the cells never see more than 2.5DCV each, so just about 50% of all watts consumed goes towards actually making HHO. in a single cell BF electrolyser, only about 9% (1.24/13.8=.089) goes towards making HHO.
why 2.5DCV? well, as chemelec will tell you, him and i spoke a few times on the subject. working with partially submerged 2" x 4" anodes and cathodes, it became necessary to raise the voltage to get any kind of amperage flow. as voltage is raised, the electrical resistance of water actually changes.
so! all in all, my system consumes 47 watts, the cells themselves receive
if you were to install this system on your car, you would be drawing 3.4ampheres from the vehicles system. the two cells in series is pulling about seven ampheres. havent tested it on a digital ammeter. but my analog one is of good quality. pretty nifty eh?
in order to raise the amps, i'd need a larger surface area for the electrodes. the electrodes are already only a millimeter or two apart and my water and KOH solution is strong, one cell in series pulls over 25 ampheres easily.

a picture of my apparatus. messy i know, just finished assembly and started testing.

radioshack 12volt (13.8 measured), 3 amp power supply. hiccup style overcurrent protection cuts off at 5 amps.

12DCV input, 0-5DCV output, 25 amphere rated power supply. up to 94.3%effecient. this is the power supply that i'd posted about. its going through a all-night test tonight.

DC Ammeter, analog of course. shows approx 7-8 amps. this is the amps that the cells are drawing.

the two cells, four smacks plates each. that makes a total surface area of 64 square inches. optimum plate life would be (64*.25= 16amps), while pushing the cell a bit harder would be (64*.5=32 amps). more surface area is needed at such a low voltage to hit the higher amps.