How Do I Leak-Check the Gas Delivery System?

I use three methods to check a gas delivery system for leaks. First, determine if there are any gross leaks by performing a pressure-drop test. Pressurize the gas lines and operate at normal flows at the instrument for a few minutes. Then turn off the flows at the instrument and mark the position of the high-pressure gauge needle by attaching a sticky note to the gauge face with its edge aligned to the needle. Next, turn off the high-pressure tank valve. The high-pressure gauge should remain steady. If it does not then there is a serious leak somewhere. If the gauge appears steady, then wait 5 min. Now, while watching the high-pressure gauge needle in relation to the sticky note, turn the high-pressure valve back on. The gauge needle should remain steady. If it moves upward slightly then there is a low-level leak somewhere.

Even if no apparent leak is indicated by the pressure-drop gauge test, use an electronic leak detector as a second way to verify critical gas lines such as carrier or make-up gas, as well as any hydrogen lines. The pressure-drop test usually will suffice for air lines, as an electronic leak detector does not work for air. If a leak is indicated for an air line, use a few drops of pure liquid water — not soap solution — to try and find it. If you are really concerned, it is possible to pressure the line with helium for the purposes of a leak-check with the electronic meter.

This question is closely related to the adoption of hydrogen carrier gas. So many have asked about it, however, that I'm including some responses here.

Gas generators provide a high level of convenience over cylinders, but at a higher initial cost; they are an excellent way to replace gas tanks in laboratories that consume them frequently. How often is "frequently"? That depends on the trade-offs between the cost of the individual tanks taken against the cost and maintenance of a gas generator over its useful lifetime. The advantages of gas generators also depend strongly on the real safety and convenience advantages gained from having fewer heavy, high-pressure cylinders to manage. In some situations this advantage alone makes changing to gas generators an easy decision. The following examples are based on full retail prices for gas generators taken from one chromatography supplier's web site in November 2012, and while the numbers are certainly representative, it would be crucial to work with a supplier to arrive at realistic figures for a specific laboratory.

Suppose a 49-L tank of high-purity helium costs $400 in quantity under a supplier contract — if you can find some, of course. A GC carrier-grade hydrogen generator with 500-mL/min capacity might cost $10,000–$15,000. Each full helium cylinder of this size contains around 8 m³ of gas, which would last for about 11 days at a 500-mL/min delivery rate to two or three GC systems with split–splitless inlets. In one year, this laboratory would consume around 32 cylinders of carrier gas for these GC systems, at a cost of $12,800, plus any extra demurrage for keeping spare cylinders on-hand. A more accurate return on investment (ROI) calculation can be obtained by considering the cost of the funds to purchase the generator and the tax advantages its depreciation would produce, but it's clear from the above quick estimate that the hydrogen gas generator saves its cost in helium cylinders within a year or so. Of course, this assumes that the laboratory can replace expensive helium carrier gas with hydrogen within the scope of their operating procedures and methodology. There is a cost of changing to hydrogen because of translating and validating methods, but in most cases this is not too difficult. The good news is that lots of support is available for this one-time activity.

What about a zero-grade air generator — really an air purifier — for flame ionization detectors? This device actually consists of two components: a zero-air purifier and an air compressor to supply it. Together, they might cost around $4000 for a 1-L/min system that could supply two flame ionization detectors. Zero-grade air tanks might cost about $75 a piece, and such a tank would last for about a week at 800 mL/min. So, in a year the cost of the 52 air tanks consumed would be around $3900. Again, ignoring the finer points of an ROI calculation, it is clear that replacing air tanks with an air generator for FID consumption would pay for itself in about one year.

Gas generators do have recurring costs for filter replacements, but this is minor compared to the savings that will result over a 5-year service period for split–splitless carrier gas and FID air generators, not to mention the labor savings and safety gains from not having to haul nearly 100 tanks into and out of the laboratory in a years' time.

When might it not be a good idea to install a gas generator? Consider replacing only the hydrogen tanks used for FID while not switching to hydrogen carrier gas. Around 45 mL/min hydrogen is needed for each flame ionization detector, so the time for a single detector to consume a full hydrogen tank would be about 125 days: that's just three cylinders per year. A hydrogen cylinder for FID use is less than $100, or about $300 per year per detector. By comparison, a smaller capacity 165-mL/min hydrogen generator, which is suitable for three flame ionization detectors, costs about $8000. The three detectors would consume a total of nine cylinders in a year, at around $900 cost, so the ROI period is something on the order of 10 years.

Clearly, the payback on an FID hydrogen-only configuration is not as attractive as the initial gain when moving directly to hydrogen carrier-gas generation from helium cylinders. What about going to hydrogen but without a generator — on cylinders only? Hydrogen cylinders to replace the 32 carrier-gas cylinders in the first example above would cost around $3200, so the net gain in a year compared to using helium tanks would be about $9600. That's close to the cost of a hydrogen generator, so it makes good sense to put the first year and some month's of this savings directly into a hydrogen generator, and save the intermediate transition from helium to hydrogen cylinders for an initial test exercise in method translation, if there are any questions about the switch over.