edatoakrun wrote:To eliminate what is almost certainly the greatest variable in early capacity bar loss, climate, I'd suggest you might try graphing only the one or two-bar loss LEAFs reported from the Phoenix area.
And also, perhaps, graph only the first-hand reports from LEAF drivers, and only those reporting who received delivery, as opposed to orphans, demos, and pre-owned LEAFs, with much less certain use/abuse histories.
If there is still a low correlation, then other factors (amount of time spent a high rate of charge, for example) might prove more significant than time and miles driven.
Manufacture date would also probably be better than delivery date, but I realize you don't have that for most of the reports.
And total kWh use (which no one seems to be reporting) would probably be a better indicator than miles driven. A leaf that has driven a given number of miles at 3 m/kWh, has done about twice the battery cycling as one that has averaged 6 m/kWh.
Very good thoughts, but my time and energy (and the data) are limited. Besides, my Leaf is cooling its battery in my 70 degree parking garage where it spends most of its life, so I am not personally affected to any great degree (although the hot San Fernando Valley where the Leaf spends its time during work days is somewhat of a concern).
However, here is the latest data on one bar capacity loss in the Phoenix area. Plot is annual mileage vs. months to first capacity bar loss. There are 26 data points.
Slope = -1,274
Intercept = 29,095
Correlation coefficient = 0.51
Edit: mean annual mileage = 13,894 and standard deviation of annual mileage = 3,967 for Phoenix area.
Interpretation: There is a moderate correlation between annual mileage and number of months to first capacity bar loss. This data suggests that some of the capacity loss may indeed be due to cycling losses rather than calendar losses.
Note: when I did the graph previously in another thread
, I did not plot against annual mileage,which is the correct figure. I believe the current analysis is the correct one, but comments are welcome re: whether the Y-axis should be total miles (previously plotted) or annual mileage (the current, and probably the correct choice).