The question centers around how to read the capacity loss graph on page 7
At first glance, the 75-65 cycle seems to be the best for the battery, but you need to normalize it for work done: a 50% depth of discharge cycle does 5 times more work, 1000 cycles of 10% is 100 full cycles. This is labeled as work done in the table below.
SoD / EoD: start and end of discharge.
DoD: depth of discharge, SoD - EoD.
Cap % at 4k: capacity lost at 4000 DST cycles.
Work done: 4000 * DoD / 100%.
Cap loss per 1k work done: capacity loss per 1000 full cycles. 1000 cycles is about the full life time of a phone, about 3 years of 1 full charge a day.
SoD | EoD | DoD | Cap % at 4k | Cap loss | Work done | Cap loss per 1k work |
100 | 25 | 75 | 81.5 | 18.5 | 3000 | 6.2 |
100 | 40 | 60 | 83.5 | 16.5 | 2400 | 6.9 |
85 | 25 | 60 | 85 | 15 | 2400 | 6.3 |
100 | 50 | 50 | 85.5 | 14.5 | 2000 | 7.3 |
75 | 25 | 50 | 87 | 13 | 2000 | 6.5 |
75 | 45 | 30 | 89.5 | 10.5 | 1200 | 8.8 |
75 | 65 | 10 | 93 | 7 | 400 | 17.5 |
As you can see, in this model, the loss is with the 100-25 cycle, or alternatively, the 75-25 cycle. It is unfortunate this information is so hidden in this paper.
Also, note that this test was done with a 3.3V nominal voltage cell, which is not a standard Li-ion cell, that chemistry is 3.7/3.8V, so findings do not always transfer.
References:
Xu, B., Oudalov, A., Ulbig, A., Andersson, G., & Kirschen, D. S. (2018). Modeling of Lithium-Ion Battery Degradation for Cell Life Assessment. IEEE Transactions on Smart Grid, 9(2), 1131-1140. doi:10.1109/tsg.2016.2578950
Comments
2 comments
I am impressed at how well you guys know your stuff. Most app developers have no idea about the tech or science behind thier products. I am very impressed with your knowledge level. Thanks.
Hi there,
thanks for taking your work seriously and data-driven.
I particularly like the normalization vs. work done you guys do.
Could you please clarify on this statement?
"As you can see, in this model, the loss is with the 100-25 cycle, or alternatively, the 75-25 cycle. It is unfortunate this information is so hidden in this paper."
As I understand, the lowest "Cap loss per 1k work", the more work you can do without loosing capacity. Why is the best case not the one with the lowest value in this ratio?
Thanks!
Article is closed for comments.