Model 3 SR+ LFP Battery Range, Degradation, etc Discussion

[ucmndd]

It figures that it might struggle with this maths at either end of the SOC.

You’ve got it backwards - it’s easiest for the BMS to accurately measure SOC at the extremes because of the voltage curve of the cells.

LFP cells in particular have a very flat voltage curve in the middle of the range - meaning it can be hard to measure SOC based on voltage alone because the cell voltage difference between say 40% and 60% is so small.

This is why the car wants you to fully charge with some regularity - it provides the fixed reference point necessary to more accurately infer SOC in the middle.

 

[MikeRadio]

Tesla's wording has always bothered me:

If your vehicle is equipped with an LFP Battery, Tesla recommends that you keep your charge limit set to 100%, even for daily use, and that you also fully charge to 100% at least once per week.

That says: Charge to 100% every day and also once per week. Logically, that doesn't make sense.

I think they mean:

If your vehicle is equipped with an LFP Battery, Tesla recommends that you keep your charge limit set to 100%, even for daily use, BUT IF YOU DON'T DO THAT, THEN YOU SHOULD fully charge to 100% at least once per week.

That’s probably what they meant. Or keep the charge RATE set to 100%. It’s written very poorly and I’ve noticed there are other things that are written by engineers and are hard to decipher…

form7nstance with FSD… you get 5 strikes you are suspended for a week. Very simple. However the manual text and release notes text makes it sound complicated and confusing. It says…

The FSD Beta feature can only be removed per this suspension method and it will be unavailable for approximately one week.

that’s so strange…. Only by this method? Why would you assume there is another method. And it’s exactly one week to the minute you lose it not approximately. Just weird way of saying something very simple.

 

[MikeRadio]

I charge LFP battery regularly to 50% that’s all I need an no reason to stress the battery more than needed. .
once a week I go to 100% and since I don’t do that much driving in may stay between 50 and 100 for most if the week.
prior to a road trip I have it also go to 100%.

i have 23K miles now and the battery is and has been for a whole at 264 miles (down from 272) so only lost 8 miles.

 

[AAKEE]

I’ve noticed some latency when trying to drain the battery down to zero miles: I’ll drive the car to the top of the hill I live on and turn back down the hill when it reaches 0 miles. Pulling into my driveway after going down the hill, at first I was fooled by the display still showing 0 miles (even though I had just gone downhill for a mile). But by the time I had the car parked in the garage, it was registering the re-gen miles and showing that it now actually had 2 or 3 miles instead of zero. So it could easily take a couple of minutes or more for the bms to finalize its estimate of how much power remains…

When driving the displayed SOC is kind of a preliminary, estimated SOC.
This because SOC cannot be measured during the drive. (This goes gor other battety chemistry as well).

So the displayed SOC during a drive is calculated by that the SOC was known before the drive, and SOC x estimated capacity = energy in the battery.
Driving consumes energy so 1 kWh used would equal 2% on a 50kWh battery. (Disregard the buffer ting for now).

So using 10 kWh would equal 20% on that car. If the starting SOC was 80%, we now See 60% on the screen.
When parking and the battery get to rest the SOC can be measured (battery/cell voltage measured).
When the car get battery voltage it now know the real SOC and adjust the displayed SOC from this. If it increases after a drive, the BMS probably underestimates the capacity and vice versa.

For the LFP cars, the voltage measuring is harder specially in the 20-80% SOC range or so due to a flat voltage curve.
The initial SOC might be wrong because of this.

 

[AlanSubie4Life]

I’ve noticed some latency when trying to drain the battery down to zero miles: I’ll drive the car to the top of the hill I live on and turn back down the hill when it reaches 0 miles. Pulling into my driveway after going down the hill, at first I was fooled by the display still showing 0 miles (even though I had just gone downhill for a mile). But by the time I had the car parked in the garage, it was registering the re-gen miles and showing that it now actually had 2 or 3 miles instead of zero. So it could easily take a couple of minutes or more for the bms to finalize its estimate of how much power remains…

When driving the displayed SOC is kind of a preliminary, estimated SOC.
This because SOC cannot be measured during the drive. (This goes gor other battety chemistry as well).

This specific behavior is actually a well known feature of continuous large regen events. There is a hidden credit for regen that builds up, and it gets dumped out in approximately 3-rated-mile chunks. You usually won't see the regen credited continuously (I've never seen that happen actually).

I have no idea why it is done this way, but this is what I have observed in the past on very long downhill drives with continuous regen. I assume the observed behavior could change at any time with a software change, since it is non-physical.

I actually don't think this specific issue is due to the BMS estimation (actually dead reckoning) exactly, since the BMS isn't doing true estimation while driving, it's just "dead reckoning." as @AAKEE has described. Since the BMS is dead reckoning, and it definitely knows the regen has occurred, and knows with high precision how much has occurred, it SHOULD credit it continuously (one mile at a time, rounding to the nearest rated mile). But it does not. It's just the way Tesla sets it up. It waits for about 3 rated miles to build up, then it dumps it out and increases your rated range by 3 miles instantaneously.

I guess if you were using percentage you'd never notice this because it probably DOES credit each 1% increase due to regen, though I'm not sure. The thing that would be different than normal (presumably) is that rounding of the percentage would not occur as it normally does. (This may be the reason for the rated-miles behavior - for some reason they don't apply the fractional percentage increases to the rated miles until the percentage is actually increased...which doesn't happen as normal, because the percentage is not being rounded as normal).

This last paragraph can be ignored. It's too late and it's nearly incomprehensible as a result.

Anyway it's a curiosity. Easy to verify...if you have a VERY large sufficiently steep hill nearby to drive down.

 

[tgrobbo]

You’ve got it backwards - it’s easiest for the BMS to accurately measure SOC at the extremes because of the voltage curve of the cells.

LFP cells in particular have a very flat voltage curve in the middle of the range - meaning it can be hard to measure SOC based on voltage alone because the cell voltage difference between say 40% and 60% is so small.

This is why the car wants you to fully charge with some regularity - it provides the fixed reference point necessary to more accurately infer SOC in the middle.

View attachment 971701

You’ve got it backwards - it’s easiest for the BMS to accurately measure SOC at the extremes because of the voltage curve of the cells.

LFP cells in particular have a very flat voltage curve in the middle of the range - meaning it can be hard to measure SOC based on voltage alone because the cell voltage difference between say 40% and 60% is so small.

This is why the car wants you to fully charge with some regularity - it provides the fixed reference point necessary to more accurately infer SOC in the middle.

View attachment 971701

Easiest when it’s parked for a while yeah. But this is when driving innit. I was just hypothesizing was what subsequently confirmed by those with far greater knowledge that me The car ain’t measuring SOC from the battery when driving, it’s basically a very well calculated guesstimate lol.

 

[tgrobbo]

What we need is a few more battery tests on LFP cars. To date I’ve just seen that one example. I’m sure there’s plenty more results out there. Going to be really interesting to see how a proper battery test compares to the displayed range in these cars.

 

[hgmichna]

I charge LFP battery regularly to 50% that’s all I need an no reason to stress the battery more than needed.

Charge it to 70%, because the related calendaric degradation stress is the same from 40% through 70% for LFP batteries. It begins to rise above 70%.

Even that rise is very gradual, hence Tesla's recommendation to charge to 100%.

 

[kvanriper]

I am almost at the one year mark around the 26th of Sept. i started with 272 miles. 100% home charging. After 3,800 miles I am at 264 miles. So 8 mile lost. This seems in line with the others reporting here at around the one year mark.

I do have a second home and when gone i leave the car plugged in the garage at around 50% or so. Sometimes 60% then the night I am back I charge to 100% and drive the next morning. Any thoughts on what I do when traveling?

 

[ucmndd]

Any thoughts on what I do when traveling?

you’re doing exactly what you should be.

 

[Baluchi]

60,000 mile (96,600 km) update for my Sep 2021 SR+ LFP. The car is now 24 months old and was originally rated at 253 miles on a full charge. The Tessie app shows a battery capacity of 51.4 kWh (down 5.8% from my original Oct 2021 post of 54.6 kWh), and a max range of 238 miles (down 5.9% from my original range of 253 miles, and down one mile since my 55,000 mile update). I've had Tessie since my first day or two or ownership, so this data shows the entire life of the car.

According to the car's screen, I'm averaging 211 Wh/mi over the life of the car (same it was at the 55,000 mile update). Seasonal temps and driving style are HUGE when it comes to the car's efficiency. In the winter I can expect 240+ Wh/mi when it's below 30f, and in ideal temps (75-85f) I routinely manage under 200 Wh/mi on my 100 mile round-trip commute. Assuming I could tap into the current 51.4 kWh battery at my lifetime average 211 Wh/mi efficiency, that gives me a real-world range of 243.6 miles. You can see in this screenshot that Tessie is predicting a real-world range close to that calculation.

My charging is mostly Level 2 from a Grizzl-E delivering 24 amps on a 40 amp circuit in my garage. I charge almost every day/night due to a long commute, typically to about 60% a few times per week and a 100% once or twice a week. I do fast charge about twice per week on average because I'm a beekeeper and I make honey deliveries across a few counties most weekends...historically I've used both Superchargers and CCS chargers like Electrify America or Chargepoint, depending on which are more convenient at the time.

Driving the same distance in my old Ford Focus would've cost $5,682 in gasoline. My car is mostly sun-powered since I installed solar at my home. If I'd kept my 2012 Ford Focus (37mpg), I would've used 1,623 gallons of gas to travel these 60,074 miles. At about 19 pounds of CO2 per gallon, that would've been 30,848 pounds of CO2. Prior to getting solar, my carbon emissions were about 35% of what I would've released in my efficient little Focus. But now that I have solar, that number is near zero, and it's pushing the car's lifetime average down every day.

I had a long talk with an older gentleman at a Supercharger last weekend. He drives a '23 Model Y Performance, and was asking if I also get way less miles than the rated range. I told him I routinely get more than my rated range this time of year. According to him, he charges to 80% as suggested, and gets about 200 miles per charge. Which is less than I expect from a daily charge in my '21 Model 3 SR+. Just thought that was an interesting comparison. That being said, his MYP is $44K right now after tax breaks, which is an insanely good deal for that vehicle, IMO.

I'll try to post another update at 65,000 miles.

 

[mariod]

Our model 3 was charging at around 410km at 85% for years. Then one day recently it dropped 50kms to 360. For no apparent reason. We tried discharging to 5%, and then charging to 100%. Made no difference. Has anyone else noticed this type of behavior. The car is a 2018, with approx. 83,000 kms on it.

 

[3sr+buyer]

My 2022 Model 3 with LFP battery is about 20 months old now with about 25,500 miles. Scan My Tesla says that its current capacity is 57.7 kWh, 4.6% down from 60.5 kWh nominal full pack when new. The rated range display shows 261 miles when charged to 100%, though the drive tab on the energy screen shows it starting out at 260.5 miles (4.2% down from 272 miles). My driving usually gets slightly better than rated economy and range when there is nothing on the roof rack, but worse when there is something on the roof rack (not exactly surprising).

Mostly, I use AC charging at home to ~70%, with charging to 100% approximately every week. I recently took a ~1,500 mile road trip, recharging with SuperChargers most of the time after using up the initial charge. Most of my SuperCharger use was using the method typically recommended by ABRP, which is charging from ~10-15% to ~50-60% (unless more is needed to get to the next SuperCharger).

 

[3sr+buyer]

Our model 3 was charging at around 410km at 85% for years. Then one day recently it dropped 50kms to 360. For no apparent reason. We tried discharging to 5%, and then charging to 100%. Made no difference. Has anyone else noticed this type of behavior. The car is a 2018, with approx. 83,000 kms on it.

Range Loss Over Time, What Can Be Expected, Efficiency, How to Maintain Battery Health may be a better place to ask about degradation of a non-LFP battery. Most Teslas (including all in 2018) use non-LFP batteries. LFP batteries became available in Model 3 standard range cars starting 2020 in China, and were in all Model 3 standard range cars by 2022.

 

[mariod]

Our model 3 charged to 415kms consistently at 85%. Recently, it dropped to 360kms at 85%. Like in one day. Has anyone experienced this? The car is a 2018 long range.

 

[GtiMart]

Does it indicate that it will be 360km when it reaches 85% are are you now sitting at 85% with 360km indicated? You shouldn't trust the prevision, you should let it charge and see. Even better, charge it once to 100%, let it sleep a bit, and then see what it indicates.

 

[johnny_cakes]

I believe your original rated range was 499km at 100% SoC, is that correct? Then at 85%, your original rated range was 424km, so previously when the BMS was reporting 415km at 85% SoC, the BMS was estimating only a 2% degradation after 5 years. That seems unusually low, so it was probably over-estimating the remaining capacity and it made an adjustment to make it more accurate. 360km at 85% would put you at 15% degradation, which doesn't seem unusual for a 5 year old car. As a comparison, mine shows 14% degradation after 4 years.

You can try the technique described here to recalibrate the BMS, but there's no guarantee that it'll do anything: How I Recovered Half of my Battery's Lost Capacity

 

[smogne41]

This is probably not degradation, but just a BMS calibration issue (very common). The BMS lacks sufficient information to correctly estimate pack capacity. Try to bring your pack to both lower and higher SOC once or twice. Say down to 10% (or a bit below) and up to 100%. Don't leave the pack in those high/low charge states for very long (get back to 20-80% range quickly to limit pack wear), but on a trip is fine. My car also had a large calculated range loss recently as well, but after taking it on a trip (and going to high/SOC) it came back to about what I expect.

 

[EV_Travels]

As others have pointed out, miles (or km) is a very unreliable measure of degradation. Set your display to % and just worry if it won't charge to 100%.

 

[AlanSubie4Life]

As others have pointed out, miles (or km) is a very unreliable measure of degradation. Set your display to % and just worry if it won't charge to 100%.

It is pretty unusual for there to be enormous swings in estimated capacity. (Of course, it does happen.)

Usually the BMS is quite stable with a variation of a percent of two.

It’s not clear whether there was any error in this case (we don’t have any pictures or specific information on the scenario). If it was a LR RWD and it actually happened as described, it went from 71kWh to 62kWh overnight, which is a huge change. This large change can be a sign of imminent pack failure (hopefully not in this case).

Using rated miles is a pretty good way to measure degradation! You can meter your pack energy to crosscheck it but it will usually be within a percent or two.