I found this off of Physics.SE, and I'd be genuinely interested in what others here think.
But the consistent advice I've seen is to do just what @dmtaylor suggests: get to know your own boil kettle and tweak your recipes/plans accordingly.
The two biggest factors contributing two boiloff rate that I could find are the diameter of the kettle (hence surface area exposed to the air) and the amount of heat transfer going into the kettle (boil vigor). In the case of my super-simple stovetop setup, this computes to how "high" or "low" I have the dial set.
So I conducted trials. The constants in every trial I ran:
- Same 16-quart brew kettle used each time
- Same burner (I have different sized burners on my stove) used each time
- Lid placed in the same 7/8 position each time
- Dial turned to the MAX setting which produces a violent boil
The only thing that changed was the volume of water I started with each time. In every case I hit more or less the same rate: 0.68 gallons per hour (GPH). There was some variation (which could have been chalked up to a million inconsistencies in my non-laboratory, kitchen-based "experiments"). But 0.68 GPH was the average. For that kettle, under those conditions.
I will also do this for whatever dial setting I need to maintain a smooth, rolling boil as well as a vigorous boil (there are differences in all of these, apparently). All different amounts of heat transfer.
That way, if I know I need a rolling boil, a vigorous boil, or a violent boil in that 16-quart kettle, on that burner, and with the lid placed at the 7/8 position, I will know what setting to place the dial at and I will know the boiloff rate in GPH to expect.