Tesla Battery Degradation - The results using real data and why the Plaid is the worst Tesla
ฝัง
- เผยแพร่เมื่อ 15 พ.ค. 2024
- We've 100s of thousands of data points from 10's of thousands of cars and we've charted the numbers to show whats happening with battery degradation
For most, the news is great, for Plaid owners, the news is not so good compared to others Teslas and not a car we'd want to buy when it reaches high mileage
Part 2 of this series looks at ultra high mile cars:
• Tesla ultra high milea...
And we've also looed at the accuracy of the EPA ratings
• Tesla Real World Range...
00:00 Intro and method
00:44 Battery pack
01:07 Model 3 Long Range
03:23 Model 3 Standard Range
05:30 Model 3Y Performance
06:20 Model Y Long Range
07:05 Model SX 2021 Facelift
08:15 Model S 85D
08:50 Model S 90D
09:45 Model S 75D
10:40 Model S P100D
11:30 Conclusion - ยานยนต์และพาหนะ
As a PhD student working on batteries life cycles I wish there is a link for these data to be publicly available
I would fully support this.
awesome PhD work! best of luck with your candidacy
in the interest of saving the world and everything we're told - that we have to buy a 100,000$ car or it's our fault the polar bears are dying - Tesla sure is preventing competition to learn about the data, manufacturing and maintenance practices.
Wouldn't we expect to know about all this data so we are better suited to decide the best ways to help reduce carbon???
@@captainLoknar with that logic all data should be public. Should all patents be revoked, so everyone could manufacture low carbon items?
@@meSOOlame if someone demonstrates they can do it better they should be issued a new patent.
Just off the top of my head I think I've achieved more than the whole IMF, the green initiatives of the neoliberal parties around the world and Gretel Thornburg the past decade.
But my whole point was to expose Tesla and their owners as hypocrites. Capitalists virtue signaling as environmentalists
2021 model s plaid here. 66k miles. Fully charged I get about 330 miles. Which is about 5% loss. I believe the plaid versions are simply driven harder and probably charged beyond 80 more frequently to achieve max performance.
That was my guess. People buying the more expensive car don’t care about battery life
Oddly enough, take an ICE powered performance car, use that performance a lot, and it wears out sooner too :)
@@PiDsPagePrototypes buy BMW M, drive it like a mad man and it costs the price of Plaid in repairs during that time.
Just cope in the comments
Btw Plaid. Pronounced Plad. Not Played.
Nice info. My own 2015-MS85D has 210,000 miles and has 89% battery still. It lives in a cool, but heated garage .. I Supercharge 10% and my L2 is 3 kW .. keep your battery cool guys.
L2?
Jesus.
I had a M3 long range in Hawaii. 35k miles and was down to 282 at full charge
Sounds like my '15 90D. Bought with 50k miles and 264 RM. At 140k, it is down to 258-259 RM. 50% Supercharge, Level 2 charge at 8 kW.
@@ssing7113 this is common mistake. Tesla will show you range at full charge based on your driving habits - not actually battery degradation. This is very important to know.
On top of that - dont charge your care above 85% unless you are going to long trip right away. After all you are at Hawaii - how much range you need per one trip?
Tesla doesn't take into account driving habits on the range when charging or at the top of the screen. We have another video debunking some of the myths about the range display.
Worth remembering that the bigger the battery (i.e. LR model) the fewer cycles needed for given lifetime distance covered. Just think how many "100 mile" full charge cycles a 24kWh Leaf needs to cover 100,000 miles.
The bigger the battery, the longer it´s lifetime is. It is the same as with Laptops and other devices.
The bigger the battery, the more of its energy is needed to carry it self around.
Leaf has horrible tempeeature management...
@@ABa-os6wmThe Leaf has NO temperature management at all.
Yes. You are also always pushing a little battery as full and as empty as it can manage. Also for the same motor power the smaller battery is pulling a higher 'C' rating meaning its working a lot harder. I think even with temperature management Leafs would still have degraded horribly.
Great info, from the UK, got a 2014 Model S 60, 100K miles, now at 51.6kWh (usable). 10% degradation if you include the 2.8kWh buffer.
I was a technician at a dealership.
Every once in a while, they want the old part back after a warranty replacement.
I would imagine the outliers would show good information to manufacturers about the parts.
Nice analysis-especially how it clearly clusters battery chemistries.
1. Be interesting to see the long term LFP lifespans.
2. And similar studies for different brands. There are rumours that VW-Audi MEB lifespans aren't great -but I've not seen large datasets to confirm or deny this. And without that: all we have is "anecdata"
I think the MEB lifespans should be fine cause most of these cars are around 170-300 HP so maximum like the Standard Range Model 3. So I dont think they will have more than 5% over 100k miles.
Is the database public? Great information!
I'm curious about the extreme outliers in the graph. My previous Model 3 dropped to 94.35% after nearly 6 years and 78,737 miles. So far, it looks like my Ioniq 5's battery is tracking the same pattern.
Hi, nice video. is there a link to those charts?
Thanks so much for the video!
In germany we have both Fremont and China Versions of the Model 3 LR, so what you want to buy used is a 2022 Model 3 from China which still has the parking sensors but also has the Matrix LED.
And yeah the LFP cells in the standard range have very little degradation, it would be so nice to have them in the LR models, but I think it would be too heavy...
Are these graphs accessible to everyone? Can you provide a link?
It's somewhere on tesla info. 6:24 shows headline and date. You can move frame by frame with , and .
It is, but if you go to the website, this data is nowhere to be found. It seems not to be publicly avaliable.
@@Piimentayeah, I was looking for it myself and I thought already I’m stupid ot what.
Great video. Nice to see some data to illuminate the issue. An interesting point was the variance in the smaller sample sizes - could indicates some consumers still be exposed to very high early costs / performance issues. Would be great to see something that looks into vehicle vehicle age and mileage at age vs failure probability.
Very reassuring to see the data. Great work
Thx for your work! Very interesting.
Beside battery deg. it would be nice to know how the internal resistance of the packs has changed bc. that effects your range too (especially going into the low end between 1 and 10% SoC)
Amazing data set and great dissection of it for us. Much appreciated!
On future videos, would you consider adding kilometres?
Also, why use KW when SOH% would give you consistent numbers across battery sizes?
kWh
My car is a S60D with a 100kWh battery swapped in. Both had been crashed prior to me getting them. My car would really mess up your chart but its no longer on Tesla's network so luckily you don't see that anomaly =D
Hello Tesla Info, Thank you for this very instructive video !
I'm planning to buy a S90D but unfortunately, i couldn't find the battery degradation for this model on your video...
What if you had clicked on the "refresh results" button when you switched from the S85D to the S90D ?
Thanks for sharing!
Good info but I think you should have added trend lines and it would have been very interesting to have tracked individual cars if you have that data. Some cars battery capacities obviously plummet and it would be good to see them more closely.
Excellent video. Now we just need this data for all the other manufacturers. Are the 2022 LGCHem and Panasonic batteries, with the bimodal distribution, the same chemistry?
Also one English usage point: 'depreciate' is about cost, but you are using it here for capacity. Capacity can 'degrade' or 'deteriorate' or 'reduce', only value can 'depreciate'.
seems to fit my experience: after 40000km 4.5%, at 62000 km now 5%. LG (Model 3 LR MIc 12/2021)
Great info, but what about the ambient temprature? Eg. the cars that has been in the northern countries with an average annual temperatur of +8 vs. +20? Has that any impact on the battery capacity?
We look at some of the outliers briefly in the second video in the series and temperature doesn't account for them, but we'll be looking at it again in the next video
From the research I’ve done, high battery temperatures lead to physical cell casting stress and higher chance for pack failure. Supercharging in the afternoon on a hot day to 85-100% is the absolute worst torture to the battery pack and delivery drivers who do it daily have had to replace their packs around 100K miles in a few cases I’ve seen.
Wondering the same thing. In Australia here temperatures are extreme, I will never buy outright an electric car however I would lease one so I am not stuck with a dude car at the end, I can just give it back.
@@mylesgray3470 Link to the research? I'm seeing anecdotes whereas TeslaInfo's "not really" answer is derived from the data. Very interested in the further deep dive myself as well, but I'm curious what makes you dissatisfied with this answer.
How do you get the raw data for these plots?
No link in the video description. Why? Is it secret? Inquisitive mind like to know.
Thank you. Any opinion on the effect of only supercharging
Unfortunately there's no way to see that via the API data we have access to.
There were a few outliers that were way off and that may be the dude with the model 3 in an apartment..so mostly supercharging? Apartments usualy have level 2 chargers now so if you are driving a model 3 and still in the hood, with no home charging options, you were made for the hood.
Not only is "only supercharging" most likely bad for battery life, it costs a lot more than charging at home ( unless you are grandfathered into the free supercharging like my wife's 2016 model X is ), so that's two reasons to avoid doing that more than you have to.
Have you any data on the time factor with regard to battery aging? I’m putting about 12,000km a year on my car and plan on keeping it a while.
My 2014 model S had a warranty replacement battery at 83,000 miles, another replacement at 104k and a third at 112,800. Every time, a pcb charging control board failed. Tesla repair used replaced warranty batteries for future warranty replacements, so it is always an old repaired battery that is fitted. A "new" battery is £17,500 retail when the 8 year warranty runs out, slightly over the value of the car... 🙄
Plan to sell mine about a year or 12-15k miles before the warranty runs out.
@@jruerph That's the inevitability...
Very nice data comparison, keep it up!
Excellent cohort study! The scatter graphs are very clear and it is interesting to see how the batteries from Panasonic and LGChem for example, differ in degradation, particularly on the M3LR. There seems to be a small proportion of samples that appear to have suffered significantly higher degradation than the bulk of the cohort. Is it safe to assume that this is mainly errors due to unreliable sampling (e.g. battery too cold/recharge depth too shallow)? Or are those mainly cases of rapid degradation (perhaps due to failed cells, or some form of "battery abuse" - deep discharge to under 0%, or always supercharging to 100%)?
Tesla did a service on my late 2021 M3LR (Shanghai factory model) at 52000km (30k miles). I asked them if they could do a battery diagnostic. The battery had negligible degradation (around 1%). I generally charge at home at 13A/9kW to 66% and the car has done about 5% supercharging (to less than 90%). Quite impressed especially compared with other manufacturers of EVs.
We're doing a few follow up videos to answer questions raised, the next one in on hyper mileage cars up to 250k miles, but a more detailed look at batterys and factories is also in the pipeline. As for the outliers, we've taken a look at a few low readings on the hypermile video and there's nothing obvious about them, I guess there could just be bad batteries in the mix, and we do see the same cars reporting low capacity repeatedly, but we also know the process for calculating the capacity extrapolates some values and at the extremes can cause noise. What does seem to be the case is that the 30% degradation limit is rarely being threatened and I would be surprised if many have their battery replaced because of degradation and not because of signs of serious pack failure.
Great video. Co you do this for Kia and Hyundai?
Thank you. Sadly not as we don;t have access to the data in the same way on those models.
Nice analysis. Thanks for providing this data. As you point it it shows the general trend of early drop and then leveling.
Electric Viking on YT has data on many 300,000+ mile batteries with 88% capacity.
Interesting video. I think I had a lemon or something with my Model 3 SR+. After Tesla remotely diagnosed it around just 20k miles, I was at a 13% loss already. The originally listed 240miles of range couldn't get me 165 miles of real world range even with an average of 210 wh/m. I got rid of it shortly after the diagnosis and it was very insightful and taught me what to look out for with EVs.
Thanks for letting me know most of the used market are full of duds.
@@altSt0rm I would say if you're looking, check what the mileage says at 100% charge. Don't just look to see it say "100%". You have to switch it to miles and see what that says.
The Tesla community also isn't helpful here because they all say the same thing "just switch it to percent and don't focus on the miles" but my car had abnormal battery degradation but Tesla wouldn't do anything until it was under 70% because they considered 87% of the max capacity "normal" at 20k miles.
Also, Tesla is the only EV manufacturer that exaggerates their range figures. So I've learned to subtract 20% from the max capacity for daily use because they only recommend charging from 10% on the low side, up to 90% on the high side. Then subtract an additional 15% if you plan to drive on the highway at speed. If you do the math there, that's closer to a real-world figure for range.
Tesla is by no stretch of imagination the only company that exaggerates the range of their vehicles ! 😂 Also what ICE vehicle,for that matter, actually returns the mpg advertised ?😂
@@stephencrowther524 I'm a Honda guy, my previous cars like my Civic that was rated at 27mpg hwy would easily get me 32mpg hwy if set to 70mph. My Acura MDX was rated at 27mpg hwy would get me 27.5mpg set to 75mph on the hwy. My other civics that were rated at 36mog would easily get 40mpg+ on the hwy. Maybe other ICE brands don't advertise accurate fuel economy, but that's not what my comment was about. My Tesla was drastically worse than advertised even at better than the rated 220 wh/m. As for other EV brands, I can only go off of reviews I've seen online for range tests. Kia, Chevy, Hyundai, etc all seem to do better than or at least meet the range expectations set by the manufacturer in the real world driving.
@@RobertoMezquiaJrwhen Stephen brings up ICE cars - totally irrelevant to anything you said - he reveals his childish anguish lol.
Nice thx for sharing
What I wonder and can't find online if, when you put in a new battery, if your charging speed also increases to the newer model speeds.
Im shopping for a used 21 or 22 M3lr at the moment. How do i tell if a car has the LGchem cells? As a used buy they seem a much better bet.
The 22 uses AMD HW instead of Intel, that's more important
There are more differences. LG Chem is better at cold temperature. Pana can deliver more power, this also means it will super charge a bit faster.
would (super) capacitors help? I wonder if since extra degradation seems to be linked to the performance variations, presumably due to launching the car more etc - I wonder if a bank of capacitors would help. charge could be moved to capacitors in as little as 20 seconds or so, but then allow the launch without taxing the battery as much.
Very nice work guys, impressed 👍😊
Is it possible for you to compile Standard Range Model Y 4680 battery data?
We'll put it on the todo list. If we can identify the battery we should be able to do it.
@@Tesla-Info The MY 4680 pack is the worst Tesla battery with respect to charging. It has anecdotal evidence indicating higher degradation so far. Looking forward to any data you can put together. We have one if you need any details. Model Y AWD standard range 4680.
We own 5 Tesla cars and we can say that they dropped more the first years and less now some of our car have mainly been charged at SuC and been driven 360.000km . But the most important part is the charging speed at SuC if you are on a roadtrip. Best so far is our Berlin build Y with BYD LFP almost flat curve until 60% state of charge and that is by far more than you need to the next SuC.
Why do you have 5 Tesla? I'm being serious, I don't understand. The overall efficiency of the power grid is 33 percent, a diesel is more efficient before you even plug in your EVs. There's no surplus renewable power so any extra load on the grid comes from fossil fueled perker power plants. As far as economical, you'd be better off driving a ICE getting 14 mpg vs the replacement costs for the battery. I've got 396,000 miles on my Toyota Corolla without any repairs except a $50 water pump at 365,000 miles. Plus look at all the environmental damage from copper mining and lithium mining. I'm just curious why anyone wants an EV.
Miles and cycles are the least of causes of battery degration the most siginificent fators are the BMS, time and temp. exposure.
What is the percentage of the normal battery degradation in the Tesla Model 3 year 2021 with 11000 km? For me it is almost 11%. Is that acceptable ?
You don't say which model or how you got an 11% figure, but the charts a 2021 LR would be in the range 71-76kwh using the API method we, tessie etc use which is a wide spread and 11% would be at one end of that. We'd suggest you follow the advice here to calibrate your BMS and balance yur sells which might help tesla-info.com/guide/tesla-bms-calibration.php
@@Tesla-Info Thanks. SR + , LFP
Interesting. I got my wife a 75 kWh model X in 2016. We've kept it limited to 80% charge for daily use, and drive it cross country once a year and use the superchargers. In that 7.5 years, it has dropped from 225 to 220 (or maybe it was only 215 last summer) miles of range. I got a model 3 SR in late 2022 with the LFP batteries, and Telsa says to charge those to 100% once a week, which I have done. In that time it has already dropped from 265 to 260 miles. In terms of miles, they have 70k and 8k respectively. We live in Orlando where it is hot most of the year.
Capacity fade is one metric of battery wear and tear.....what about the frequency of individual cells faulting modules, that fault the pack?
Excellent video! I love the way you are showing this by year and also miles traveled. I would also love to know the pack failure rate. People like to say that is extremely rare and not worth worrying about but when I look at used cars for sale a surprising number have had batteries replaced. What is the mean time to failure for a Tesla pack?
Is there data on older vehicles? I would expect that at some point during aging there is an inflection point and the battery starts to degrade rapidly over a short time frame.
If you can track the data points to individual cars, it would be interesting to do functional data analysis on them to get a better idea of the function over time
Wow, foundd this info nowhere else. Thank you for your efforts.
It would be useful to add the error bands to your analysis. I.e., the battery capacity is x kwh, +/-y kwh. Can you estimate how much of the data result from changes in battery capacity measurement? Thanks.
We're working on some updated charts including rated v displayed range and looking at % differences as well as absolute. We'll see if we can show the % or kwh change from new along side that.
I'm sorry but this analysis is completely ridiculous. These trends suggest near infinite battery life if you attempt to extrapolate beyond 120k miles. The rate of degradation at the 100k mark appears to be zero for many of the trends but there is no way this is actually true. The reason is that the battery does not die in a linear fashion. At end of life, the battery will perform reasonably well and then suddenly start to degrade rapidly. Also, the KHW capacity is as reported by the computer system and is not independently measured. How do we know the computer system is reporting accurate data, especially as the mileage gets up there? Again, the rate of degradation at end of life is expected to accelerate rapidly, so it is completely improper to make claims like 9.5% degradation going from 50k to 100k miles. What if going from 100k to 130k experiences another 10% degradation, then going from 130k to 170k experiences a 70% degradation to failure? Do you see how in hindsight, the 9.5% degradation of performance figures cited for the early mile cases are completely incorrect? That 9.5 % reduction in reported capacity actually represents 50k miles out of a total expected life of 170k miles for a true reduction in life of 50/170 or 30%. So although the battery voltage only dropped a small amount, and therefore the total maximum reported capacity will seem like a small amount, this can represent a much more significant reduction in overall battery life.
You have to be really careful performing these types of analysis on nonlinear systems because you will produce highly misleading results.
Degradation definitely levels out to minimal post 100k miles, drops off a lot after 50k.
@@wakeywarrior so the battery life is nearly infinite? Please explain why that should be the case. A new battery degrades faster than an old worn out battery? Please explain why that should be the case. Both of these statements are false. An old battery wears out faster than a new battery. It’s just that the wear does not manifest as loss of capacity. It occurs at the molecular level and performance remains relatively constant for most of the time until it really starts falling off a cliff or dies suddenly.
There is a huge difference between useful capacity and remaining life or degradation. Imagine a hypothetical battery that gets exactly 100 cycles and experiences zero capacity degradation during that time. The lack of capacity loss will make the analysis in the video trick you into believing there is zero degradation after cycle 99. But obviously after cycle 100 the battery will completely stop working. In fact the battery experienced 99% degradation after cycle 99.
Again, so you believe tesla batteries degrade then stop degrading and last forever after a while?
People want to know how many miles total to expect out of the useful life of the battery. They want to know when they will need to buy a new battery.
Imagine if the batteries that appear to stabilize are actually hitting their failure point. What if the reason why you don’t see the trend continue downwards is because after a battery reaches 60kwh or whatever, it fails before it hits 59 kWh and hence the trend doesn’t have much 59 kWh data because the batteries are all failing?
You have to be really careful making statements about degradation and remaining life of systems that are highly nonlinear.
I do battery pack development testing, one thing to consider is there is a customer/technician facing capacity, and then there is the real capacity for internal use only. Beginning of the the max and min voltage range is tighter, and opens up as Ah degrades. This is to compensate for capacity degradation. A specification may say the car needs 100 kWh after 10 years, for this to happen you run a 110kWh pack and software limit it to 100 kWh. So after 10 years, youd be running a degraded 110kWh pack that puts out 'closer' to the og capacity to meet some specification that says the 'pack' cannot degrade X amount, where the cells still are allowed degrade.
My Toyota Corolla has 396,000 miles and still runs great without any repairs except a $50 water pump. How well do these vehicles do at 400,000 miles. My Toyota Corolla will still go 380 miles per tank, how about the battery cars at that mileage?
@@Bryan-Hensley Just like my 1994 bmw 318 tds takes 5 minutes to refill and i refill it anywhere; and it gives off good heat in winter, i don't need to wear a blanket with the heater off, like many do in the cold...
@@josepeixoto3384 we will probably get shadow banned from this channel. Many of these types of people live in echo chambers. They don't allow logical intelligence people to intervene and mess up their fantasies.
@@Bryan-HensleyBut can you make your own fuel for your corolla at home for free?
@@JustifyJustin nope and you can't generate power for your EV for free. I tried that crap. Each of my solar panels do good to produce 75 cents per day in power. They only work well for about 4 hours per day. I live in East Tennessee and we have hazy days 80 percent of the summer, which cuts the output of my 250 watt panels down to 70 watts. I'd need a half acre of solar panels to even think about changing an EV, and on top of that, I'd only be able to charge in the afternoons.
controlling for location and average temperature could be revealing since batteries don't like extreme temperatures hot or cold
Comforts me into buying a SR with LFP battery, on the used market i have around 25k euros budget, ideally 17-18k but finding a model 3 for that price won't be easy
Fingers crossed i won't face any big issues not covered under warranty cause I'm taking a big risk putting 20k+ in that instead of putting it in savings, the apartment etc.. x)
For the LG Chem battery in the 2022+ models, the answer is NCMA chemistry. Much better cycle life and charge retention than NCA from Panasonic and standard NMC.
The large spread is interesting and I wonder if the power quality of charging has a significant affect. People who charge their cars at work on industrial sites with lots of inverters are exposed to higher levels of High Frequency emissions.
This is great! is there a way to get access to the raw data? I'd love to do some machine learning modeling to determine how each factor affects degradation, and (thus) be able to predict degradation of any combination of cars/factors.
Sounds like you could write a wonderful research paper on this
You dont need ML. Just basic statistical regression.
For Model 3 ‘17+, pretty much consistent with mine, I’m around 9% +/- 2% degradation (bms calibration)
sounds like the S75 is using 85 or 90 kWh batteries and they simply shift the charge/discharge window to maintain spec capacity.
This data is fantastic! Love that you can see clear banding around battery chemistries. People who downplay battery chemistry advancements ignore that they've been happening for quite some time now. There are tradeoffs, always, but we're moving forward. Degradation, in a word, seems overblown. But still something to watch, especially for the small number of outliers. Also, the 2023 cars that have already done 50k miles are probably doing a lot of road trip miles and thus supercharging a whole lot - good to see that they're not significantly off the trend line. Interested to see those hyper mile examples!
We've another video that looks at the hyper miles in more detail, including examples up to 200k miles
@@Tesla-Info Yep! Watched it already. I'm assuming this isn't all drivers, yes? Just people who opted into sharing data with Tesla?
Very interesting patterns emerging. All factors baked in, I do wonder how much of a "driving" style and charging habits differs for the average model, that is, demo and age of driver similarly to the way insurance companies use that info to set rates. Such that the Hardware would play X% and the driver's profile play Y%, really curious what would the split be... thanks for the data, receive warm regards!
Good point. A Tesla usually gets tons of heavy acceleration the first 5,000 miles because of the instant torque sensation. Once the first set of tires is done things usually calm down. So the rather steep decline at first could be an indicator for the toll that aggressive driving takes on the battery. Should have gone with a super capacitor…
Is there any data on batteries with 120K+ miles?
It would be nice to have a deeper dive breakdown of recharging habits compared to degradation. People who charge regularly to 90 versus 70% etc
My 2011 Nissan Leaf lost 50% of its battery capacity after only 20,000 miles and 5 years because I left it unused each of the 6 month winters. I wish I was warned. It sold new for $32,500 and I got $3,000 on a trade in. At the end when charged it only had a 50 mile driving range in winter.
Oh my god!
Leafs are notorious for their battery degradation but it sounds like letting the car sit unused for months was what actually killed the battery in your case.
@@FrostyAUT good to know. Will stay away like the plague
@@FrostyAUT yes
Maybe a nice metric to compare is the %capacity loss in comparison with the kWh discharged. This came to my mind after you said that the plaid batteries lost capacity faster, so its probably related to the fact that the plaid consumes more per km.
Tnx for the analysis!
Pls add new ModelY LR RWD
How do are u able to get the data sets
Can u cite your data source
API data
Excellent work. Thank you.
I just come to reaffirm the total nonsense of BEVs. A real Tesla car would run on wireless electricity and have no battery.
Like you commented at the end, sensible and less extreme performance users get a better battery life. Performance also includes people who actually need the extreme range and use it, thus doing more fast charging. Another
Interesting to see that all the big batteries (> 50 kWh) show twice the level of degradation (up to 12%) compared to the smaller (yet average size by international standards) 45-50kWh packs (4%-5%).
Any idea why? Could it be because they go through less cycles than smaller ones, and that somehow don’t get to regenerate as well/ as often?
We'd be speculating, but the LFP batteries which are now frequently used as the smaller battery is a different technology, and the larger batteries are often in the more powerful cars and maybe subjected to harder use.
Whats the average annual degredation of a model y at average 10 degrees c, averaging 5000 miles per year, only driven on weekdays and parked facing north? Andni listen to the radio
awesome data analysis, thanks for sharing
I'm watching this video and it's talking about the range over the lifetime of the battery but I'm not seeing any failure data. It seems like as far as power charging reliability that the data backs up that it's robust but I'm still seeing many many reports that are not seemingly represented in the data regarding complete failures. There's a lot of antidotal conversation about it though.
Why not zero the axis on the left?
I am wondering how the data was sourced, I heard it was through ‘the api’ but I don’t really understand that. My own car is a 2019 performance and I had the service centre measure it just before it went off (4 yr) warranty a few months ago. They said it had 84% capacity which they said is average for the fleet. That seems well below average. The car has 49k miles and has not had a particularly hard life other than being subjected to cold temps (Canada) charging very rarely over 80 and near never under 20. How can I find the battery degradation on this car?
You can benchmark your car using the guide here, we do use the API and we explain the results as part of the output. (There's also a video which shows it in action) tesla-info.com/guide/tesla-battery-capacity.php
You’ll NEVER get Tesla to share this data. Tesla doesn’t even share repair data.
I wonder if the china/LGChem data is actually in Kilometers instead of miles.
Would explain why they seem like they degrade slower and are being driven farther
Very interesting and useful
We should always be remember a internal combustion engine losses range in a similar way. And also the way you drive and the way you take care of your battery and charge it can make a big difference. Big difference between me a grandad not in a hurry, and my son always in a hurry, but that’s also the same for the ICEs, the only real problem is still charging and the time it takes, hear in Scotland there is a problem with the the lack of charging points
Wow, the STD is just 48kWh 😮 I always thought round 58.
Nice info.
8:00 It seems that Plaid is tweaked to the level where inverters will take more power than the battery can provide without heavy degradation. Looking at the data, the long range without "performance" option seems to result in best battery lifetime which is expected because it has most capacity compared to the peak power taken by the inverters so it causes least stress for individual cells.
What you didn't talk about is how the data was collected. Is this from cars coming to service (this would explain the gap in the beginning)? And what do the dots mean? Is each dot an individual battery or might the same one show up multiple times during its lifespan?
And survivership bias. I guess, cars with better performing batteries will be driven longer which might explain the apparent flattening because lower performing batteries don't contribute anymore.
Ideally one would have a line graph for each battery to see what is the case.
Nevertheless still interesting to seeb these bifurcations indicating a performance comparison between two different cohorts of the same model.
The only way to know is force Tesla to open up their API so open-source community can collect informative data to form an accurate conclusion which will never happen.
@@GBR9794 they might tell somewhere what the data actually represents. But it's in their interest to make the data appear favorable, of course... And obviously they don't want competitors to get too much insights.
The big question is how many have a 0kWh capacity i.e. died and at what mileage? It should be added to the dataset.
My 2019 MSP has 89.9% at 89.7k miles, so this tracks. All Supercharging since 66k miles.
And have you been charging past 80%?
@@rkgsd rarely but not never
@@thedownwardmachine I've always thought constant Supercharging is a little harder on the battery than say level 2 40amp/10kw charging. Unfortunately, many EV owners don't have access to home charging. It also depends on how far down the battery is drained. Better not to go too much less than 20% remaining.
@@rkgsd I agree, otoh a 300mi battery charging 20% to 80% effectively only has 180mi range between fill ups, which sucks. This is why 500mi batteries are actually important and I want one.
It's so easy to gather data from EVs, since their vitals are very easy to monitors. I really, really wish we had comparable data for IC cars on their performance - dyno data, fuel consumption, and ideally - dyno data, since power and torque curves are closely tied to condition of mechanical components.
For me outtake is whatever you're buying, you're getting ~58/45 kWh battery, everything above that is just a gratis from the factory.
My P100D (JUL 2018) according to Tessie: 91 kWh, 93.7%, 287 cycles, 487km range at 100% @ 64000km. 6.3% degredation after 6 years is OK. I don't feel anything. About 90% Supercharged. Charging it up to 100% like once every 2 months.
How do you check your battery degradation - by the estimate range at 100% charge vs EPA rating?
@@PaulPushkarov The Tessie app calculates this with Algorithms. It's not 100% exact but it's astonishingly close to reality. I think that's also how the data was collected for this video.
If you visit our website you can provide some data and we'll calculate it for you, we've been doing it for several years and Tessie have either copied us or come to the same conclusions completely independantly..
Very interesting data, thanks!
Do we know exactly which cells are being used in each of these models? I've dug into a lot of research on different li-ion chemistries (I build solar/backup systems). It's abundantly clear that some will outlast others by an order of magnitude or more - LiFePO4 is the most prominent example. It seems like we'd have all the tools necessary to predict the effective lifespan of an EV battery with access to something as simple as the cell manufacturer's data sheet.
You could probably research it, it’s likely hidden by their respective companies though
Thank you for this info.
I have a 2015 85D. It wasn't clear to me what you saying about that data set. It sounded more negative than about other data sets. Could you explain/clarify please? Thank you. I wasn't ever good at math.
I actually meant the opposite. The 85 battery cars suffered from something known as Batterygate where a software update dropped the range and charge speed on a lot of those cars giving them a bad reputation. Tesla have reversed much of that and the battery now looks pretty normal compared to others.
Interesting data, thanks for putting it up. Where I live used Tesla's just don't sell, an example is a 2019 Model3 standard range which has done a fairly low 85,000 km. The owner is selling for a little more than 50% of it's original retail price and it has 3.5 years battery and powertrain warrantee left. Given the high initial price, the asking price for this car is slightly more than a brand new, mid spec Toyota Corolla Hybrid with a full warrantee and 4 years free servicing. For me it's a no brainer, the Toyota wins every time and I think the vast majority of people feel the same. The Model 3 is a private sale and it has been for sale for over 3 months now. Even though the data supports EV batteries lasting quite well, the used/second hand market and appetite for them is non existent and that makes them a poor choice.
My 30 year old japanese racecar will get damned near the same mileage, Id love to see all these electric cars after 30 years, especially as when the range starts degrading, the charging frequency increases.
Pair this with the batterys degrading from simply time itself.... we certainly wont see "classic" electric cars around and theyre a legal nightmare to "recycle" in many places. However I do like these graphs, thanks for this.
The LFP batteries in the RWD are great as they will all have been charged to 100% as well, but have low degradation.
After having my wife's model 2016 75 kWh model X and keeping it limited to 80% for daily use, per Tesla's recommendations except when going on road trips, I was a bit surprised when I got a 2022 model 3 SR with the LFP cells and they said to go to 100% once a week.. Hers has 70k miles now and mine has 8k miles and hers has gone from 220 to 215 or so miles of range, but mine ( never supercharged ) has gone from 265 to 260 in the much shorter time and mileage.
@@phillipsusi1791 the curve isn’t linear. So don’t worry too much about that. It will probably drop off to hardly anything. Also be aware the LFP battery due to chemistry has a problem calibrating and determining exactly how much battery is left and ‘slips’ in its estimate quicker. So run it right down and charge it right back up again. Also the mileage estimate isn’t a great measure of battery health.
@@wakeywarrior My wife's model X didn't lose those 5 miles in the first year, or even two, or three, so I'm a little concerned that it seems to be faster so far on my model 3. When I'm driving around town, I'm less worried about how accurate the range is and more worried about long term life, so I might just decide to screw keeping the range accurate and top up to 85% a few times per week to get longer life. And what do you mean the mileage estimate isn't a great measure of the battery health? That is *the* measure of the battery health.
I do have 2 problems with those data presentations. First I do not see dead betteries mentioned anywhere... As my understanding is that batteries that malfunction may actually be a bigger problem then degradation and anyone I hear about with high milage Tesla has exchanged battery several times. And second problem is this reported capacity - where does it come from? Is it the battery controller/electronics. In case of laptops or phones - so many times I have seen electronics report almost full capacity and then the device suddenly died at 60%. Which means the data is often inaccurate and never for the better. A good data would come from measuring the last full power cycle of the battery.
Where is this data from?
We've answered this a few times, you can provide your token and we'll contact your car and calculate your battery capacity and save a summary (without personal imformation) in a fleet database. It's taken over 2 years to amass the data. tesla-info.com/guide/tesla-battery-capacity.php
Related to Plaid do you think it is related how these Cars are used? I mean if you Drive normally and use 1000+ HP only occasionally vs you pull out all the performance frequently as daily habit?
Based on the data, just like you said, it likely can be correlated to aggressive use vs battery makeup with exception of the China battery which seems superior for the data set.
Be sure. My Model 3 LR with 100K miles /165K km degraded only 5,5% at this moment and the degradation even went down from 6,5% to 5,5% in the last 30K driven! I am only driving in Chill mode, never accelerate fast, my lifetime average usage over the 100K mi is 150Wh/km or 242Wh/Mi.
Great info. Thanks !
I’ve got a 2019 M3 SR+ with about 9% degradation at 65k miles… hopefully it doesn’t drop much further
Makes sense about the plaids, you're looking at miles not charge/discharge, any time you loose efficiency via wheels or going 100+(drag) you're pulling more power and have to recharge so you have more cycles/transfer for the same mileage as a car that just drives "normally"
Data Source?
No Data source named, so ignore. Any muggins would place the source on every slide.
we collect data via our website tesla-info.com/guide/tesla-battery-capacity.php
@@Tesla-Info Seriously, well done for the data science part of it, but...
You collect this data without telling the user! Neither on this site explicitly, nor in your privacy policy. As a user, I would like to know, and in the best case have to opt-in. Just add a checkbox to it whether one wants to contribute.
@@jonathanmucke5937 I thought we'd made it clear that we kept copies of such things. I've made it more explicit in our privacy policy,
@@Tesla-Info I did not expect such an instantaneous, or any, response or action. I wish everyone took feedback that well and put in effort like you do!
I apologize for being so nit-picky, but it's anonymous, not anonomous.
Just a technical question: Do you only store a pair of mileage and capacity, or do you also store the car (in an anonymized) form with it? Asking, because of skew in the distribution, if values are not normalized by car. To be clear, this is nothing to worry about, just curious.