Problem in GC/MS: unusual He consumption & no leaks detected

[lmarsan4]

Dear all,



We've been experiencing some problems with our GC/MS system (GC Agilent 6890N and MS Agilent 5973), coupled to a double-shot pyrolyser (model 2020i; Frontier Laboratories). Under normal conditions and regular use, the He bottle used to last between 6 and 12 months, whilst ca. 5 months ago it only lasts 3 months or even less. We all are conscious of how costly ordering frequently a He bottle is. On November I asked a related question in this forum.

Once we detected that the He bottle was depleting to an unnormal rate, we started to replace and adjust any consumable that could affect this. Furthermore, it was observed how the needle of the pressure gauge slightly fluctuates while the device is on. This problem was solved when we changed the Cu tubing from the bottle to the GC, solving the welding problem, but still, the rate at which He bottle drops out is not usual by far.



Here's a list of the replaced consumables:

-Pyrolyzer:

*O-ring for a simple base of Py-2020iD

*Interface needle

*Quartz tube y corresponding o-ring and washer.

*Graphite vespel ferrule

*Interface Union needle



-GC:

* Cu tubing that goes from the He bottle to the GC, along with the SWAGELOK 1/8” fittings.

* DB-5 GC column was cut in both extremes, inlet and MS line zones, in case there was any obstruction. At the same time, fittings and correspondent graphite ferrules were replaced.

* Liner Splitless: Gold Plated Inlet Seal with Washer, injector line septum, and Cu tubing from Splitless.



Ultimately, we tried two internal leak tests according to the manual, and both resulted negative. Everything seems to be right but the abnormal He consumption keeps going.



Can anyone that experienced a similar problem suggest me where could be the problem? I read in some forums that it could be a problem related to the EPC; if so, then what we could do?



Thanks in advance and kind regards,



Layla

[Consumer Products Guy]

In the automotive world, they call this "throwing parts at it", just hoping by chance to find/fix the leak. Get an electronic leak detector and find the source of the leak, and don't overlook the helium regulator and its connections too.

I doubt that the MSD is the issue; concentrate on the plumbing and the GC.

[uzman]

If you can not find any leak outside the GC , you may concentrate on EPC.

I faced an EPC problem on a Thermo equipment , which was giving 25 times more gas to the column , because of malfunction.
I locate the problem by measuring the gas directly at the outlet of the column.

[James_Ball]

If you have a flow meter, check the flow from the septum purge and split outlets on top of the GC. If there is a problem in the EPC it could be allowing too much flow to pass through but not reporting it on the display.

[Steve Reimer]

If you are looking for another common leak location, check the connection at the tank. The nipples are easy enough to replace, not too expensive and are easy to damage if there is foreign material between the sealing surfaces.

[HawkEye_Analytical_1]

Simply perform a pressure decay test, by turning the instrument flows down close to off. Only maintain column flow at a low rate, suitable for the MS. (1.0Ml/min should be fine)
Make sure split flow / purge flows are minimal.

Turn the tank off, and watch the pressure of the gauge. If it drops rapidly, you may have a leak in your plumbing as suggested above. 50/50 IPA/Water make a good bubble producing leak test liquid, as it will not leave a residue. This can be applied to the plumbing connections, taking care not to touch electronics.

Barring any plumbing leaks, the split trap is a common location for gas leaks on a 6890. Check the trap located at the left rear of the 6890 for the inlet in question. This will be a cartridge style and pencil style trap. The cartridge may need to have the housing tightened. This is of course after the septa, o-ring liner and gold seal have been inspected.

[Peter Apps]

Do not use liquids to check leaks - if there is a leak you will contaminate your plumbing. Get yourself a proper leak seeker.

Peter

[LALman]

I second getting an actual leak detector. I've saved the cost of the detector in extended lifetime of He cylinders. Also, I suggest checking the three o-rings where the weldment interfaces to the EPC. I was having an air problem on my 6890N/5973inert recently and that was where the air came in, and He was escaping.

[LALman]

I've also talked to my tank supplier about rough surfaces on the UHP tank fittings. I showed them how they leaked right there at the tank fitting no matter how I or they tried to seal it. They now check the tanks before they bring me one. I suppose someone wanting to use He for welding is not pick about loosing a little He over a week. But my setup goes nearly 3 months to run around 180 samples on a single tank of He.

Prior to getting a leak detector, some cylinders were lasting as little as a month.

[James_Ball]

I've also talked to my tank supplier about rough surfaces on the UHP tank fittings. I showed them how they leaked right there at the tank fitting no matter how I or they tried to seal it. They now check the tanks before they bring me one. I suppose someone wanting to use He for welding is not pick about loosing a little He over a week. But my setup goes nearly 3 months to run around 180 samples on a single tank of He.

Prior to getting a leak detector, some cylinders were lasting as little as a month.

We use about two cylinders per week, but at one point we were going through almost one per day. Found it was leaking at the flexible line from tank to the switching regulator in the tank farm. Good thing it was outside or we would have all sounded like Mickey Mouse

As mentioned above, a digital leak detector has helped me find many leaks I could not find even using a bubble solution. Best bubble solution to use in a pinch is just plain old Windex or any spray window cleaner, just be sure to wipe it all away quickly if you see bubbles or it can actually enter the system at the leak. A little counter to what one would think, but even with gas pushing out the leak, other things get drawn in as well.

[Izabela]

Dear lmarsan4,

did you solve the problem? I have the same Problem with the unusual He-consumption and I can't find the reason. I didn't find any leak (I have Restek Leck-Detector and Leck-Spray).
Is it possible, that it's something alse than Leck?
Best Regards,
Izabela

[tlahren]

Another possible source is if you have two injectors/columns installed, but are only using one. Or other detectors. On Agilent Chemstation it is easy for some other user to load their method and the software will make changes to the system (eg., turning on gas flows to other inlets or detectors). This happened to me on a SSL inlet that was not being used. A different method had turned the septum flow on and leaked out a bunch of He without making any errors. This was more prominent on a Thermo system but my 6890s can do it too if it is configured with other detectors or inlets that you are not currently using.