-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Range Loss Over Time, What Can Be Expected, Efficiency, How to Maintain Battery Health
- Thread starter KK_RedM3
- Start date
My last post on this thread was in February/March (I think).
I was a bit warried and not happy because of the NFP displayed by SMT.
Battery is a 82.1kWh Panasonic E3DL on my Model 3 Performance 2021 (march) .
In February it was low to 73,9 kWh and it stayed to 74/73,9 for a full month ( or 5 weeks).
Tried to discharge to 0% , tried to go 100% , I tried to "move " a bit my routine ( 20% to 55% and let it sleep under 55%).
Nothing ... the 82,1 kWh battery (actual never showed more than 79,6) it was stabilized to 74 kWh.
here a pic of march 2022 (1 year old car and 22500 km):
Then I was convinced . The batt losts about 8-9%.
Then I don't
know what happend:
I was a bit warried and not happy because of the NFP displayed by SMT.
Battery is a 82.1kWh Panasonic E3DL on my Model 3 Performance 2021 (march) .
In February it was low to 73,9 kWh and it stayed to 74/73,9 for a full month ( or 5 weeks).
Tried to discharge to 0% , tried to go 100% , I tried to "move " a bit my routine ( 20% to 55% and let it sleep under 55%).
Nothing ... the 82,1 kWh battery (actual never showed more than 79,6) it was stabilized to 74 kWh.
here a pic of march 2022 (1 year old car and 22500 km):
Then I was convinced . The batt losts about 8-9%.
Then I don't
know what happend:
scottf200
Well-Known Member
My son had something similar with his 2018 TM3.
Seems around 2021.44.* OTA is when it went up.
Seems around 2022.4.5.* OTA is started going down again.
Last edited:
You have about a 6% degradation so looks normal. Did the Service Center tell you why you got that error?
AlanSubie4Life
Efficiency Obsessed Member
It’s a precise indicator of the energy your car estimates it has left. There is a very well defined amount of energy per rated mile displayed.
However, the BMS estimate CAN have errors and drifts for a variety of reasons, so a couple % variation is not unusual at all. The 4% shift example just above is more unusual, but certainly not unheard of.
It’s important that the BMS estimate be within a few % of the actual value, otherwise it risks stranding you due to thinking it can make a trip when it cannot.
So overall, the estimates are fantastically good and consistent.
That’s why it shows so many miles used when driving uphill - it takes an enormous amount of energy to move the car uphill. This value can be calculated (PE = mgh, and Google is your friend) and cross-checked via the energy values indicated by the BMS, and it will be found to match extremely well. In fact, this is what the car does when using the trip planner. It knows about hills. (It does not know about winds, in most places, or rain, etc…but it has algorithms, of limited efficacy, to limit error due to those factors.)
For a Model 3 2022 AWD, the energy per displayed rated mile is probably around 210Wh, (corresponding to 220Wh/rmi rated, 79kWh/358rmi), but this value can extremely easily be cross-checked by an owner (I haven’t seen the value calculated, and am using knowledge from the 2021, so this may not be exactly right).
Well, if this were the case...
...then this makes absolutely no sense. Does anyone consistently get 210-220 Wh/mi on road trips? The number it uses should be extremely conservative, something like 300 Wh/mi.
It uses the results of the EPA test which is what they are required to advertise on the Monroney sticker. Which isn't based on high speed highway driving.
AlanSubie4Life
Efficiency Obsessed Member
My point wasn’t that the constant accurately represents “normal” real-world use. It only accurately represents the (blended and scaled) results of the EPA test cycles (all five of them), according to the formula.
My point was that each “rated mile” is a parcel of energy, which is remarkably fixed and very consistent, no matter how you drive (and even if you’re not driving at all!). It’s a unit of energy, not distance. The size of each parcel is determined by the model year and variant of your vehicle, and nothing else (except wheel size, in one special case).
The trip planner knows what is the likely efficiency you will get, and will (usually) accurately predict the distance you can go. It’s not directly using the miles displayed to figure out whether you can make it. Note I said “it [car/BMS] risks stranding you thinking it can…” - obviously it is still very possible for a user to look at the rated miles and end up stranded, if they don’t use the trip planner…and it is also possible for a user to buy a car thinking it has a lot more highway range than it actually does.
For that part, my point was that if the BMS error were very large (in the wrong direction), the car could not correctly determine whether it could make particular trip segments. So the BMS estimate has to be quite accurate and large errors (more than a few %) are quite rare. This is a largely separate topic from how this estimated energy is denominated (rated mile energy content).
Last edited:
My M3P did get a 315 mile range on the EPA test.
The ”normal” energy in the battery is calculated slightly below a perfect new battery to account for variations in new cars battery capacity. They did set about 80.5 kWh for the M3P.
80500 Wh / 315 = 255.6 Wh/mile.
This is the energy content in each mile on the range on the screen.
The rate of used miles or km on the range indication will vary with the actual consumption. It is possible to drive (slow) below the rated consumption and thereby get a longer range than the on screen range.
The in-car range estimate is not required to use the same result displayed on the sticker AFAIK. Nor are car manufacturers prohibited from publishing a 2 dimensional range chart with speed on the x-axis and range on the y-axis.
The EPA test is just a minimum set of requirements for a sticker. If they want to make sure people don't get stranded, they can either give a more conservative estimate or a 2 dimensional chart.
Hello AAKEE, in your graph of battery degradation...are any of these actual Tesla batteries? And if they are..are they still connected to the Tesla BMS?
I understand that it’s only the chemistry that you are explaining...but I was just curious if Tesla’s engineering can out perform the chemistry
I have done a couple of discharges to try to lower my NFP. This because I was more or less certain that the BMS overestimated the capacity because of my charging schedule.
First one, late last summer down to 0.4% did not change the NFP noticeble.
Second one down to -1.8% did change the NFP down with about 1.5 kWh.
Third one -2% did set it down to the same NFP as the second one.
What I learned was that after the second one the NFP climbed back quite quickly from about 79.5 kWh to 80.5-81 kWh.
After the third one the NFP initially stayed at about the same value, then I realized that the charging schedule was very tight to the drive to the work. Earlier I always had one and a half hour or two with the 55% charge finished, letting the car sleep with 55%.
When changing the schedule back to give a sleep after the charge the NFP started the climb up immediately.
I also have found that not using Sentry at work increase the NFP further.
I am at the opposite side, normally having too high NFP, and the factors that keep the NFP up cannof course be used in that purpose.
- Let the car sleep as much as possible, after charging.
- Shut of Sentry when possible
In my case, these actions and my charging schedule will cause an overestimate of the capacity, but it might take some time.
The research reports for NCA chemistry mostly use Panasonic NCA cells. There is not very many manufacturer of NCA cylindric cells.
As Tesla do not exactly specify exactly what cells they use we can not be 100% sure most of the times.
These Panasonic cells is not cheap to buy, so we could imagine that if any changes is done, its most probable to keep the costs down.
This is an example of an research report where the Panasonic NCA cells was used:
NCA test
We recognize this picture from that one:
As all lithium batteries more or less behave in the same manner but to slightly different degrees, we can be sure that the basic principles will be correct.
Any cell will degrades from high SOC at high ambient temps, but as we do not have (as an example) an actual 2170L cell tested we can not say for sure if it will loose 4, 5, or 6 percent after 9.6 months at 25C and 80% SOC.
We know that the chemistry changes, cobalt is expensive and also not popular to use. Cobalt stabilizes the chemistry so reducing cobalt basically destabilises the battery, reducing life. But research by the battery manufacturers find ways to compensate for this. We could imagine that the goal is to reduce the cobalt without reducing the battery life.
To sum this up, the charts shown is probably quite close to the real life values we see. In some cases we perhaps have slightly less degradation and in other slightly worse.
I often get shot at in a Facebook forum, as the cars “have a BMS” so the graphs is invalid.
Also, “A Tesla battery always degrade 5% in the first 50.000km and then another 5% after another 250.000km”
First of all, the chemistry of the first Model S had about 6-10 times the cobalt of today.
Secondly, the graphs we see at teslafi and teslalogger.de do not really support the 5% and 10% theories.
There seem to be some warship about BMS:es.
A BMS can not threat a single battery cell better than it does by it own when it comes to calendar aging. The SOC x time will set the level of degradation. In theory a BMS could cool a battery with the help of an climate control, but this is not done(nor a good idea, in most of the cases)
A lot of cells together can be balanced, so they are at the same state of charge, or more exact, the same voltage at rest.
Balancing a battery pack is nothing else than discharging the “high cells” to reach the same voltage level as the lower voltage cells. Balancing is done at high SOC and not for long time. When the balancing is done, the car either sleeps with the battery disconnected with no input from the BMS, or Sentry is on and there is a small consumption but the BMS can not perform any battery “life saving” actions.
Summarized, a single cell at calendar aging( sleep) or cyclic aging (driving/charging) is always happy as it sets its own limits.
A lot of cells will need help by a BMS to experience the same happy life as the single cell would do. Without a BMS the lowest cell might be overdischarged or overcharged from regular use.
When the BMS is used, in perfect theory, each cell will have the same happy life as it would if it was on its own.
—> The BMS can not change the calendar or cyclic aging.
Tesla is clear that the range at the battery symbol is showing the energy content and not actual range. This is the choise Tesla have made.
The real range is shown at the energy app, and that one is good on estimating the actual range from the average of the last 10/30/50 km (and equivalent for miles).
There is some charts showing consumption vs speed on the net that seems to origin from Tesla. They are not in the manual but as you have the energy app, you basically do not need it.
OK I am 2000 miles into my M3LR and LOVING IT! I cannot make heads or tails of the energy widget (not sure why it breaks down into 5, 15 and 30 mile increments or what it means)
I'm trying to figure out my average range and wondering if I am doing it correctly.
My wh/mi over 2k miles is 251. I believe I divide this into 10000 (3.984) and then multiply by battery size which I think is 82 (82x3.984) = 326
So is 326 an accurate range estimate over these 2k miles or am I doing it incorrectly?
I'm trying to figure out my average range and wondering if I am doing it correctly.
My wh/mi over 2k miles is 251. I believe I divide this into 10000 (3.984) and then multiply by battery size which I think is 82 (82x3.984) = 326
So is 326 an accurate range estimate over these 2k miles or am I doing it incorrectly?
ATPMSD
Active Member
The energy screen shows the average consumption over the range you select, so the calculation is based on your driving over the last 5, 15 or 30 miles. Projected range is based on that average and based on the current SOC IF you select to display average range.
Hopefully you set a “trip” counter to zero when the car was new as this is the only real way to know the answer. If not, set one now.
AlanSubie4Life
Efficiency Obsessed Member
Correct, though clarification is that each displayed mile does not include the buffer (when you get to 0 miles you’ll have ~4.5% of that 80.5kWh left). So each displayed energy unit on the screen is 95.5% of that 255.6Wh, or about 244Wh. For a Performance 2021; other values apply for other vehicles.
Similar threads
