. Using common technique of smaller tube inside larger tube and a reducing "T" on the tubes to separate in / out flows.

For example, 1/8 inch tube inside a 1/4 inch tube. Use a Swagelok reducing "T" having one end 1/4 and one end 1/8. 1/8 tube goes through 1/8 end (swaged) through 1/4 end and through the swaged 1/4 tube. The 1/ 4 tube attaches to the bladder and the 1/8 tube ends up deep inside the bladder. Flows are into the 1/8 tube, through the bladder, and out the "T" side arm. Pinch cocks on rubber hoses on inlet and outlet tubes.

This would also allow the bladder to be part of a process sampling loop.

Same result as fill / squeeze empty type flushing. Quicker, but more complicated (more chance for error).

[avrc]

Thanks Carrol for your explanation. I will try that out and inform you.

But finally you have told that this is complicated and there is more chances for error.

Do you mean to say that there is a chance for more air to get into the sample if I use as you have suggested? Or do you mean the older system i'm using (fill, squeeze & fill again), is error prone? Please tell.

[chromatographer1]

I think I understand your situation.

Your sample is being diluted by air entering your sample during sampling and/or possibly during introduction into the GC.

Why don't you calculate the amount of air that has entered the sample from a calibration using a std mix and then recalculate the amount of hydrogen and hydrocarbon sample actually injected subtracting the volume of air that was accidentally introduced.

You can also correct for the deviation from ideal gas laws if you really wish to be diligent about it.

I assume your hydrogen/hydrocarbon test sample does not contain air.

[arie2044]

hi
to avoid air in your samples try to use bladers with inlet and outlet tubes, so your sample will flow throuh, or use glass sampling bulbs with inlet & outlet. we had this problem far in history so we switched to glass & steel cylinders. in some types of bladers there is a diffusion of hydrogen out of the blader.

Arie

[avrc]

Thank you chromatographer1 & arie.

chromatographer1, my sample should not have any air in it.
How can I calculate the volume of air that has entered even after calibrating for air? Please explain.

SORRY FOR MY DELAYED RESPONSE!

avrc.

[arie2044]

Hi Avrc
in your first message you wrote " But almost every time I analyse a sample, due to the unwanted inclusion of oxygen and nitrogen(which gets in during the time of sample collection by the plant people), "
that is the source for air. Calibrate for Oxygen & Nitrogen and then use it as a multiplier ( or dilution factor)or what ever your name is for it, to calculate the Hydrogen %.

My question to you is: Who is your blader supllier ? i'll be thankful for their address
Arie

[chromatographer1]

AVRC,

I assume your balance is hydrogen, and you are measuring C1-C5 hydrocarbons. I assume your sample loop is 1mL.

You run a sample. You have calibration values for Hydrogen, C1-C5, and air from a certified standard mix.

You calculate your air peak and it gives you a value of 2 mole % (vol %).

You measure your hydrocarbons and they each measure 980 ppm or 0.098 mole % based on a 1mL sample (0.98µL), but the actual sample is 0.98mL gas and 0.02mL air.

You deduct your air peak from the volume of your sample.

You recalculate your mole % values for the hydrocarbons.

Your hydrocarbon peaks are each actually 1000ppm or 0.1 mole %, 0.98µL/0.98mL.

Correction for non-ideal gases would make it more accurate but this is close enough, no? ...... yes?

Rod

[avrc]

Hi arie,
I was just answering Rod's question whether air is meant to be present in my original sample or if it gets into the sample accidentally.
Arie, are you serious in knowing where to get bladders? then mail me at avrcgc@yahoo.com.

Hi Rod,
I think you are asking me to normalise other components to hundred percent leaving out O2 and N2. Am I right?

Can I just avoid N2 and O2 peaks by switching off integration while these two are getting detected?

Thanks for your sugestions.

[chromatographer1]

Yes, that is one way of looking at the issue.

But if you use external stds your calibration will not be applicable.

For example, if you had 50% hydrogen and 50% air in your sample (from handling errors) and you calibrated using an external std @ 100% then your report would indicate that your sample was not 100% hydrogen but 50% even though the sample was 100% before contamination.

Since you have to account for different gas TCD response factors, I wish to repeat it is more accurate to measure the impurities based on an accurate external std and then calculate your balance by difference.

You are welcome to call me directly if you have additional questions or if I am not clear.

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

814-359-5737 voice
814-359-5459 fax
rgeorge@sial.com

[Les]

I deal with the same problems with samples. However, I don't remove the air from the sample and assume the make-up as Hydrogen. I only report what was presented in the sample. The reasoning : What if the operating unit actually had problems and air was present?
I allow the unit supervisor deal with the sampler. If the supervisor has no problem removing the air, then that decision is his, not mine.

[Les]

The bags we use are polypropyl gas bags with a hose and pinch clamp. The bags are evacuated with a vacuum pump and returned to the operating unit. The operators are to fill the bags and press out by hand a couple of times. Then fill the bag and pinch off the hose. I attach the bags to the sample line and actually sit on the bag in a chair for an extended purge time. Then allow to equalibrate and inject the sample loop. Most of the time have no problems with air.