an average of .424 GPH or 63.1 ounces per hour! this is a lot less than your first hour run in sounds like the generator is a lot better than expected or your loads have not been nearly as high as the run in load
Frank - one note on this - my initial fuel consumption test was a constant load to see what peak fuel/kWh efficiency would be under ideal loading conditions. Real world is always different. I did another test in real world with battery charging (which is more controlled conditions than load run) and got .44 gal/hr at an average 3.04 kVA output. So the real world load run fuel consumption is still very close to what I got with battery charging during my one test. The generator has not run for battery charging since that first test I did, so I'm confident that it will continue to be more efficient than the propane generator was by about 3%.
This is mainly due to the higher thermal efficiency of the gasoline fuel and the engine that burns it. An LPG engine would have to run a compression ratio of around 12-13:1 to match it. The Honda's iGX electronically controlled engine is obviously a masterpiece of engineering in small engines. If you read about what Honda did with the thing, they made it so the engine's ECU communicates with the machine it's powering , whether it be a generator or a power trowel on a construction site. On the generator application the CT (Current Transformer) sensor adjusts throttle opening and ignition timing before the load can affect engine rpm. And it is VERY fast - there is no throttle linkage on it because the ECU is integral with the carburator. So there is no "play" or "droop" that you would get even with conventional electronic governors.
Honda's system is strikingly similar for gasoline engines to what we developed for the Cummins N14 called the Celect Plus system. On the Cummins Celect Plus engine management system the throttle pedal merely operates a potentiometer and injectors are directly controlled by the ECU with no mechanical linkages or interaction. This allows very precise metering of fuel and injection timing - and the Honda's technologically advanced engine is doing the same thing with a gasoline unit.
I think this generator is more of a showcase of Honda's engineering prowess than anything else. And believe me, when it comes to engineering, few companies on earth can match Honda.
When comparing LPG fuel to gasoline, you're going to burn at least 50% more fuel (gal/hr) with LPG than you will with gasoline just because the BTU content of LPG is so low compared to gasoline. And then you look at volumetric efficiency, and the LPG gets worse. Because of the low air/fuel ratio with LPG there's not as much room for oxygen molecules in the combustion chamber.
So the pumping losses in the four-stroke cycle are going to be higher with LPG, along with poor cylinder scavenging because of the volume of fuel that has to re-introduced to the cylinder on the intake stroke. You can try to reduce the amount of raw fuel lost out the exhaust valve by reducing overlap on the valve timing events. But then you trap some exhaust gas (poor cylinder scavenging) in the combustion chamber with the fresh charge - and once again you get reduced thermal efficiency because of a diluted charge.
There is basically no way to make a spark ignition engine as efficient on LPG as it can be made on gasoline because of the nature of the fuel, and the basic parameters that the four-stroke cycle uses to scavenge the cylinder and recharge it. So people that put LPG conversions on gasoline engines are shooting themselves in the foot.
--
Chris