Teslanomics Examines Tesla Battery Degradation: Interactive Owner Spreadsheet Shows Results


APR 15 2017 BY MARK KANE 31

Teslanomics by Ben Sullins recently took some time to analyze Maarten Steinbuch’s Tesla Model S battery degradation data, and then upgrade it with an interactive graph that shows the percentage of battery capacity in relation to mileage, charge cycles and a car’s age.

One of the handy options also enables the ability compare results from particular regions.

As it turns out, some models noted a great 93% capacity retention after 220,000 miles. There were also some worse scenarios like a 85% result after just over 32,000 miles, but in general most are around 95%, which is very positive.

Teslanomics encourages others to join the research and add their own data to build a more comprehensive data sheet.

What is the Lifespan of a Tesla Battery and How Long Will it Last?

Tesla Model S and Panasonic lithium-ion battery cell

Haters love to claim that the Tesla car batteries, the ones that power the car, will only last 2 years or less.

Well, now we have some data coming from actual owners and their experiences.

Use our code and get $1,000 off a new Tesla Model S or X – https://teslanomics.co/td

TMC Dutch Forum – https://teslamotorsclub.com/tmc/threa…

Source Data – https://docs.google.com/spreadsheets/…

Data Viz – https://teslanomics.co/what-is-the-li…

Maarten Steinbuch blog about Tesla Model S battery degradation data – https://steinbuch.wordpress.com/2015/…

source: Teslarati

Categories: Tesla, Videos

Tags: , ,

Leave a Reply

31 Comments on "Teslanomics Examines Tesla Battery Degradation: Interactive Owner Spreadsheet Shows Results"

newest oldest most voted

Good looking initial data set. Would love to see a comparable ICE sedan data set for 200,000 miles. I would wager that the overall lost efficiencies were similar though the initial 3% would be different. The ICE data would be more susceptible to proper maintenance too over the long run.

M Hovis, there was a government study I saw awhile back that said, as I recall, that an average ICE car has about a 20% degradation in gas mileage compared to when it’s brand new, from just a combined wearing down of the various components of the engine and so forth. I’m sure a little Googling and the study can be found, but I’m on my phone right now.

But the key point is, Teslas actually compare favorably to ICE vehicles in that regard.

According to this data, one in twenty Tesla owners have had their battery replaced. I’m not sure if the data reflects pre-replacement batteries, or not.

My own X 90D range is down 10 miles, to 247, after 11K/one year. That’s only 4%. If it doesn’t decrease any further I’ll be happy.

The battery will continue to degrade and so will the range, it’s inevitable. The question is at what rate. This data seems to present a very positive outlook.

If after four years you still have over 90% battery capacity, as others have demonstrated above, then there’s really nothing to complain about, IMO. 🙂

Not that I would ever believe anything serial Tesla basher “Four Electrics” says about his (probably) imaginary Model X, but the good news is that bench tests of li-ion battery cycling show that degradation actually gets lower over time. That is, it seems to be a percentage of the remaining capacity, rather than the full capacity when the pack was new.

However, li-ion batteries do have a “shelf life” and will degrade over time even if they’re never used. So far as I know, no modern EV is yet old enough to show any capacity loss due to battery shelf life. And since battery makers keep tweaking the chemistry, changing it from year to year, the shelf life of any given batch of batteries may not be known for many years, perhaps well over a decade, after they’re made.

“According to this data, one in twenty Tesla owners have had their battery replaced.”

One is a double entry. 4 total of 127 = 3.15% or 1 in 31. Most were replaced at about 20k miles (1 at 40k). Doesn’t seem like a battery degradation issue, but I am surprised at the high percentage.

I should have checked more thoroughly, there are many duplicate entries, in fact only 46 different owners in the list. this makes the replacement 4/46 or 8.7% or 1 in 11.5

What is construed as a “charge cycle” (i.e. is charging from 75% SOC to 85% SOC (10% uplift) the same charge cycle as 20% SOC to 90% SOC (70% uplift). Thanks. Cheers

I believe it’s a calculated number from mileage, and each cycle represents a theoretical 0-100% charge.

So assuming 3.5 miles/kWh efficiency, a 75D with 100k miles on the odometer has done 381 cycles, whereas a 90D with the same odo reading has done 317 cycles.

Yes, Mint. This is how it’s described in the youtube description.

If only based on miles, and battery size, this says that depth of discharge is not factored in? I thought this was a fundamental factor to control for, if possible.

Yes, it’s very strange to see the term “charge cycle” referred to as if there is a fixed number of miles per charge.

Some EV enthusiasts recommend recharging your car every chance you get, and keeping it as close to 90% charge level as possible. Others recommend you don’t charge the car at all unless the charge level drops below 80%.

The reality is that you can’t describe the way the battery pack has been used (or abused) with just one or two numbers. A full analysis of charge cycles would have to log the SOC every time the car was put on a charger and every time it stopped charging.

All charge cycles are not created equal!

* * * * *

I’m curious to know what sort of use (or abuse) would cause a Tesla battery pack to lose 15% capacity in “just over 32,000 miles”. There will of course always be outliers in data like this, but that seems to be a pretty extreme outlier. Is that within the parameters of normal distribution, or does that indicate some sort of malfunction in the battery pack?

With careful charging/discharging and temp control it seems normalizing everything to “full cycles” is a reasonable first-order approximation.

So a 1000 cycle battery will do 2000 half-cycles, 5000 20% cycles, etc.

You’re right, a battery engineer will go into individual battery logs and such. But this is decent top-level data.

I’ve read charging Li-ion to 100% and letting it sit is bad. Might just be another bat-myth.

But a 1000 cycle battery can do also 4000 50% cycles. Therefore the charging type/range is very important.

“I’ve read charging Li-ion to 100% and letting it sit is bad. Might just be another bat-myth.” I’m pretty sure that’s no myth, but on the other hand I doubt any modern production EV actually allows the driver/owner to charge the pack to 100% of the battery cells’ rated capacity. For example, older articles on the Tesla Roadster said that Tesla reserved 5% on the top, so even when the car’s readout said “100%” it was actually only 95% of Panasonic’s rated capacity for the battery cells. The old rule of thumb was 80%/20% for li-ion batteries; that to maintain battery life as long as possible, they should never be charged above 80% or discharged below 20%. But plug-in EVs, especially newer ones, seem to be able to tolerate a larger depth of charging and discharging without significant loss of capacity. That’s not to say that you can’t abuse a battery pack, and I don’t mean just the Leaf’s passively cooled battery pack. If you repeatedly charge a battery pack rapidly to the car’s maximum capacity and run it nearly out of juice before recharging, then it almost certainly will wear out faster than average. Letting your EV sit out… Read more »

Interesting data.I have been tracking my Leaf’s battery degradation since June of last year. Of course it’s a 2012, so it has been pretty severe. Heat seems to be the biggest driver as the degradation slowed to almost nothing over the winter. I am approaching 5,000 charge cycles and deplete the battery almost daily. It has just over 48,000 miles. I say all this to point out the obvious. Larger batteries result in fewer cycles and less overall stress. Keep the temps down and a battery can last a good long while.

Yeah, Nissan’s batteries are a turd compared to Tesla’s in terms of rate of capacity loss. Even at moderate temps, the LEAF loses capacity very quickly. After almost 6 years, my 2011 LEAF in moderate San Diego only has about 60% of the range it had when new. It appears that newer batteries are somewhat better, but still not heat tolerant enough.

Folks need to get past what happened with the 2011 and 2012 Leafs. The new Leaf batteries are far better. The point of my post was that if you manage the temps and number of charge cycles a battery will last much longer. I, for one, am skeptical that Tesla has any magic bullet when it comes to battery chemistry. It’s all about temperature management and charge cycles. None of the Teslas in this data set have over 800 charge cycles.

It’s all about temperature management and charge cycles.”

I don’t know that Tesla has any “magic bullet”, but it certainly has engineered its cars to “baby” the batteries to extend battery life… and Nissan has done exactly the opposite, ignoring the need for active cooling and letting them degrade as they will.

But certainly part of the reason, possibly the main reason, that Tesla battery packs degrade so slowly is simply that they have a larger capacity, so don’t need to be charged as often, and therefore have fewer cycles for the same distance traveled than cars with smaller capacity battery packs.

How did you get to 5000 full cycles? 5000 cycles in roughly 5 years is 3 charges or minimum 150 miles/day driven electrically for 5 long years.

Do you really drive 55.000 miles a year or did you count not full cycles?

The 200K miles Tesla within 1 year (as long distance hwy rental) did get its pack replaced with “claimed” 6% degradation.

Math error: actually you have substantial less than 5000 cycles. 48,000 mi / 75 mi EPA range per cycle = 640 cycles.

My model S was manufactured in March 2014. It has 34,000 miles on it. The last 100% charge showed over 265 mile range which is the rated distance. For local driving I charge it to 65% and discharge it down to about 30%. I expect me to degrade well before my batteries the degrade.

serial anti Tesla troll Thomas

Some interesting News translated from Germany’s most popular newspaper BILD:

The American electrician Tesla is threatened with a strike at Tesla Grohmann Automation in Pruem, which was taken over at the beginning of the year.
The union had demanded the opening of collective bargaining for the company with 680 employees. “We received an unsatisfactory response from the company,” said Patrick Georg of IG Metall Trier (p.s.: IG Metall is the most powerful German industrial union). Tesla plays on time. “We’re going to check next week whether strikes are possible,” Georg said. The machines manufactured in Germany are important for the production of the Model 3, which is scheduled to start in the summer. IG Metall calls for the adoption of the collective agreement for the metal industry as well as workplace guarantees. At the current level, the wage level at Tesla Grohmann Automation is about 25 to 30 per cent below the tariff level, said Uwe Herzig, chairman of the works council. An offer from Tesla to increase the wage of all employees by 150 euros monthly is not enough.

PLUG IN AMERICA also tracks capacity loss. They show capacity loss for each different size battery. They are reporting higher capacity loss that shown here.


I believe most of the difference is due to reporting the battery management system’s guesswork, rather than real degradation. For a Tesla battery system, doing a 2-3 full charge cycles will recalibrate the system. The BMS does not balance the cells unless the charge is set to be over 93% and it also cannot have an accurate guess without a cycle down below 15 miles of range. Most people just report the guess from the BMS without balancing and calibration. Ironically, the vehicles with the most usage tend to report less degradation due to more constant use of the higher and lower edges of the SOC, even though the actual degradation is higher.

This is great data and excellent to see. Pretty much solidifies my intention of upgrading to the larger battery option on the Model 3. Seems like cycle count is the best measure for battery life and a larger battery cycles less frequently for similar mileage. If the upgraded battery on Model 3 is <$5k, I'd say it is a worthy investment. You have more range, decreased degradation over time, and even with degradation, likely more range than the base battery. There's other factors of course to having a fully degraded (65% capacity battery) such as the lack of regen available to you, so 65% capacity will feel like 55% or so. Either way, I think with this data I could expect to get 300k miles out of a model 3 with a larger battery until its capacity gets to unusable range for my lifestyle. This is more than the average ICE car engine or transmission and by that point I'm guessing that battery replacements will be much much cheaper than they are today.

It’s not the best set of data. Some cars degrade to 85% even at 1.2 years. If that was indeed a valid data, you’d expect to see many data points below 85%, yet that’s the only one. I don’t think they cherry picked the cars showing the best data, but I have to wonder how valid this might be.

The unfortunate aspect of all of this user data is that high mileage cars (>100,000 mi) do not correlate to calendar degradation. In fact, there is likely a slight inverse correlation. For example, taxis or limousines may exceed 100,000 mi in just two years while the average, or below average, drivers might take 10 years. The best comparison would be between a 2, 3, 4, 5, 8, 10 year old S85 Tesla all at 100,000 mi. Furthermore, the highest mileage cars can be “outliers” which force correlation and further confuse the data analysis. For example, a battery cycled 5 times per day for a year is not equivalent to one cycled once per day for five years.

“For example, a battery cycled 5 times per day for a year is not equivalent to one cycled once per day for five years.”

Agreed. Assuming everything else is equal, the battery that cycled once per day for 5 years will have additional “aging” due to time that is minimized in the 5 cycles per day for 1 year study.

Yeah, due to the charge cycles being more ‘ballanced’ in the PluginAmerica survey, I’d tend to give them more weight.

Interestingly, although a bit dated, it does seem the Tesla drive units are becoming more reliable. The sample size is too small for the newest (2016 and on) models, but the trend seems to be going in a good direction.

Neither survey mentioned specifically what was done by the owners to unwantingly cause more horrible degredation. It would be interesting to query the owners to find their charge/discharge habits as well as the ambient temperature (I wouldn’t have thought that would have made quite so much difference in an “S”, but who knows) that the car is exposed to, to see if there is any effect good or bad.