Estimating Battery Degradation

[mpulsiv]

I vaguely remember reading that max regen on the e-Golf is around 40 kW. While the high regen found in B mode will send more 40 kW spikes into the pack, I don't believe it would generate the same amount of heat from DCFC charging where 43 kW (at a 125 amp DCFC station) is sustained continuously for 20 to 30 minutes. Heat is what speeds up the pack degradation.

VW warrants the NET capacity. Based on a European e-Golf manual I got my hands on, it says the starting net capacity is 64 Amp hours. AT 323 V nominal pack voltage, that translates into 20.7 kWh starting net capacity. Therefore, 70% of 20.7 is 14.5 kWh. But I could be wrong about the correct starting net capacity.

I suspect that the "reserve" capacity is automatically activated by the battery management software once the health of the battery degrades to a certain level. Of course, once that happens, the battery health will then decline even faster than before.

Based on my net capacity measurements, my car has about 19 kWh of usable capacity (about 10% loss), after 24,000 miles and almost three years. I'm not sure the warranty will be triggered but since it's getting hotter in the SF Bay Area, I think the battery will be taking a beating over the next several summers, and the capacity may drop faster than it already has.

Correct me if I'm wrong, regenerative braking, specifically aggressive B mode could heat up the battery?

 

[f1geek]

B mode will heat up the pack more than D, but it is probably a small amount. If you are in stop and go traffic, B will probably heat just about the same as D because pressing the brake pedal provides regen until the required pedal force exceeds the stopping power provided by regen. If you use B mode from high speeds, then B will heat up the pack more than coasting in D, but again, if you need to slow down quickly, you are using regen via the brake pedal. I really don’t believe B vs D will add much more heat to the pack. The best way to heat the battery pack (other than soaking the car in desert like ambient temps) is sustained high speed driving or DCFC.

 

[mpulsiv]

Can someone chime in about duration of charge vs. frequency?
Doesn't the battery take a toll with L1 charging pattern? Needless to day L1 is ~12% less efficient.

30% - 90% SoC (~100 miles range)
L1 = ~ 22 hours
L2 = ~ 4 hours

I can get by with L1 charger but would love to know the long term consequence when it comes to battery health.

 

[f1geek]

Very high temps, high temps and high SoC, and very low SoC are the main ways to degrade the cells. L1 charging generates very little heat (joule hearing varies with the square of the current), so is better than L2 (typically double or triple the current as compared to L1) for battery health since less heat generation means lower pack temps for a given charge duration. That being said, unless the pack is already hot prior to charging (mostly due to desert ambient temps) then I don’t think L2 charging will cause any measurable degradation.

 

[mpulsiv]

Very high temps, high temps and high SoC, and very low SoC are the main ways to degrade the cells. L1 charging generates very little heat (joule hearing varies with the square of the current), so is better than L2 (typically double or triple the current as compared to L1) for battery health since less heat generation means lower pack temps for a given charge duration. That being said, unless the pack is already hot prior to charging (mostly due to desert ambient temps) then I don’t think L2 charging will cause any measurable degradation.

High/low SoC and temperatures totally makes sense. Yes, L2 generates more heat but this car built for it. Much of Europe is pushing 220v. My concern is extended runtime with L1. Long story short, our batteries are being used 4x more with L1. Total charge runtime in 5 years will be substantial.

 

[f1geek]

I don’t know the heat capacity of the e-Golf pack, but based on the manual’s directions about low power charging, I bet the pack has a high heat capacity. Combined with the rate at which the pack loses heat (due to pack shape/exposure to air that provides some small convective heat loss at a standstill and conductive loses to the car frame and air) , L1 heat input will not lead to high temperatures even if left on charge for 24 hours. But, the only way to know for sure is to hook up OBDELEVEN and record temps over time. Or get ahold of a VW engineer who can give you test reports.

It is all about temperature, and if VW did their homework, which for the e-Golf, I think they did, the rate of heat loss during L1 charging is near the rate of heat gain due to joule heating and thus the rate of temperature increase is very slow so that a long charge on L1 doesn’t lead to dangerously high temps. Again, to verify you will need to run an experiment. I don’t believe you need to worry.

 

[mpulsiv]

I don’t know the heat capacity of the e-Golf pack, but based on the manual’s directions about low power charging, I bet the pack has a high heat capacity. Combined with the rate at which the pack loses heat (due to pack shape/exposure to air that provides some small convective heat loss at a standstill and conductive loses to the car frame and air) , L1 heat input will not lead to high temperatures even if left on charge for 24 hours. But, the only way to know for sure is to hook up OBDELEVEN and record temps over time. Or get ahold of a VW engineer who can give you test reports.

It is all about temperature, and if VW did their homework, which for the e-Golf, I think they did, the rate of heat loss during L1 charging is near the rate of heat gain due to joule heating and thus the rate of temperature increase is very slow so that a long charge on L1 doesn’t lead to dangerously high temps. Again, to verify you will need to run an experiment. I don’t believe you need to worry.

It's inevitable that L2 will generates more heat. It's common sense to avoid charging with L2 when ambient temperature hover over 80 degrees. Back to curiosity about extended runtime with L1 vs. short runtime with L2. I have a reason to believe that L1 extended runtime will take a toll. After all, the battery is being used all 4x longer vs. sitting idle with L2 utilization.
Hypothetical scenario over the course of 12 months:
L1 = ~700 hours charge cycle runtime
L2 = ~150 hours charge cycle runtime

 

[f1geek]

Cycle number, not cycle time is a factor in degrading a battery as there are a finite number of cycles you can expose a cell before you see degradation (depending on temperature and starting/ending SoC, of course). If there is no published correlation with cycle duration and battery degradation, then there is no need to worry. I don't see an issue with long charge duration based on what I know about lithium ion battery science. If you can find out someone who was able to show such a link, please let us know.

 

[mpulsiv]

Good to know

Sharing relevant links:
30-60% is much healthier for the batteries over the long-term -> https://electrek.co/2018/05/04/are-you-killing-your-lithium-batteries
https://batterybro.com/blogs/18650-wholesale-battery-reviews/103090502-why-do-lithium-ion-batteries-die-long

https://www.prba.org/wp-content/uploads/Exponent_Report_for_NFPA_-_20111.pdf
https://batteryuniversity.com/index.php/learn/article/types_of_lithium_ion
www.plugincars.com/eight-tips-extend-battery-life-your-electric-car-107938.html
https://batteryuniversity.com/index.php/learn/article/how_to_prolong_lithium_based_batteries

https://pushevs.com/2018/04/27/battery-charging-full-versus-partial
Informative chart here to show longevity
Cycling from 100 to 0 % we get 500 cycles
Cycling from 100 to 10 % we get 500 cycles
Cycling from 100 to 20 % we get 1.000 cycles

Cycling from 90 to 0 % we get 1.500 cycles
Cycling from 90 to 10 % we get 1.500 cycles
Cycling from 90 to 20 % we get 2.000 cycles

Cycling from 80 to 0 % we get 3.000 cycles
Cycling from 80 to 10 % we get 3.000 cycles
Cycling from 80 to 20 % we get 3.500 cycles

Cycling from 70 to 0 % we get 5.000 cycles
Cycling from 70 to 10 % we get 5.500 cycles
Cycling from 70 to 20 % we get 6.000 cycles

 

[mpulsiv]

I spent hours on https://insideevs.com, https://electrek.co and https://batteryuniversity.com/learn/article/how_to_prolong_lithium_based_batteries
Some sources pointed to Jeff Dahn, a world class expert on Tesla's battery tech.

The takeaways:
From chemistry perspective, closest to 50% is still optimal. The further they are charged above 50%, and the further they are discharged below 50%, the worse it is for the batteries. Lower SoC = lower voltage = benefit of increased # of charging cycles.
Charging from 20% -> 100% yields 1.000 charge cycles
Charging from 20% -> 50% yields over 12.000 charge cycles

Source: https://www.myvwegolf.com/forum/viewtopic.php?f=11&t=1507&start=30
@f1geek: correct me if I'm wrong.
100% SoC = 4.07V
80% SoC = 3.94V
70% SoC = 3.87V

Don't wait until your battery is depleted to 30%. Lucky drivers with short commute could benefit short and frequent charges.
If you only deplete 10% per day then charge to 50% daily. If your commute is longer (e.g. 20% per day) charge up to 60% daily. Remember, ~50% is least stressful state that increases longevity.

Some folks argue that we could face a capacity loss from unbalanced cells. Based on my slew of data and driver's feedback, uneven cell wear is a myth. Those who are living in fear, can certainly charge to 100% to let the batteries re-balance itself.

This is what I have gathered and would like to hear more feedback how to prolong our batteries Monday - Friday for short commutes

 

[f1geek]

Here is the data I got from my car about a year ago:

100% SoC - 4.07 V
80% SoC - 3.94 V
70% SoC - 3.87 V

Thanks for posting this info. Good advice. I’m not sure I’ll keep my car for 12,000 cycles (33 years assuming one charge per day), but someone else will be able to get use out of the pack.

 

[srue]

Agreed. mpulsiv, everything you're saying fits with my understanding. I believe this is the best practice for maintaining battery life.

I will note, however, that the difference between this best practice and a less cautious 70% average charge (for example) is likely insignificant. That said, some people really enjoy maximizing things like battery life and durability, so go for it.

 

[Newev]

Good to know

Sharing relevant links:
30-60% is much healthier for the batteries over the long-term -> https://electrek.co/2018/05/04/are-you-killing-your-lithium-batteries
https://batterybro.com/blogs/18650-wholesale-battery-reviews/103090502-why-do-lithium-ion-batteries-die-long

https://www.prba.org/wp-content/uploads/Exponent_Report_for_NFPA_-_20111.pdf
https://batteryuniversity.com/index.php/learn/article/types_of_lithium_ion
www.plugincars.com/eight-tips-extend-battery-life-your-electric-car-107938.html
https://batteryuniversity.com/index.php/learn/article/how_to_prolong_lithium_based_batteries

https://pushevs.com/2018/04/27/battery-charging-full-versus-partial
Informative chart here to show longevity
Cycling from 100 to 0 % we get 500 cycles
Cycling from 100 to 10 % we get 500 cycles
Cycling from 100 to 20 % we get 1.000 cycles

Cycling from 90 to 0 % we get 1.500 cycles
Cycling from 90 to 10 % we get 1.500 cycles
Cycling from 90 to 20 % we get 2.000 cycles

Cycling from 80 to 0 % we get 3.000 cycles
Cycling from 80 to 10 % we get 3.000 cycles
Cycling from 80 to 20 % we get 3.500 cycles

Cycling from 70 to 0 % we get 5.000 cycles
Cycling from 70 to 10 % we get 5.500 cycles
Cycling from 70 to 20 % we get 6.000 cycles

I don’t know if you (or others) are following this information any more, but I have a basic question. Very interesting reading in the links above leaves no doubt that limiting the charge extremes, especially at the top end, can greatly increase the battery life. One point that is not made clear is the definition of charge cycle. I know Apple defines a cycle as the charge and discharge of their battery completely; if you charge your phone from 50% to 100%, then use it down to 50% capacity, that’s only a half cycle. Battery University glossary says “ no standard exists as to what constitutes a cycle.” So I wonder about the discussions of studies in the links posted that quantified the number of charge cycles at different top and bottom charge capacities. If charging from 80 down to 20% gives 3000 cycles, does that mean 3000 of those 80/20 cycles, or 3000 full cycles, which would be almost 5000 80/20 cycles. Either way, that adds up to a lot of miles of driving before battery degradation is an issue, but I am curious if there is consistency in what the studies are calling a charge/discharge cycle.

 

[f1geek]

The cycle is defined by the end points. For example, when the cell was tested from 80% to 20%, that is considered one cycle. So you now probably realize that if you cycle from 70% to 30%, that is fewer miles than cycling from 80% to 20%. The additional cycles you gain from "short cycling" the pack end up giving you much more mileage from the pack before you hit the 70% state of health that is defined as "end of life" for the pack. Of course, there is nothing to stop you from continuing to drive the car once the pack gets to 70% state of health, but typically in the industry 70% is thought of as "end of life".