Weatherman said:
I'm not sure that's the appropriate direction to take, since it makes the heat-induced degradation rate in Honolulu very close to the rate in Phoenix.
I'd be more inclined to steepen the heat-induced curve by doubling every 5 or 8 degrees but assume "normal" is a lower percentage of the total loss per year. The cycling losses would, then, represent a larger percentage of the total loss and the sum total loss rate would be more linear.
I agree that it isn't really the right direction, but that is what I found. I don't think steepening the curve will help, because I couldn't make the numbers work even with the original factors--I had to make the battery aging factors somewhat lower in order to get the numbers to work properly. The ratios weren't too far off, since normalizing on "Normal" produced:
Phoenix - 1.5
Normal - 1.0
Boston - 0.83
That number is really kind of a side alley, since the model currently works quite well with numbers that are at least fairly reasonable.
As I said before, if anyone wants to play with the spreadsheet, just PM me. I have an almost perfect fit to Nissan's data (including TickTock). I am going to try adding a few other factors I think may be important (solar loading, percent of time spent at high SOC, perhaps DOD), then need data to calibrate it (and test it). Unfortunately, I no longer trust the Gid meter as a proxy for usable range. Only a range test will suffice, although going to VLBW and then extrapolating would certainly be acceptable. My next task is to set up the spreadsheet so I can just plug in some number and the predicted range will pop out for comparison with the measured range.someones measured range.