Purging is important when dealing with hydrogen systems because hydrogen can ignite in the presence of relatively little air. A mixture of H2 and only 23% air (5% oxygen) can ignite, as opposed to (for example) propane which requires a minimum of 89% air to ignite. Also, the friction caused by gas being forced at high pressure through a valve could theoretically create enough heat to ignite an appropriate H2 / air mixture. It is thus vital that no air be allowed in any piping or storage tank containing hydrogen.

Happily, it is a simple matter to get rid of the air. The technique is one of repeated dilution. Suppose you have a tank full of ambient air at 1 atmosphere (≈ 15 psi). You add, say 5 atm nitrogen (bringing the tank up to ≈80 psi). The concentration of air is now 17%. You bleed this gas out of the tank until it is back to just over 1 atm. Repeat the process twice more, and the concentration of air is down to 17% of 17% of 17%, or 0.5% air. Obviously, during bleeding the tank pressure must stay high enough (>1 atm) to prevent air re-entering. Three purges are desirable, to add a margin of safety in case one purge goes awry.

You must keep a tank of N2 on hand and make careful purging a habit, always. This is easy to do. It is clearly desirable to isolate the main storage tank from all pipe lines with valves, because while purging a line is a trivial matter, purging a large tank takes a lot of K-cylinders of N2. Also include a bleed valve and a gas inlet (eg compression fitting and valve) at opposite ends of each run of pipe.

We found that residual N2 in the tank (from purging) slightly diluted the H2 concentration in the fuel cell feed, lowering output proportionally.


Left: Stephen at work with a nitrogen K-cylinder.
Right:N2 tank with pressure regulator valve. The H2 K-cylinders have a separate but similar valve.
    All website materials © SIEI. We accept no liability for any use you make of this website and information.