I ran 10-gas mix instead of Helium in carrier gas. Damage?

[john.gardner]

My Daniel 500 gas chromatograph developed a leak, draining a cylinder of Helium from full to empty in two days. I don't have a Helium sniffer but I do have an LEL sniffer, so I plumbed my 10-gas calibration cylinder to the carrier gas input and pressurized the GC with that for a while. (10-gas has the same components as natural gas, but with elevated quantities of some of the lower-percentage components to ensure those peaks actually appear). I ran the 10-gas out the measurement vent to make sure the whole unit had that gas in it instead of Helium so I could sniff out the leak, then I capped the measure vent and gas flowed into the unit and out the leak. Analysis was halted for this effort. I eventually found the leak, fixed it, sniffed again, and put the Helium and 10-gas back where they belong. When I started it back up I obviously didn't have good readings, the detector was very noisy, I was getting preamp alarms, etc. I left it running on the cal stream overnight hoping to flush all the 10-gas out of the carrier stream. This morning the noise is gone, but the peaks are wrong, and no peaks at all in the second half of the run. We are now more than 19 hours later, and no good readings.

I have a case open with Emerson, waiting to hear back, but I wanted to ask the group: Is it possible that I poisoned the columns, or otherwise damaged the GC by flooding it with natural gas (more or lesss) where it expects to have carrier gas and only minute quantities of hydrocarbons?

Here is what a chromatogram used to look like:

Chromatogram Before by gardnertoo, on Flickr

And what it looks like now:

Chromatogram After by gardnertoo, on Flickr

[rb6banjo]

What is your detector? What is your column? Is it possible that the flow through the column is too fast (head pressure too large)?

[john.gardner]

Detector is a pair of thermistors, one measuring the temperature of the carrier gas, the other measuring after the columns.

The instrument has four columns. All are 1/16" tubing, packed media. The two long ones, seven footers, are labelled "silcoport P", don't remember what the other two are. All four columns are coiled up in an oven.

Flow through the columns is set in the center of the range specified by the manual.

[chromatographer1]

It looks like the porous polymer column (#4) pores are partially blocked with pentanes and hexane plus components.

Your propane peak is 17 seconds late and the Isobutane peak is over 20 seconds late in elution time.

You will have to remove the last PP column and replace it, or condition it in another oven @ 150C for several or more hours.

The easiest quickest solution might be to replace the column set as a whole and reset your timed events.

Good luck,

Rod

PS yes, I was a chromatography method engineer for Daniel Industries in the 1990s.

[john.gardner]

Rod, thank you for that, I will run it by the Emerson engineer who is helping me troubleshoot the problem.