We chose a pair of Hogan-GC electrolyzers made by Proton Energy. These are small PEM (proton exchange membrane) desk-top units, designed to make hydrogen in the lab.

The electrolyzers’ output is small, about 0.6 liters/minute, and they draw a lot of power, about 600 watts AC each. Their rating is 1200 w, but they seem to draw less. Their poor efficiency works against them in this job. However, the next size up we could find, eg the Hogan-200, was too large and too expensive for our project. We could not find a product sized just right for a residential installation.

We also chose the Hogan-GC’s precisely because we could afford two of them. We wanted redundancy in case of breakdown, and we wanted to be able to hack the units without shutting everything down. As it turned out, one of the units did fail (Proton replaced it). And we needed to modify the units for remote operation, which we have been doing one unit at a time. So for our purposes getting two units was a good move.

However, since the efficiency of the Hogan-GC’s has been about 18% (see Efficiency ), they are not the tool of choice. Other units could do better, since electrolyzers have an internal economy of scale, max-ing out at about 30-40% efficiency .Proton’s Hogan-HE series, for example, perhaps attains this efficiency. An electrolyzer which ran off DC power would be better, especially for remote operation. The Hogan GC’s also have a one hour warm-up period, in which they draw power but produce no hydrogen. This is a nuisance.

We needed to hack into the Hogan-GC’s in order to set them up for remote operation. Because our electrolyzer output is so low, our system depends on using solar power to make hydrogen when no-one is about (since at these times all power is surplus, and sunshine is free). In absentia operation makes up for the inefficiencies of the electrolyzer-fuel cell loop by increasing access to the solar power.

One electrolyzer bug we have not yet solved: The Hogan-GC’s have a pressure sensor of some kind which is used to detect leaks. For some reason, sending hydrogen to a large tank triggers this sensor. The electrolyzer then flashes a “leak” warning and starts beeping. We have confirmed that there is no leak, but we have not figured out how to prevent whatever pressure fluctuation the sensor is responding to. Adding a pressure regulator downstream (just visible in picture at top of page) did not help.

The electrolyzer table. On the far left is a water deionizer (see Balance of Plant).
An electrolyzer modified for remote operation. In the lower left, wires from voltage-activated relays turn the machine on and off, bypassing the push-button controls.