-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Wiki Sudden Loss Of Range With 2019.16.x Software
- Thread starter Dutchmeeuw
- Start date
faughtz
Model S P85DL
I posted this:
Range and acceleration loss | Tesla
And it’s still up, but doesn’t have any replies.
Can you see it?
NickleNiceness
Member
here are my data points, i do not have any tool to look at cell voltages, but i plotted my SC charging. My range was very stable as well as my charging on several superchargers was pretty much on top of each other. Then my charging rate dropped significantly when at a low SOC. Orange is may and June, Blue is the most recent
lightningltd
Member
Yes. I just posted to it and put in a link to here. We will see if it remains.
IMHO there are far less users on that site than there are here. Answers are found quicker here as well.
DJRas
Member
Tesla forum has not deleted a related thread... but it is barely read.
Quick drop in range after 2016.16.2 update | Tesla
Quick drop in range after 2016.16.2 update | Tesla
dhanson865
Well-Known Member
Did you not even click on the link? It's to a page on this forum, in this very thread. Exactly why I asked if you didn't read half the pages of this thread.
How are the admins on the other forum going to delete posts on this one? You are talking complete nonsense.
So I have to assume you aren't reading much of anything in this thread, just posting whatever you want to say without considering anything that anyone else posted in a thread with 1500+ posts.
Hi,
I also own a S85 with about 75.000 miles, but it lost only about 5 miles after the update, so far.
I think degradation of the battery wasn't defined very well, so far.
We all thought that the current amount of energy available at full charge of 4,2 V divided by the amount of energy of a new battery charged to 4,2 V subtracted from 1 and multiplied by 100 will be the degradation in percent. In pure physical terms this may be true, but technically this doesn't take the internal chemical changes in the battery into account, that have influence on usability (internal resistance) and safety (lithium plating). Up to the firmware update, causing a drastically lower range being displayed, it seems to be that Tesla, when calculating the state-of-health of the battery in form of the capacity for full charge, didn't take the usability and safety much into account. So over time the lithium plating of the cells, caused by high charge power at high SOC and low battery temperature, is increasing more and more, until the separator is punctured, causing internal shorts and my finally cause battery fire. That this happened only a few times up to now, even there are billions of cells in use, is a wonder. This shows that Tesla has done a good job in improving the BMS over time. The last changes definitely go into the same direction.
As has been already written, Tesla may now detect the health of the battery regarding safety better, by detecting the lithium plating through differential voltage analysis, when charging to 100 %. If the voltage starts to rise with a higher slope than normal, charging is stopped. The reason for this sharp rise is that the lithium ions cannot be intercalated into the graphite anode fast enough, which finally increase lithium plating further.
So I agree with those that say that the degradation (capacity, usability and safety) was already there. We just didn't see it until Tesla introduced the new software. In my opinion Tesla is not at fault regarding making the real degradation visible here, but it is certainly at fault regarding the missing communication. If only a few owners are hit very hard, by losing more than 10 % of there range, then it should be also possible to replace the batteries by refurbished ones. The parameters should be transparent.
Taking this into account, I can only warn people to constantly test the battery capacity with 100 % charges. They may make the situation even worse, because nobody knows if Tesla is already stopping charging early enough to prevent further lithium plating. It is better to use either ScanMyTesla or TM-Spy to read out the BMS values, but charging to lower SOC and dividing the displayed range by the SOC percentage, will also be good enough.
How can the risk of lithium plating reduced?
In winter it is better to charge immediately after the trip, when the battery is still warm. But avoid charging to more than 80 % if possible.
Keeping the car at SOC of 10 - 20 % most of the time, gives the chance to reverse lithium plating a bit.
BTW, I don't see the loss of capacity and the lower charge rates at the SuC as separate issues. Both work together in reducing the chance of lithium plating. So at least all owners of vehicles with S85 cell type may have to live with more visible degradation and lower charge rates.
Finally, I think this problem will not be that visible with cells with silicon in the anode, because there can the lithium ions be intercalated more easily.
I hope this helps,
Emil
I also own a S85 with about 75.000 miles, but it lost only about 5 miles after the update, so far.
I think degradation of the battery wasn't defined very well, so far.
We all thought that the current amount of energy available at full charge of 4,2 V divided by the amount of energy of a new battery charged to 4,2 V subtracted from 1 and multiplied by 100 will be the degradation in percent. In pure physical terms this may be true, but technically this doesn't take the internal chemical changes in the battery into account, that have influence on usability (internal resistance) and safety (lithium plating). Up to the firmware update, causing a drastically lower range being displayed, it seems to be that Tesla, when calculating the state-of-health of the battery in form of the capacity for full charge, didn't take the usability and safety much into account. So over time the lithium plating of the cells, caused by high charge power at high SOC and low battery temperature, is increasing more and more, until the separator is punctured, causing internal shorts and my finally cause battery fire. That this happened only a few times up to now, even there are billions of cells in use, is a wonder. This shows that Tesla has done a good job in improving the BMS over time. The last changes definitely go into the same direction.
As has been already written, Tesla may now detect the health of the battery regarding safety better, by detecting the lithium plating through differential voltage analysis, when charging to 100 %. If the voltage starts to rise with a higher slope than normal, charging is stopped. The reason for this sharp rise is that the lithium ions cannot be intercalated into the graphite anode fast enough, which finally increase lithium plating further.
So I agree with those that say that the degradation (capacity, usability and safety) was already there. We just didn't see it until Tesla introduced the new software. In my opinion Tesla is not at fault regarding making the real degradation visible here, but it is certainly at fault regarding the missing communication. If only a few owners are hit very hard, by losing more than 10 % of there range, then it should be also possible to replace the batteries by refurbished ones. The parameters should be transparent.
Taking this into account, I can only warn people to constantly test the battery capacity with 100 % charges. They may make the situation even worse, because nobody knows if Tesla is already stopping charging early enough to prevent further lithium plating. It is better to use either ScanMyTesla or TM-Spy to read out the BMS values, but charging to lower SOC and dividing the displayed range by the SOC percentage, will also be good enough.
How can the risk of lithium plating reduced?
- Don't charge to high SOC
- Don't charge when battery is cold
- Don't charge with range mode on when battery is cold
In winter it is better to charge immediately after the trip, when the battery is still warm. But avoid charging to more than 80 % if possible.
Keeping the car at SOC of 10 - 20 % most of the time, gives the chance to reverse lithium plating a bit.
BTW, I don't see the loss of capacity and the lower charge rates at the SuC as separate issues. Both work together in reducing the chance of lithium plating. So at least all owners of vehicles with S85 cell type may have to live with more visible degradation and lower charge rates.
Finally, I think this problem will not be that visible with cells with silicon in the anode, because there can the lithium ions be intercalated more easily.
I hope this helps,
Emil
Blu Zap
Grinning member
@egn1 I have to say that Tesla is at fault for not communicating as you say. But the cause of the fires in the 85kWh batteries is real, as we saw in Hong Kong and Singapore. They are being responsible by releasing a BMS software update to prevent future fires. But for heaven's sake, they need to tell the owners what's going on. We all have our set of stories how unannounced range reduction effected us, reduced car value and ticked us off. But that is the issue. We all know how to charge our cars and manage the batteries. That is why this flies in the face of you, me and the other Tesla owners here that Tesla would throw range reduction back on us. To lose 15-20% overnight with a software update has no reflection on how an owner cares for their Tesla's battery. Further insult is their comment to "Roadshow" that "most" owners won't notice it. WHO'S IN FOR THE CLASS ACTION!!??
Thanks for this additional information.
From what I read in your post, you believe the real degradation rate has always been much higher than what the crowd-source data has shown for years because that data has been based on a misleading BMS reporting, showing the degradation much lower than what has been for the older cars; and that the new BMS reporting is more aligned with the real degradation rate. Am I understanding you correctly?
Blu Zap
Grinning member
Yes. I think if we stick to kWh capacity it is a cleaner discussion. Anytime "miles" or "percentage" come up as a measurement metric, it is subject to additional formulas and calculations. I bought a car with 77 kWh and now have a battery with 60 kWh. It was about 73 kWh before 2019.16.2. Plain and simple.
Degradation visible in the crowd sourced data took only account the range calculated from the calculated full range by the BMS at the time of reading. The BMS algorithm has changed slightly from time to time.
But even then the crowd sourced data showed a lower degradation than the reading from ScanMyTesla. I once discussed this with Matteo, who managed the data, but we didn't come to an agreement.
I am also wondering why only the plotted average was in discussion, may be because it looked so good and everybody wanted only to see to positives in the chart. But there are also a lot negatives in there by having a lot of data point below the average curve. For me this either means that the production process for cells has a very large deviation, or the usage has a high influence, or both.
It would be great if Tesla owners could continue to update the writable copy of the MaxRange Tesla Battery Survey again. This would help to see the further trend of degradation.
As mentioned above it should be avoided to charge to 100 % for determining the range. Best to use ScanMyTesla or calculation from lower SOC.
But even then the crowd sourced data showed a lower degradation than the reading from ScanMyTesla. I once discussed this with Matteo, who managed the data, but we didn't come to an agreement.
I am also wondering why only the plotted average was in discussion, may be because it looked so good and everybody wanted only to see to positives in the chart. But there are also a lot negatives in there by having a lot of data point below the average curve. For me this either means that the production process for cells has a very large deviation, or the usage has a high influence, or both.
It would be great if Tesla owners could continue to update the writable copy of the MaxRange Tesla Battery Survey again. This would help to see the further trend of degradation.
As mentioned above it should be avoided to charge to 100 % for determining the range. Best to use ScanMyTesla or calculation from lower SOC.
DJRas
Member
The problem with using kWh capacity is NOT plain OR simple. Tesla does not display kWh capacity anyywhere visible to the owners. The "77kWh usable" of a new 85 kWh pack was not known without reading the CANBUS.
The ONLY data most owners know is State of Charge in percent or Rated range which are subjet to manipulation of the calculations.
I HAVE been affected by the range loss and only AFTER that occurred did I acquire the tools and knowledge how to read the kWh capacity. I do have a long history of charge capacity data in the form of rated range via Teslafi. But how many kWh that represents is subject to calculation.
Further complicating the calculations is the "buffer". People keep speculating whether it is or is not included in range calculation.
MY calculation shows the buffer is NOT included but that Tesla at some point has manipulated the Wh/mile constant used.
Further muddying this is the speculation whether these changes are or are not related to the fires. There was a single press release stating that software updates would be made to reduce the risk of fires.
After that, however, Tesla has ONLY officially announced that the changes were made to improve longevity of the battery. Tesla Service Center personnel explicitly said this is NOT related to the fires.
My inference is that Tesla wants to separate this as NOT a safety issue which could then come under NHTSA recall order.
Last edited:
dark cloud
Active Member
Great post Emil, thanks for contributing to the thread. That is a great summary of what this whole thread is about. Might I ask your education/training regarding lithium battery chemistry?
Tesla BMS is considered as good as it gets, so I question whether charging a cold battery matters; does the software not warm the battery first before charging if required?
Similar threads
