From pg 326:
I have updated the Wiki with surfingslovaks table of estimated relative rate of battery capacity loss based on Arrhenius equation for rate of chemical reactions that 10 degree Celsius increase in temperature doubles rate of battery capacity loss:wiki/index.php?title=Battery_Capacity_Loss#What_To_Do
Additions, corrections, etc. are welcome (especially from surfingslovak).
Arrhenius Factor: Exp(-(DeltaE)/kT) where
T is absolute temperature
DeltaE is activation energy.
I agree the Arrhenius factor is very relevant, but how fast it varies with temperature depends upon the activation energy of the chemical process that is causing our degradation. A higher activation energy reduces the absolute magnitude of the factor, but it increases the relative change in the factor for a given change in temperature. This makes sense, since we are dealing with a very slow chemical process.
60 F is 540 Rankin (absolute). A 40 F change in temperature (60 vs 100 F) represents only a 40/540 = 7% change in absolute temperature, yet we are seeing perhaps a 5 to 1 change in relative degradation rates for folks in different climates.
Your rule that "10 degree Celsius increase in temperature doubles rate of battery capacity loss" implies a certain activation energy. The wide disparity between degradation for folks on this forum suggests a higher activation energy might be closer.
The following published paper studies lithium-ion battery loss as a function of both temperature and SOC:
"Correlation of Arrhenius behaviors in power and capacity fades with cell impedance and heat generation in cylindrical lithium-ion cells"
from Sandia National Laboratories.http://126.96.36.199/ft/641/92454/1607538.pdf
This 2003 paper is obviously not talking about the LEAF's particular Lithium chemistry, but I believe the behaviors it describes are typical. Capacity fade is discussed on pg 7, Fig 5, which I display below:
The chart shows that capacity fade slows for all temperatures as the SOC is reduced from 10% to 80% to 60% SOC. At high state of charge the Li-ions are concentrated on the graphite electrode. My understanding is that the primary loss process takes place at this electrode, so it seems reasonable that this process would slow as the SOC lowers.