I have never had a cell phone battery degrade that fast. I've had my phone for 3 years. Leave it plugged in all night. Has to be charged during the day because I use it for work all day. At about 1000 cycles with a 13% loss.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Getting pissed with degredation
- Thread starter KetchupKilla
- Start date
No need to torture yourself over "battery degradation", a windy or cold day will impact your true range much more than your change in battery capacity. Even low tire pressures will impact your true range. Rarely will the drop in capacity continue at the same rate to where you will need a battery replacement.
I get that it should taper off, but when? A 10% drop in a year is not normal regardless what other people are saying. I expect degredation but not to this extent. I got a drop around 4800 miles and it stayed pretty steady for a long time. But a drop in 3% over a 1000miles Is not normal.
if it weren't so sad it would be funny.
Folks get a computer on wheels which lets them dig right down into the bowels of the system without having enough information to be able to accurately interpret the data being seen and are being misled by the stats apps so badly.
That label that says battery degradation is misleading rubbish and is based on nothing more than simple math. Bad enough that they imply it is a constant - even worse that they label the variable number as degradation.
The play number that the stats apps show is the battery capacity for the snapshot in time that the reading was taken, throwing away all the real variables needed to interpret the number. Battery capacity will change based on many factors and will fluctuate up and down depending on those factors - but it isn't degradation, its just current capacity compared to starting capacity.
In the 35k miles I've done in my car the not-degradation has fluctuated from an initial 309, max at 315 a few months later, a low of 285 miles about 6-8 months ago - currently its at 305 miles.
If it were real degradation it could not have recovered - therefore it isn't degradation at all and those flawed stats apps need to stop labelling it at such, unless they add disclaimers like the stock apps do.
for info, a non-exhaustive list of factors that affect not-degradation
amount of last charge (or series of charges)
starting charge
finishing charge
time since last charge
battery drain since last charge
temperature
TLDR
If you're not going to trust something, don't trust the permanency of the purported degradation number from those stats apps.
The stats apps are really useful, I love TeslaFi for so many things - just not in this case.
Just use percentage display and move on
Folks get a computer on wheels which lets them dig right down into the bowels of the system without having enough information to be able to accurately interpret the data being seen and are being misled by the stats apps so badly.
That label that says battery degradation is misleading rubbish and is based on nothing more than simple math. Bad enough that they imply it is a constant - even worse that they label the variable number as degradation.
The play number that the stats apps show is the battery capacity for the snapshot in time that the reading was taken, throwing away all the real variables needed to interpret the number. Battery capacity will change based on many factors and will fluctuate up and down depending on those factors - but it isn't degradation, its just current capacity compared to starting capacity.
In the 35k miles I've done in my car the not-degradation has fluctuated from an initial 309, max at 315 a few months later, a low of 285 miles about 6-8 months ago - currently its at 305 miles.
If it were real degradation it could not have recovered - therefore it isn't degradation at all and those flawed stats apps need to stop labelling it at such, unless they add disclaimers like the stock apps do.
for info, a non-exhaustive list of factors that affect not-degradation
amount of last charge (or series of charges)
starting charge
finishing charge
time since last charge
battery drain since last charge
temperature
TLDR
If you're not going to trust something, don't trust the permanency of the purported degradation number from those stats apps.
The stats apps are really useful, I love TeslaFi for so many things - just not in this case.
Just use percentage display and move on
I'm pulling data directly from the Tesla. No stats app, not Tesla fi
All kidding aside, so you have "some data thats not teslas" saying you have 9.38% degradation. Lets assume its accurate for sake of discussion.
Then what? Tesla's warranty is for retaining 70% capacity or having 30% loss over the warranty period of 8 years X number of miles depending on which model 3 you have. They dont even have to talk to you with "9.38%" degradation, reported in some non tesla approved app.
So, you go to tesla and tell them 'I have this app that reports 9.38% degradation" Lets even say they believe you (which is unlikely). What are you expecting them to do? They are not going to swap out your battery unless or until it reaches 30% in their measuring tools. What is your expected outcome from submitting a service request? Even if they agreed with your data (which is unlikely), what are you expecting other than "yep, that looks like what our data shows too, come back and see us when it drops another 20%".
This is a real question, not trolling or anything. What is your expected outcome from engaging tesla?
Ok, I need to step in a bit more aggressively than usual. As a guy struggling to make an app like ScanMyTesla, but taking a long time because of threads/complaints like this that I need to be very careful with...
No.
This is why I don't have an app released yet. I refuse to allow my users any opportunity to take a screenshot of my app to Tesla Service and say "See?! It's bad!". Hell no, dude. That's my app's interpretation of the data that you have interpreted further, not Tesla making a statement of degradation.
Tesla didn't give me a guide on how to interpret their CAN frames. A fantastic community of people has been able to decipher a lot (to the point where we're confident on the raw numbers being "correct"), but we still have little clue on what the numbers mean even if they are technically correct.
Take the information from the car's screen and bring that to Tesla. Gather data the old-fashioned way. Run an experiment, note the results. Charge to 100% somewhere that is above freezing temps, note the final range, and compare it to the advertised range. For 2020 vehicles especially, this is accurate enough for the purposes of a warranty claim, because displayed range is that "Expected remaining" value on SMT multiplied by the efficiency constant for your car, and that value matches "Nominal remaining" at about room temp on the battery. That is it. There is no need to be fancy about anything anymore to determine a degradation number. It's right there as the range available on the center screen. This is how Tesla does it for warranty claims!
Bringing a third-party representation of data that would be highly suspected to be an incorrect representation (and rightly so, these are internal data signals that have context we do not understand) will only get you resistance from Tesla Service. Not to mention this is the sort of thing that Tesla starts to classify as a problem. Do you want encrypted CAN networks so that we can never see data like this ever again? Because that's how you get encrypted CAN networks.
/rant
I apologise for being so blunt, but it needed stating. Just do it the simple way, with data directly from the car's screen at 100%, and you'll not only get the same number but come across much better to Tesla Service.
Thanks for posting this, it's what reminded me that we need to be careful about third-party data representations being brought to Tesla.
No.
This is why I don't have an app released yet. I refuse to allow my users any opportunity to take a screenshot of my app to Tesla Service and say "See?! It's bad!". Hell no, dude. That's my app's interpretation of the data that you have interpreted further, not Tesla making a statement of degradation.
Tesla didn't give me a guide on how to interpret their CAN frames. A fantastic community of people has been able to decipher a lot (to the point where we're confident on the raw numbers being "correct"), but we still have little clue on what the numbers mean even if they are technically correct.
Take the information from the car's screen and bring that to Tesla. Gather data the old-fashioned way. Run an experiment, note the results. Charge to 100% somewhere that is above freezing temps, note the final range, and compare it to the advertised range. For 2020 vehicles especially, this is accurate enough for the purposes of a warranty claim, because displayed range is that "Expected remaining" value on SMT multiplied by the efficiency constant for your car, and that value matches "Nominal remaining" at about room temp on the battery. That is it. There is no need to be fancy about anything anymore to determine a degradation number. It's right there as the range available on the center screen. This is how Tesla does it for warranty claims!
Bringing a third-party representation of data that would be highly suspected to be an incorrect representation (and rightly so, these are internal data signals that have context we do not understand) will only get you resistance from Tesla Service. Not to mention this is the sort of thing that Tesla starts to classify as a problem. Do you want encrypted CAN networks so that we can never see data like this ever again? Because that's how you get encrypted CAN networks.
/rant
I apologise for being so blunt, but it needed stating. Just do it the simple way, with data directly from the car's screen at 100%, and you'll not only get the same number but come across much better to Tesla Service.
Thanks for posting this, it's what reminded me that we need to be careful about third-party data representations being brought to Tesla.
jelloslug
Active Member
Exactly. How you drive has way more of an effect on range than perceived or actual degradation.
its all the same data - what ever you're using in your screenshot to show "Battery degradation" is misleading at best.
The number itself is real - its just not the degradation you are interpreting it as.
[edit]
just adding this after reading @camalaio explanation. I feel for you sir, you have been working hard to present all this data, for which there is no simple way to present it in a general car user format.
its almost like it needs a health warning. I don't envy your task.
Last edited:
I get where you're coming from, but the BMS is more accurate than you're giving credit for I think. It accounts for all sorts of variables, and I'd bet good money it's within 2% of the truth on every vehicle out there. The fluctuations reported are real fluctuations a lot of the time. I think part of that is your point - there are real fluctuations in usable capacity, so representing any particular point as "degradation" is potentially dishonest. That I can agree with.
But if the thing is 9+% out, clearly the bulk of that is probably real degradation. Even imbalance is a real form of degradation (some classes of which are partially correctable and can result in true increases in usable energy capacity). I'll go further to say that even the correctable portion of imbalance should be interpreted as degradation, and it would be fantastic if Tesla Service could note that and recommend productive steps to correct that. In that case, even at "only 9%", the correct approach should be to ask Tesla Service what's up (because clearly the pack isn't getting to balance itself via normal usage and something is up, likely with usage and charging habits).
EDIT: Didn't see your edit until mine, lol. I feel like I need to give myself a health warning tripping over all this stuff. As much as I disagree with Tesla on many things, the main thing I've discovered in all this data digging is that I never had to dig, and probably no one needs to. The range display is an incredibly honest and accurate representation of usable energy. I think that finding is worth it, since it gives credit to a lot of what Tesla shows and I can personally trust them better than before (about some things).
Last edited:
I understand where you're coming from.
My real complaint is that all these degradation report things all present it as an absolute instead of what it is - a point in time variable.
As you say, the BMS is tracking a whole bunch of metrics, but even with all of that the best it can do is estimate, which is why it fluctuates so much. Lots of small, low differential charges throw it off as do lots of similar charges with little variation.
Following my historical TeslaFi charge data, I can see an almost 17% variation in reported capacity, which makes labelling it degradation seem even more meaningless.
I see the current increases I've seen in capacity so fall back over time as my usage change.
Interestingly - I did find out that the Nav uses the BMS data in the Energy app to provide more accurate information for trip arrival SOC etc. I find it fascinating where Tesla applies this stuff.
"If it were real degradation it could not have recovered - therefore it isn't degradation at all and those flawed stats apps need to stop labelling it at such."
This^^^^^^^^^^^
If the battery truly degraded, then how does that number rise. Do you believe in battery ressurection?
The degradation crowd is absurd. I ask them OVER and OVER again to state how many real miles they get on a charge and they never answer the question. They constantly hide behind third party apps that have convinced them they're getting screwed.
This^^^^^^^^^^^
If the battery truly degraded, then how does that number rise. Do you believe in battery ressurection?
The degradation crowd is absurd. I ask them OVER and OVER again to state how many real miles they get on a charge and they never answer the question. They constantly hide behind third party apps that have convinced them they're getting screwed.
uujjj2
Member
Tesla batteries quickly lose ~10% of their capacity, and then stabilize for a long time. If you buy a Tesla with a certain rated range, expect to get 85-90% of that range for most of the life of your car. A 322 mile range new means 273 to 289 miles for the majority of the life of your car. With current battery technology, that's just how it is.
There are multiple 'degradation' types, some a permanent, some are recoverable. (Same as an ICE car)
By 'degradation' I am referring to the car reporting less mileage available than it did in the past. I am referring to 'Available Range Degradation', rather than 'Battery Degradation'.
In a similar way to an ICE car referring to the MPG getting worse, you get less miles out of a tank of gas than before. All kinds of reasons that may cause that, but at the end of the day, if you are getting 15% fewer miles before the gauge reads empty, you are getting 15% fewer miles. (It might be that the gauge is not reading accurate any more, there's a hole in the tank, or the engine needs tuning...)
Examples:
1) Permanent physical/chemical battery degradation.
2) Unbalanced packs/blocks/cell. Which can be re balanced
3) BMS estimation in-accuracy.
Can't do much about #1 (apart from following the charging guidelines long term, to some extent) The initial degradation I believe is mainly just down to chemistry/luck.
But #2, #3 and others can be effected by behavior. Looking from the outside, it's not possible to distinguish the cause of the 'degradation', just that the car is reporting less available range
My answer:
When I got the car driving at approx 80% average efficiency with a reported range of 307 miles, which is realistic driving in a P3D, I would get 250 miles till it indicated I would need to fill up.
After approx 12 months, the reported range was down to 272 miles, and at that same 80% efficiency, I would get 215 miles til it indicated I would need to fill up.
That's a significant difference, whereby on some days, for the same trip, I would need to a quick splash/dash near the end of the trip.
I could drive also drive at 100% efficiency (i.e. at rated energy usage), which would squeeze out a few more miles.
However, Now, after changing charging habits and not commuting, for several months, the reported range is back up to 296 miles. So I can guess that the loss of range wasn't largely due to #1, it was due to #2, #3, something else, and probably a combination of all three. The 'Range Degradation' has reversed itself to a large degree, but the 'Physical Battery Degradation' almost certainly hasn't.
But #1 being less significant, and is what Tesla measures when they are measuring 'Physical Battery Degradation'. (vs Reported Range Degradation)
Now Knowing what affects each of the factors, enables, if you want, to modify your behavior to gain more value out of the car. You do it in an ICE car, why not in an Electric. Different factors
No, I have never tries to run the car down to 0(-ve) miles or percentage to see if the missing miles were just there hiding, that's not realistic. If the car is telling me I have 250miles available, I have 250miles available. Doesn't matter what the cause is, still can't drive more than 250miles.
-------------
As far as labeling goes, 'Battery Degradation Report' may not be 100% accurate (in Tesla FI), it's really measuring a whole 'system' not just the physical batteries. I'm not interesting in the 'physical batteries' themselves anyhow, I am interested in the 'Battery System', and how that Battery System is affecting the 'value' of my car to me.
Not obsessing about every little change, but being able to see/track a trend is useful/valuable information.
To be completely fair, it is absolutely expected that there is more variance in the TeslaFi data than the data from ScanMyTesla. Any API-based service, like TeslaFi, is using the temperature-corrected usable range value (this can change ~3% even without the snowflake icon showing up in the car) and suffering from rounding error for the 100% range estimate (error is worse with lower SoC, more accurate for high SoC readings). The SMT data can avoid both of these issues, depending which values you are looking at.
More accuracy without CAN bus data (like with SMT) requires carefully controlled measurements, something these third-party services don't seem overly concerned with. And to be fair to them, they're trying to look at the forest, not the trees -- they amalgamate many vehicles' datapoints into fleet summaries. It would be better if they were accurate to tighten up the apparent spread, but their fleet medians are probably near reality. Of course, that spread means comparing any particular vehicle to that median results in poor accuracy.
Thanks, you said many of the things I was thinking. Only correction I'd make is that some forms of imbalance do cause permanent and irrecoverable loss (e.g. a brick that genuinely has lower capacity than the rest).
Otherwise, I agree it's fair to have different classifications of degradation. With the largest hand wave I can muster, daily fluctuations are real and due to "individual cell differences" that have a lot of complexity. And the BMS is actually reporting on that complexity! Accurately! It's a fantastic feat. My personal belief is that the 1-2% actual fluctuations we see are due to these factors. The larger fluctuations we see on third party services based on the API are due to calculation errors and temperature effects.
It's completely absurd for someone to actually measure 0-100% range experimentally. Timing it so that you have exactly zero range left but also happen to be near a charger is a huge logistical hurdle. Someone like Bjorn Nyland could do it by circling around an Ikea parking lot that had a charger, but the average person is not likely to go to this extent (and is likely to be kicked off the lot as well).
If you have 99 days that your car is "10% degraded" and one day it is "7% degraded", that doesn't mean it was actually only 7% all those other days. It very likely was 10%. Batteries are incredibly dynamic. And as you pointed out, that degradation does have a real impact on usable range!
Tesla has done an outstanding job of measuring and predicting available energy, in my opinion. I'm far from a fanboy (I have many issues with Tesla as a company), but they got this right.
JaeTheDev
Active Member
I appear to get about the same, and I have the aero cap kit on my wheels (so no hub) and I dont drive that nicely either. LOL. Glad to hear were in the same boat!
Im at a little over 15k miles and rocking 11% degregation per the scanmytesla app. I also have around a 6mV imbalance although that seems to vary depending on if Im moving or stationary.
Presumably the degregation is based off the capacity when new vs the current full capacity but Im not sure how the car reports full capacity and if that is affected by the discharge "fixes" people like to do.
Presumably the degregation is based off the capacity when new vs the current full capacity but Im not sure how the car reports full capacity and if that is affected by the discharge "fixes" people like to do.
I guess we could use Tesla's word, "retention" (also the opposite number of course). So instead of 11% degradation, 89% retention.
It still has the same issues of course if the number itself is disagreeable, but the word is better I guess?
Movement requires enough power draw that it can sag the voltages, and each brick will sag differently due to small physical and chemical differences. This is why your imbalance while moving may be smaller or greater while moving (yes, it can be smaller at times, though I'd bet this is unlikely overall). The same-ish principle applies to regen as well.
Also, despite my "nice imbalance" quote earlier, this imbalance number really doesn't tell you the full story regarding the imbalance of your pack. It's much more complicated and something I struggle to represent in a better way, but it ultimately reflects in the reported range/capacity anyways which is what we're all after.
Similar threads
