This paper presents a novel battery degradation model based on
empirical data from the Racing Green Endurance project. Using the
rainflow-counting algorithm, battery charge and discharge data from
an electric vehicle has been studied in order to establish more
reliable and more accurate predictions for capacity and power fade
of automotive traction batteries than those currently
available.
It is shown that for the particular lithium-iron phosphate
(LiFePO₄) batteries, capacity fade is 5.8% after 87 cycles. After
3,000 cycles it is estimated to be 32%. Both capacity and power
fade strongly depend on cumulative energy throughput, maximum
C-rate as well as temperature.