Wiki Sudden Loss Of Range With 2019.16.x Software

[Droschke]

I posted my P85DL+ for sale in the market place forum.....still on v8

I'm never buying a car again that has over the air updates. Not because I don't think it is a really awesome capability but because you'd have to trust the company behind the updates which in this case I do not.

That's sad. May I ask what you would like to buy?

 

[CuriousG]

It sounds like they are doing a major rewrite of your BMS based on all the changes you are seeing. I have a feeling that once the dust settles, you will have most of this sudden range loss back. Here’s hoping.

Somehow I'm not optimistic about this. Tesla had a chance to nip this in the bud and just say "we have temporarily reduced range for a small number of batteries so we can rewrite the BMS algorithms that may contribue to fire hazards".

Even if what you say is true, it sounds like Supercharging for these affected batteries will have further reduced rates. I was able to charge yesterday but the SC was so full I had to use the temporary urban charger that appears to only max out at 50kW. It did give me plenty of time to have lunch though.

Funny story to the urban charger as I pulled up to one of the two that were available, a Tesla was sitting in the other one and the owner is outside the car. I thought he was waiting on me to ask about SCing when I opened the door. He pulls out his Tesla card and says the chargers aren't working but to hang in there and he's there to repair it. He had his family with him and went around the corner to what I can only assume is some type of breaker switch that he flipped back on. This could have been an opportunity to ask about the recent degradation issue but I missed the boat and had someone already waiting for me at the lunch destination.

 

[sorka]

It sounds like they are doing a major rewrite of your BMS based on all the changes you are seeing. I have a feeling that once the dust settles, you will have most of this sudden range loss back. Here’s hoping.

I'd say the odds of that are just about zero. From the sampling in this forum, it's likely the update effects half of all older models and the half that aren't effected will be eventually as the conditions that the BMS management system is now adjusting for become a reality for those batteries too.

99.9% of everyone effected has no idea what's going on. They haven't seen this thread or read the electrek article. They noticed a drop in range and some of them will assume degradation and some will ask Tesla and be told it's normal when it isn't.

With such a small number of owners who TRULY know what's going on, why would Tesla do anything different? This will practically eliminate warranty claims for the older batteries and will accelerate upgrades to newer cars.

The only way this is going to get fixed is for enough effected owners to ban together, find a good firm, and convince them to seek the formation of a class. At that point, 100% of all owners effected will be notified via a call to join the class.

 

[sorka]

That's sad. May I ask what you would like to buy?

Z06. Possibly a GTR.

 

[faughtz]

Z06. Possibly a GTR.

That’s funny.
I sold my z06, and bought the p85dl, based on the 691hp claim.

 

[PaulusdB]

That’s funny...

It's sad, so sad (so sad)
It's a sad, sad situation
And it's getting more and more absurd
It's sad, so sad (so sad)
Why can't we talk it over?
Oh it seems to me
That sorry seems to be the hardest word
(Credit to B. Taupin/E. John)

 

[Droschke]

99.9% of everyone effected has no idea what's going on.

I'm surprise there is no thread on tesla.com forum about this. At least I can't see one.

 

[tinm]

Got a notification in the car of another software update but I’m to the point of calling Tesla before installing because of concerns the update might chop my car’s range. Tesla assured me only a handful of non-battery-related changes so I did the install.

New update is 2029.20.2.1. No problems detected. 90% SOC is still 224 miles.

I don’t like having to worry about OTA updates; they used to be fun.

 

[Target]

@IngTH has done a nice job posting here, thanks for that, you have put forth some of the better tech data around why they might be doing this.

For any that haven't see this post / video yet it would be worth watching.
Battery Degradation Scientifically Explained

@EV-Tech Exp
Can you pop into this thread and take a look around? (if you haven't already)

I currently have a 2013 S with 1014114-00-B and 49K on the clock. I only procured it from Tesla CPO a couple of months ago. I have no idea on the history or usage. I did a longer trip last weekend that should have exposed the car to enough abuse to get the degradation algorithms going if they were going to and didn't get a noticeable degradation. (I don't have hard data but rated range looks close enough)

I also live where it gets cold and this car was local, but it is very possible it wasn't used significantly during the winter based on the condition of the steel bits underneath.

I am going to speculate:
Based on what I can gather, cold/cool battery temps when supercharging sounds like a real issue due to Lithium Plating. Thinking about where I have seen the temp probes in the modules (in the middle) it makes sense the the cells closer to the edge of the pack would be much cooler than the central cells when doing pack heating. Putting this in place with winter cold soaked super charging would imply that the inner module cells would be at temp, but the outer edge cells wouldn't be yet, causing potential issues in those outer cells. The corner modules would have it worse of course. I believe the fluid connections are in the rear of the pack, which implies that front of the pack would be the last to warm up, and since the front is also the most exposed while driving it could be it wouldn't stay warm, however the temp sensors could report all is well due to their central location. Pull into a super charger and those front edge cells get the bad combination of low temp and high currents. As discussed, it doesn't even have to be Chicago winter for this to be an issue, anything below 15C could be a problem, and 10C (50F) sounds like it becomes a fair bit more likely. Parked in a garage with still air for a while before charging starts would likely reduce the issues around temp imbalances while making the pack heating more balanced and effective. Still though heat transfer will be an issue and certain cells will always lag, hopefully just less so.

To carry this thought one step further, the temp sensors in the middle of the modules during Super Charging demand chilled below ambient water where the edge cells may still be cold soaked potentially maintaining the issue all the way through the charge cycle. Based on seeing affected cars with locations all over the country I have to believe that it doesn't have to be Chicago cold to be an issue, but the data provided is that 50F is cool enough to substantially raise the chances of plating.

Data also provided in this thread leads me to believe that Lithium Plating can be detected during the charge cycle, AND could trigger voltage to rise to 4.2 even though the battery is around 80% full. Once the charging current is removed the effect goes away and the voltage drops to normal 80% levels. Alternately they are detecting the plating by seeing the voltage characteristics and, depending on the severity, trigger a cut off voltage that they feel is safe.

That was speculation. Please treat it as such.

 

[tga]

Based on what I can gather, cold/cool battery temps when supercharging sounds like a real issue due to Lithium Plating. Thinking about where I have seen the temp probes in the modules (in the middle) it makes sense the the cells closer to the edge of the pack would be much cooler than the central cells when doing pack heating. Putting this in place with winter cold soaked super charging would imply that the inner module cells would be at temp, but the outer edge cells wouldn't be yet, causing potential issues in those outer cells. The corner modules would have it worse of course. I believe the fluid connections are in the rear of the pack, which implies that front of the pack would be the last to warm up, and since the front is also the most exposed while driving it could be it wouldn't stay warm, however the temp sensors could report all is well due to their central location. Pull into a super charger and those front edge cells get the bad combination of low temp and high currents. As discussed, it doesn't even have to be Chicago winter for this to be an issue, anything below 15C could be a problem, and 10C (50F) sounds like it becomes a fair bit more likely. Parked in a garage with still air for a while before charging starts would likely reduce the issues around temp imbalances while making the pack heating more balanced and effective. Still though heat transfer will be an issue and certain cells will always lag, hopefully just less so.

In one of the TM-spy/ScanMyTesla threads, it was reported there are 32 temp sensors in an 85kWh pack, 2 per module. I'm pretty sure it said coolant inlet and outlet temp at each module, so the car knows the temp throughout the pack. From the teardown threads, it's clear that the coolant loop touches all 7104 cells, so all cells should all be heated/cooled equally.

You shouldn't see a case where the front module is too cool to supercharge, but the car doesn't know. Even if it is colder than the others, the car should know and wait until it warms.

 

[bhzmark]

(A) it's likely the update effects half of all older models and (B) the half that aren't effected will be eventually

I haven't seen anything supporting either of these claims.

 

[Target]

In one of the TM-spy/ScanMyTesla threads, it was reported there are 32 temp sensors in an 85kWh pack, 2 per module. I'm pretty sure it said coolant inlet and outlet temp at each module, so the car knows the temp throughout the pack. From the teardown threads, it's clear that the coolant loop touches all 7104 cells, so all cells should all be heated/cooled equally.

You shouldn't see a case where the front module is too cool to supercharge, but the car doesn't know. Even if it is colder than the others, the car should know and wait until it warms.

Certainly makes sense, even then the software guys are making assumptions of cell temperature based on water temperatures. In a literal and figurative corner case they could be incorrect, although not as grossly incorrect as I was speculating previously.

I went digging around to see why I thought that and I see now that what I was thinking of as a temp probe is the BMS leads.

 

[Blu Zap]

This is real. 2014 P85+ with original "D" battery. Original rated at 265 mi. Before 2019.16.xx 255 rated at 100%. Now rated 100% is 227 miles. Okay, but that is all about the rated algorithm. So lets look at facts. I charged to 100% in Boise Idaho. 227 miles rated. I drove to McDermott Nevada. 182 actual miles covered at 319 wh/mi. 7% remaining charge. 58.1 kW/h used. Okay, 319 is higher than the "rated" 295 wh/mi. But with that, I am no where close to what I thought I could get. Actual miles are less than 200 miles at 319 wh/mi. I have been careful about charging and preserving battery life. So this is a blow.

I did talk with a Model S 2013 85 owner at the Lovelock, NV. Supercharger. He had a warranty replaced rebuilt "B" battery. He is still getting 244 rated miles after this new software. I believe the software reducing range is directed to specific 85 kWh batteries, specifically the D's and probably others.

Tesla is trying to get out from under the looming warranty replacements for these batteries by limiting their range. We can discuss how they got to this, but as an owner and driver who has taken long trips, this is apparently obvious. A software update does not reduce range like this without a malicious intent.

 

[bhzmark]

One data point. My range and performance metrics all seem the same for my p85dl but I do notice that battery cooling noises seem to come on more in hot weather. Previously they would only seem to come on during supercharging or when really.pushing the car hard. . Now they seem to come on just when ambient temp is > 90.

 

[Chaserr]

My car is now the loudest in the parking lot at idle so it's definitely pushing the cooling system harder than it was last summer.

 

[DJRas]

FYI, "D" is no longer a sufficient answer - once they rev the part number, the number and letter suffix revert to "00-A"

In the early days, there was one main part number, so the letter was sufficient to identify a pack version. If you look at the degradation spreadsheet, we now have "1014114", "1074978", "1031043", "1088815", "1071941", "1107172", and who knows what else.

Your picture shows you have "1014114-00-D". You know that your battery is earlier than my "1014114-00-E", but you can't infer anything about ours in relation to another part number ("1074978-xx-y", for example).

So, my friend photographed his battery and his is the same part number . Just several thousand lower serial number. His is unaffected yet has more miles, more supercharging etc.

 

[Target]

Speculating based on what I have seen in my research.

Lithium Plating appears to happen the fastest under the combination of high state of charge, low temperatures, and charging. 50 degrees F / 10C is defined as the beginning of low temperature in this situation. Like many things 50F battery temperature isn't the end of the world, but certainly the colder it gets the more likely / faster the metal Lithium can form. Same with State of Charge, the higher we go, the worse it gets. Combine very low temperatures, high charge rates to high states of charge and you most likely have the recipe for metal Lithium forming the quickest. If your normal usage has that combination frequently then there is a good chance you have metal Lithium forming in your cells. The software recently deployed to the car / BMS most likely can detect the voltage irregularities during charging that are caused by a Lithium plated anode. I would guess this is why certain cars keep seeing range loss during charging as they are using the charge cycle to measure / estimate the amount of Lithium that has been deposited on the anode. Like many things in the world the first reading isn't sufficient to accurately determine the scale of the issue, but as the new software gathers this very specific body of data it can better assess the state of the anode. If it finds the anode to be in poor shape it takes as many of those issues away as possible. High state of charge is the one that we notice the most right now (in the Northern Hemisphere). Come winter I bet we will see more heating happening to the packs before they charge and various other adjustments that further inconvenience us.

I certainly hope that Tesla is gathering the very specific body of data from the cars and bringing it back for further analysis. We can all hope that the analysis allows them to refine their models and make them less conservative. I wouldn't say that all the range will be coming back though. The people who have lost the most KWH are likely to never see them all again. They might be allowed to get back up to 4.1 volts again someday, but 4.2V is likely off limits for those batteries forever. I wonder if this is going to impact how Tesla rates battery packs/ cars in the future? Are all of their new designs going to show up with a lower charging voltage cutoff from here on out to keep out of this mess? I don't know, but if I was them I would consider it.

On the topic of normal degradation vs software limited, I certainly feel the consumer side of this one, even though my car isn't yet impacted I certainly don't look forward to the day of having a "heavy 60". The flip side is if they would have had the Lithium Plating detection built in back in 2012 then this would be part of normal degradation. However they only did it in response to some incidents and we are all paying the price of their ignorance / lack of will to address this issue. We are paying for that in time spent charging, lost usefulness, and almost certainly loss of future value. On the positive side, assuming their detection system works, we are less likely to wake up to a fire in our garage, or worse yet, not wake up ever. I don't fault people for being unhappy. Tesla has identified an issue with their batteries, and they are limiting them to prevent fires. They don't want to give the other side any more data to sue them with so they are keeping quiet, when you are dealing with life and death it shouldn't be surprising that they will deal with unhappy and alive owners bothering them, versus dealing with unhappy loved ones of dead Tesla owners. They can probably settle all of the lawsuits we could ever bring on this topic for less than the cost of just one wrongful death lawsuit.

Tesla has recognized and started addressing this issue. What are the odds that any other EV manufacturer has recognized this also? I would say pretty low. They might be reading these articles and forum posts and adding it to their to-do list, and if they aren't they are VERY foolish. My expectation is that they still won't do it until they have this issue many years from now because they are VW/BMW/GM/everyoneelse and they "KNOW WHAT THEY ARE DOING". That is unfortunate but that is how the world turns. I am glad Tesla is dealing with this issue now and not waiting until cars are burning down every week.

Speculation over

Fact
Let go long enough Lithium Plating eventually grows Lithium dendrites, and Lithium dendrites eventually puncture the separator and short out the cell. If enough heat is generated by the short, you will have a fire.

 

[tiancai8888]

I am a Chinese tesla user. just in my local region （hangzhou，china），at least 20+ old version of tesla mainly 85 70 have reported sudden lose of battery range. Many of them using SC a lot, So Some specluate that due to frenquent use of SC and full charge(100% of charge) tesla just cut the range to protect the battery .

Here is my specluation:
I think tesla may have some sort of algorithm that counts the number of using the SC and the number of full charge. When the algorithm thinks your buttery is getting worse , It may just cut the range to protect the battery.

Also the battery and the motor have 8 years warrant in China. Tesla China won't replace the battery above 70% of range(through tesla never give the exact number of model S and X. the model 3 is 70%).

Tesla may just cut the range but also let the range stays above 70% and stretch the lifespan of the battery over 8 years, then tesla could free of any legal problems

Just my specluation. The feedback tesla give us just the normal degredation of the battery .

 

[DJRas]

This is real. 2014 P85+ with original "D" battery. Original rated at 265 mi. Before 2019.16.xx 255 rated at 100%. Now rated 100% is 227 miles. Okay, but that is all about the rated algorithm. So lets look at facts. I charged to 100% in Boise Idaho. 227 miles rated. I drove to McDermott Nevada. 182 actual miles covered at 319 wh/mi. 7% remaining charge. 58.1 kW/h used. Okay, 319 is higher than the "rated" 295 wh/mi. But with that, I am no where close to what I thought I could get. Actual miles are less than 200 miles at 319 wh/mi. I have been careful about charging and preserving battery life. So this is a blow.

I did talk with a Model S 2013 85 owner at the Lovelock, NV. Supercharger. He had a warranty replaced rebuilt "B" battery. He is still getting 244 rated miles after this new software. I believe the software reducing range is directed to specific 85 kWh batteries, specifically the D's and probably others.

Tesla is trying to get out from under the looming warranty replacements for these batteries by limiting their range. We can discuss how they got to this, but as an owner and driver who has taken long trips, this is apparently obvious. A software update does not reduce range like this without a malicious intent.

One of these software updates changed the multiplier for Wh per mile to 276. Previously it was either 300 or 295.
THIS gave a further unrealistic "rated range".

 

[sorka]

One of these software updates changed the multiplier for Wh per mile to 276. Previously it was either 300 or 295.
THIS gave a further unrealistic "rated range".

That would be entirely illegal. This is set by the EPA. They can't change it.