The most accurate way to measure battery degradation

[SergeyUser]

So does anyone know what is the best way to measure, in the most precise and practically doable way, the battery degradation? Tesla guarantees 70% of battery capacity retention under warranty, but how are we supposed to measure that? I always thought "Remaining miles" are pretty good, but it seems they fluctuate quite a lot, and battery can "gain range" while idling (check my notes in the thread: Displayed range has increased without charging). I have noticed some people think that battery is degrading in a hidden way (Battery degradation hidden in energy consumption?).

I currently use TeslaFi, as follows:
1. Charge car to a certain pre-defined % (80% (Y), for instance), make a note of the "Battery Range" value (X);
2. Let the car idle for an hour and see if the range will increase. Use this new "Battery Range" value if it is higher;
3. Calculate estimated 100% of range (X/Y) ("Estimated 100% Range)";
4. Calculate difference between highest value of the "Estimated 100% Range" (TeslaFi calls it "High Range") in %.

TeslaFi also gives you an option to compare your "Estimated 100% Range" to "Starting Range", which is the very first value of "Estimated 100% Range" ever calculated by TeslaFi. I think "High Range" is a bit better option, maybe we can only improve it a bit via averaging this value using, say, 3 highest values or something like that.

Anyway, in my case I have:
a. High Range = 264.1
b. Battery Range = ~202
c. Estimated 100% Range = 252.5

(1 - c/a) * 100% = 4.39%. My curent mileage is around 8k miles.

Is it even a good way to measure battery degradation? Do you know of a better way? Please share.

Also, if someone can hint to TeslaFi that they can improve their battery degradation report logic by taking into account the "range gained" after charges, I think it will improve quality of the data TeslaFi users are receiving (and they can do this retroactively, as new data is not required).

 

[AlanSubie4Life]

TeslaFi that they can improve their battery degradation report logic by taking into account the "range gained" after charges, I think it will improve quality of the data TeslaFi users are receiving (and they can do this retroactively, as new data is not required).

This gain is likely not related to degradation. It's just the BMS re-estimating the amount of energy available in the pack. There are a number of reasons why this might happen. It's actually "real" energy that becomes available, either due to a poor earlier estimate, a warming pack, rebalancing (?), etc. (I haven't heard a definitive reason here for exactly why this happens - these are speculative reasons.)

It's true though that none of the apps that I know of take into account vampire gain (I've noted this before - I only have Stats so can't speak to anything else), so there is some room for accuracy improvement, if they simply track negative losses (which would not be hard to do)

 

[5_+JqckQttqck]

So does anyone know what is the best way to measure, in the most precise and practically doable way, the battery degradation? Tesla guarantees 70% of battery capacity retention under warranty, but how are we supposed to measure that? I always thought "Remaining miles" are pretty good, but it seems they fluctuate quite a lot, and battery can "gain range" while idling (check my notes in the thread: Displayed range has increased without charging). I have noticed some people think that battery is degrading in a hidden way (Battery degradation hidden in energy consumption?).

I currently use TeslaFi, as follows:
1. Charge car to a certain pre-defined % (80% (Y), for instance), make a note of the "Battery Range" value (X);
2. Let the car idle for an hour and see if the range will increase. Use this new "Battery Range" value if it is higher;
3. Calculate estimated 100% of range (X/Y) ("Estimated 100% Range)";
4. Calculate difference between highest value of the "Estimated 100% Range" (TeslaFi calls it "High Range") in %.

TeslaFi also gives you an option to compare your "Estimated 100% Range" to "Starting Range", which is the very first value of "Estimated 100% Range" ever calculated by TeslaFi. I think "High Range" is a bit better option, maybe we can only improve it a bit via averaging this value using, say, 3 highest values or something like that.

Anyway, in my case I have:
a. High Range = 264.1
b. Battery Range = ~202
c. Estimated 100% Range = 252.5

(1 - c/a) * 100% = 4.39%. My curent mileage is around 8k miles.

Is it even a good way to measure battery degradation? Do you know of a better way? Please share.

Also, if someone can hint to TeslaFi that they can improve their battery degradation report logic by taking into account the "range gained" after charges, I think it will improve quality of the data TeslaFi users are receiving (and they can do this retroactively, as new data is not required).

It's a guess o meter at this point. Real way to track it is to charge to 100% and discharge it down to 10% and do the math. That number will be what comes out of the pack into km driven by total kwh consumed. Not what was put in via AC charging and estimated by TeslaFi.

 

[SageBrush]

My opinion: pick one reasonable method, stick with it, and trend over time. Any one measurement has a few percent +/- variance due to battery temperature and calibration shifts.

If I bother at all, I'll track the reported range once a year on the car's birthday after a charge to 80% and then an overnight rest.
I very much doubt that Tesla is mucking with the miles/kWh constant but that is easily corrected for if it turns out to be true.

 

[ewoodrick]

The only way to accurately measure it is to test it. Charge to 100% (probably should let it sit there for a few hours). Drive the car according to EPA standards until it stops moving. Rinse and repeat probably 5+ times. Take the best range (not the worse) of the numbers.

Or, stop worrying about the dang battery. Everything that you think that you are seeing is probably not real and is only estimates that change after software updates.

If you have a 300 mile battery and after charging to 100%, still can't get 200 miles range in warm temperatures, that's when you should start talking to Tesla.

And indeed remember, that you are going to start calculating that your battery sucks, because as the temps cool down, range will decrease 30%. On a good battery, 200 of 300 miles is going to be hard to hit.

 

[N54TT]

Get a can bus reader to see what the total kWh capacity is....compared to what it read when new.

 

[SageBrush]

Get a can bus reader to see what the total kWh capacity is....compared to what it read when new.

Why do you think that value is different from the range displayed on the screen * a fudge_factor ?

 

[N54TT]

Why do you think that value is different from the range displayed on the screen * a fudge_factor ?

I was answering what the best and most precise way of measuring degradation was. Comparing readings of what the BMS states is the full capacity..is probably the best and most precise way.....since that’s the core value we’re looking at when talking about degradation. Is that value different from the range displayed? Who knows. Displayed range is derived from calculations using total battery capacity. So probably not different? I don’t have a can bus reader. I’m just tracking what Stats app says is my total range at 100% and what Remote for Tesla states is the total kWh capacity.... No clue how they get the kWh data lol. Remote for tesla showed my capacity, when my car was 2 months old, was 76kwh. Its now showing 72.2kwh. Which is the same % decrease when I compare displayed range compared to when new.

 

[SergeyUser]

Guys, my point was to get a consensus of practical way to measure the number. Charging to 100% and drive to EPA standards is not really a practical way, IMHO ) What @SageBrush suggested is reasonable. My point is that there are some articles referring to "very small" (2% or so) battery degradation at 50k miles. Mine has more like 5% at 8k miles, and I wanted to check the math and get a good way to measure the number.

 

[rickdogg82]

I'd charge to 100%, reset the trip meter, drive until it nearly dead, and use miles driven and Wh/mile reading to calculate your capacity.

 

[AlanSubie4Life]

Mine has more like 5% at 8k miles, and I wanted to check the math and get a good way to measure the number.

Well, to do that you'd need a relevant datapoint from when you had 264 rated miles. You can do any sort of measurement now, but you really have nothing to compare to see what your degradation is.

There's nothing really wrong with the TeslaFi extrapolation - it's probably within 1%, so you have 4.4% degradation or so.

If you were able to go back in time and do your trip meter extrapolation and compare with today's trip meter extrapolation, you'd see the same 4.4% projected degradation.

You can always do a valid test now, document it, and then have something to compare to later if your range bounces back, gets worse, etc.

 

[SergeyUser]

Well, to do that you'd need a relevant datapoint from when you had 264 rated miles. You can do any sort of measurement now, but you really have nothing to compare to see what your degradation is.

My starting point is what Tesla should have delivered to me - Model 3 Mid Range, 260-264 rated miles. Which is confirmed by TeslaFi data I have since Day 1. I agree this is not ideal, but this is not nothing.

 

[AlanSubie4Life]

My starting point is what Tesla should have delivered to me - Model 3 Mid Range, 260-264 rated miles. Which is confirmed by TeslaFi data I have since Day 1. I agree this is not ideal, but this is not nothing.

Fair enough. Which means you have something to compare to - and you have 4.4% degradation as you say. There's likely nothing else you need to do to monitor it (unless they change the constants at some point in the future). It's a perfectly valid way to measure degradation (not really degradation necessarily - let's call it "loss of available energy").

 

[SageBrush]

Its now showing 72.2kwh. Which is the same % decrease when I compare displayed range compared to when new.

Indeed. So long as Tesla leaves the Wh/mile value constant the range change over time will be exactly equal to the Canbus total_kWh change over time.

 

[Uncle Paul]

Believe the issue her is that the instruments used to determine battery capacity are not precise. They are not designed to determine precisely the condition of the battery.

Kind of like how ICE cars fuel gauges stay on full for many miles, but drop much faster when almost empty.
Some people still have significant gas in their tanks, even when their fuel gauge reads empty.