Plan to Fix Decrease in "Est. Range at 100%" Problem

[Allistah]

Hey all --

I did some data analysis and thought it warranted a new thread. I'm going to post what I posted in another thread and I'd like to get some options on it. Below is my data from the post and a link to my spreadsheet with all the data at the bottom. Scroll to the bottom of the spreadsheet to see the final conclusions.

Take a look at this data... I took all the data from TeslaFi that was from charges above 50% and included incomplete charges as well to see if I could extract some information from it. The link to my sheet is below. I attempted to find out if there was something that resulted in a positive change in the Estimated Range at 100%, on average, so I could see if there was a trend of some kind. This could give me tips to what charging practices that may help us figure this out. This is what I found.

From the data in the link below, I found that on average:

• Decreases in charge % resulted in an INCREASE of the Estimated Range at 100% (by 0.7644 miles on avg.)
• Increases of charge % resulted in a DECREASE of the Estimated Range at 100% (by 0.9268 miles avg.)

The average change in "Charge %" that resulted in an INCREASE of "Estimated Range at 100%" was -0.2075%.
The average change in "Charge %" that resulted in a DECREASE of "Estimated Range at 100%" was 0.0256%.

In the data, you can also see that it was more impactful to increase the Charge % because on average it reduced the Estimated Range at 100% by 0.9268 miles which was more impactful than decreasing charges that resulted in an average increase of 0.7644. It hurts the Estimated Range at 100% more to increase the charge % than it does to lower it. Thats the theory at this point.

To boil it down: We should do more, smaller decreases in charge % and less, larger increases of charge %. If you do an increase, do it in one big charge, then do many charges with slightly decreased amounts.

See the bottom of my spreadsheet for these numbers in the link below.

Following this logic, if it's true, I should start by charging to 95% then every time I charge, decrease the charge to % by something like 5 or 10% to see if you get an increase in the Estimated Range at 100%, on average. So for me, something along the lines of:

Sunday - Charge to 95% and drive it down to 90% (So it's not at 95% very long)
Monday - Charge to 90%
Tuesday - Charge to 85%
Wednesday - Charge to 80%
Thursday - Charge to 75%
Friday - Charge to 70%
Saturday Charge to 65%

Here is the link to my spreadsheet where I came to these findings:
Tesla Battery Range Data

Let me know what you think!

 

[Allistah]

I'd also like to point out, that the word "fix" is a little misleading. If this works, it's not a fix - it would just be a workaround until Tesla actually fixes the problem.

 

[Allistah]

Here is my battery report showing the "Estimated 100% Range" that I'm trying to figure out how to get back up to where it should be. I'll be sure to post the charts again later and mark where I started this scheduled charge levels to see if it actually has a positive impact on this chart. Fun stuff! I'm really excited to see if this changes things!

 

[Allistah]

I just found something else out that is interesting.. Last night, I was down pretty low in Charge %, around 10%. I charged it to 50% overnight. Then when I got up this morning, I charged it to 60%. Then 70%, then 80%. I also noticed that in my data, that recently, I was gaining miles for small increments in the "Estimated Range at 100%" number.

My wife is out with the car right now. When she gets back for the night, I'm going to do another experiment. I'm going to charge the car 1% at a time to see if I get any gain in the "Estimated Range at 100%" number. If I do that 10 times and see nothing, I'll switch to 2% and see if I get anything there. Maybe the bug in the software has something to do with a charge completing and nothing really to do with how much the charge is. I'll post back tonight when I get some data! Ok, where's my wife and the car?! LOL

Here is the data that helped me come to want to test this:

 

[Sherlo]

I just found something else out that is interesting.. Last night, I was down pretty low in Charge %, around 10%. I charged it to 50% overnight. Then when I got up this morning, I charged it to 60%. Then 70%, then 80%. I also noticed that in my data, that recently, I was gaining miles for small increments in the "Estimated Range at 100%" number.

My wife is out with the car right now. When she gets back for the night, I'm going to do another experiment. I'm going to charge the car 1% at a time to see if I get any gain in the "Estimated Range at 100%" number. If I do that 10 times and see nothing, I'll switch to 2% and see if I get anything there. Maybe the bug in the software has something to do with a charge completing and nothing really to do with how much the charge is. I'll post back tonight when I get some data! Ok, where's my wife and the car?! LOL

Here is the data that helped me come to want to test this:
View attachment 480574

Interesting. I did that a few weeks ago (60 to 65) and noticed something similar. Didn’t think anything of it (thought it was just a data error or rounding issue) until your post. But then again, my rated range is now closer to normal.

 

[xilex]

My battery is doing the same thing, too. Currently about 9000 miles. I have to SC to 90% about once a week at 72 kW chargers because no place else to charge. Every now and then if schedule permits, I do some partial level 2 charging. But I don't go out of the way just to charge on level 2.

I did a 2000 mile road trip a month ago. I think I was around 304-307 when extrapolated to 100%. The final SC the day after I got home it went to 320. Since then it has been around 301-304. But 90% charge is around 269 miles. My drives, however, frequently give me more miles driven than rated miles used. Commute is 12 miles, and it will use around 9-11 miles. Sometimes I park the car and the next day it will have a few extra miles (eg 151 to 155).

 

[Allistah]

Interesting. I did that a few weeks ago (60 to 65) and noticed something similar. Didn’t think anything of it (thought it was just a data error or rounding issue) until your post. But then again, my rated range is now closer to normal.

Hey Sherlo!

Do you have TeslaFi? I was wanting to see if we could get others to try and see if charging by 1% at a time over and over again would have some increase in the "Estimated Charge at 100%" number. I'd do it but my wife is out with the car. If 1% didn't do anything, maybe 2% would. Also not sure if we should wait a minute or two between 1-2% sessions or not. I'll try not waiting and if I don't see any improvement, wait a minute or two to see if that helps. Charging up in small increments is easy, it's going down that's the difficult part.

Not sure if people that only have Stats for iOS could take measurements or not. I'm going to monitor both to see if it can be seem on Stats when I can get the car home.

 

[Sherlo]

Hey Sherlo!

Do you have TeslaFi? I was wanting to see if we could get others to try and see if charging by 1% at a time over and over again would have some increase in the "Estimated Charge at 100%" number. I'd do it but my wife is out with the car. If 1% didn't do anything, maybe 2% would. Also not sure if we should wait a minute or two between 1-2% sessions or not. I'll try not waiting and if I don't see any improvement, wait a minute or two to see if that helps. Charging up in small increments is easy, it's going down that's the difficult part.

Not sure if people that only have Stats for iOS could take measurements or not. I'm going to monitor both to see if it can be seem on Stats when I can get the car home.

Unfortunately I don’t use Tesafi just Stats (which doesn’t give anything close to Teslafi data).

But I will certainly try a few test increments on my next charge and report back, most likely early next week. You are right about the slow process of data accumulation with this project but happy to help.

 

[Allistah]

Unfortunately I don’t use Tesafi just Stats (which doesn’t give anything close to Teslafi data).

But I will certainly try a few test increments on my next charge and report back, most likely early next week. You are right about the slow process of data accumulation with this project but happy to help.

Cool, thanks! I'll try to take screenshots of Stats every step along the way to see if I can see any improvement there. I would think that it would show over time on the first battery screen - upper right gauge, max level number. That number moved around a lot due to the rounding errors the data has but who knows. Will see!

 

[Allistah]

Interesting. I did that a few weeks ago (60 to 65) and noticed something similar. Didn’t think anything of it (thought it was just a data error or rounding issue) until your post. But then again, my rated range is now closer to normal.

Im curious if there is something that is happening when a charge completes (no matter how big) that helps correct things somehow. Which then many mini charges can help it get it back to normal levels. If this happens, there will be no need to do a weeks long test next week. If this doesn’t work, I’ll go ahead with my series of lower-level as I described in this post to see what happens. Too many questions to be answered. ;-)

 

[Allistah]

I played around with this last night and a little this morning. You can't manipulate it over a short period of time - at least not from what I was doing. I was doing 1% charge increments and it would go up, but then back down. Essentially it went sideways without gaining much ground upward. Then I started doing one minute charges and that helped but it had a tendency to want to snap back to what it thinks is it's baseline or something. I think the manipulations need to be done over time and the car needs to be driven in between. I'll see what I can come up with.. I'm not sure if I'm going to work my way upward, or work my way downward this week. Here is a screenshot showing my efforts that didn't really get me anywhere.

 

[Allistah]

Ok, I have some info to report back on this. Don't do short charges like I did. that sort of messes with the internal calibrations in a not good way. Now the brake regeneration limiting is coming on much earlier, down in the low 80s instead of the low 90s. I think longer, larger charges are what this thing likes which is why you hear about people telling you to do some large charges from 10 to 100 three times. Not sure if I'm going to ask for a CAC reset and go from there or if I should just start charging it to 90 and leave it alone. The only thing is that with the brake regeneration coming on much earlier, my wife is going to be hating that. So, knowledge for all of you don't do what I did with the short charges. The car must have some sort of history or something with the charges which helps it do what it needs to do. I'm guessing that over time, it will correct itself.

 

[Allistah]

After driving the car today, I charged it to 100% and things are back to what it was before I started experimenting.

if I’m able to get anything substantial to report, I’ll post back. So far though, nothing good to report.

 

[toolman335]

The question I've always had is if the BMS predicts range based on driving history(specifically, recent driving history). Winter has arrived in New England so obviously my efficiency has gone south. Does the car know that I can't get 310? Since the temp dropped my 100% has gone down to 294.

Thank you for posting data. Not coming to conclusive observations is itself an observation. We're all trying to figure out these seemingly random variables. So many things work for some, but not others.

 

[JGard]

I appreciate your efforts here. Like you, I find myself annoyed by the inconsistent rating, especially my recent "predicted" 100% charge going from 317 miles to 304 literally overnight and now staying down at the 304. I know this number is "meaningless" to most, but I feel like this is about the only way we have to make sure there isn't some kind of major battery issue. Potential battery problems (like dead cells) is probably a non-issue, but being my first EV, I don't know, and battery tech is still like black magic to me.

Like you, I've tried a variety of different charging limits and small charges/large charges, etc, to see how it affects that rated range. All I can tell for now is that those small charges (5% or less increase) seem to not do any good to the BMS calibration. I've dome some larger charges and they haven't really changed anything, but I haven't done multiple in a row, mostly because it takes me a week to run through enough battery to do that, and I don't like leaving it unplugged for so long.

 

[Allistah]

Had an issue which prevented me from going in to work today so I took the car over to Tesla. They won't even talk to you unless you have an appointment. Well, they talk to you, but won't do anything without an appointment. Their system is pretty tight so if you want to have someone do something, just schedule an appointment. Most of the time they call ahead of time he said. Will look at getting the CAC reset.

 

[Allistah]

File a service request and got a call from my SC a few hours later. Talked to the guy for a long time. A few points that I took away from the call:

• Wants me to charge to 90% daily since we use down to 30-40% daily
• Says that it's good to let the car sit overnight at both ends (90% and 30-40%) because it takes time for the pack to "settle" and the system takes readings at those points which are taken into the calculations.
• He thought it was interesting that my mini-charges testing that I did made the number jump around like it did
• I learned that brake regeneration is limited on battery level AND temperature. In cold areas, it's possible to have brake regeneration limiting come on/increase when things get colder.
• At the lower end of the battery, it turns to measuring voltages only whereas up top it measures voltage and current.

So what I'll do is mix up my charging between morning and evening so that the car sits overnight at 90% and sometimes in the 30-40% range overnight. This will allow the pack to stabilize at each end which is when the system takes its readings from the battery.

 

[Allistah]

Now that I've had some time to think about this, I'm going to go this route after my call with the SC today. I'm going to alternate between charging at night right when I get home and charging in the morning. This will give the battery pack "time to stabilize/settle" and for the system to take proper measurements while its at 90% and when it's low. Right now its at 26%, so it will sit there all night until some time in the morning when it's time to start for a 6am departure. Alternating back and forth to see where this gets me.

Also, they refused to do any CAC reset for me. He was shocked that I even knew what a CAC reset was. lol

 

[KenC]

No surprise. The SC will only do a CAC reset if your battery is out of spec, like 30% range loss. Otherwise, they'd have thousands of owners begging for CAC resets after 4 miles of BMS drift.

 

[toolman335]

Keep us posted! Thank you for your data.
This time of the year I get the regen warning every morning. I park in an unheated garage that is 45-50 degrees and I still get the warning.

Did the guy say why it doesn't measure current when the battery is low? I definitely did not know that.