It uses this Component layout:
Probably not the best component layout around, certainly open to adjusting it, but I think it's efficiency per rod is good, even before gas conversion. ( I heart biogas )
Sugar is grown up top, managed by cropmatron and harvestor, pump from ocean feeds cropmatron's hydration and canning machine. All equipment powered by 1 semifluid generator, which gets fed by the machine before export of gas. Harvestor>Macerator>Solid/Fluid Canning, then fed into fermenters. There are 18 fermenters connected to 18 Liquid heat exchangers. This turned out to be overkill, there's simply not enough hot coolant to go around.
All Fertilizer is fed back to the sugar patch up top. I've got 53 Semifluid Generators to expend the gas produced. That seems about the right number.
Confusingly the reactor display says 1056hu/s, that can't be right, it must mean HU/t? I don't think it could be supply so much otherwise.
40,000 heat to make a bucket of biogas
= 1 bucket of biofuel every 37.87 ticks
32,000 EU per bucket of biogas
32,000 / 37.87 = ~845, but my numbers are a little better than that, not including feeding the separate generator that runs the system.
Also the rods seem to be lasting a VERY long time, I think much longer than the 2 hours and 46 minutes projected by the Java app.
I don't know how much bounce per ounce you get from steam. But I like biogas, and I think this is a pretty cool set up :). I swap the rods out once in a great while, otherwise it needs no intervention from me.
I suppose someone could convey energy losslessly over great distances this way, though they'd have to grow a lot of cactus first.
I'd love to hear any input.