Long Term Battery Costs, Fears, and Serviceability

[gottagofast]

While I havent seen any tendencies to this, I can asure that you do not loose real range.

It might be possible that the BMS get off track, and show less range but the battery itself lives better than ever at these low SOC ranges.

I always need longer drives at least ones a month so I havent stayed low for that long.
I might have been ~ 1.5 months at 55% and below without any higher charges but these have never showed lower range on my M3P. It has regained range by being low long time.
My MSP might “loose” a half percent/0.5 kWh after one month low. (But it always regains 0.5-1kWh that was lost from high SOC / supercharging / larger cycles.
This “recovery is probably real capacity recovered, and the possible loss after ~ 1 month would problably be the BMS getting slightly off.

I am quite sure you have less degradation from this year at 55% and below than you would have using the usual 80%.

We know that battery degradation is very predictable and does happen in a controlled rate from factors we understand.
Low SOC will preserve the battery.

You could perform a BMS calib if you like, but that only will restore displayed range, not the real drivable range. Guess you know how to perform a BMS calib?
No sentry, no perping in the app meanwhile.
The car needs to sleep at different SOC to measure the open circuit voltage (no load on the battery goves the perfect SOC reading).
The BMS would need to see high SOC and low SOC. You could do 90-100% SOC and let it sleep for three hours or so, and also single digit SOC for three hours or so as well. Several times might be needed and it could be good with sleep at 70-80% as well. (This means inducing slight extra degradation to reach your goal of a BMS calib)

I use 11% on the way to work and 11% on the way home so I was thinking of doing a

100% charge
Then sit at work at 89%
Then sit at home at 78%
Then the next day will be 67-56 etc

I’ll do this over 3 days without charging until I get below 20%

Should I do a 100% charge like every 3 months since I don’t have any long trips? Even if it’s just the estimator and not a real loss it still bothers me lol

 

[dhrivnak]

I use 11% on the way to work and 11% on the way home so I was thinking of doing a

100% charge
Then sit at work at 89%
Then sit at home at 78%
Then the next day will be 67-56 etc

I’ll do this over 3 days without charging until I get below 20%

Should I do a 100% charge like every 3 months since I don’t have any long trips? Even if it’s just the estimator and not a real loss it still bothers me lol

I would and I do. Below 80% the cells are unable to balance. So range is at the lowest calculated brick. A 100% charge calibrates the BMS and allows cells to balance. I am at 127,000 miles and am at 93% of new.

 

[AAKEE]

I would and I do. Below 80% the cells are unable to balance. So range is at the lowest calculated brick. A 100% charge calibrates the BMS and allows cells to balance. I am at 127,000 miles and am at 93% of new.

Earlier we needed -> 90% (more precise a specific voltage like 4.1V / cell) for balancing to start. Some people here state that the car can balance the pack at any SOC. I do not think it really looks like that.

My M3P pack always had ~4 mV imbalance even after 4-6 weeks between 55% and 10-30 before the charges. Imbalance was the same during long times at low SOC cycling.

Small imbalances do not matter and top balancing would modtmy be fone to maximize the energy.
The different capacity of cell bricks will cause the imbalanxe to increase at low SOC specially when newly balanced in top.

 

[gottagofast]

Earlier we needed -> 90% (more precise a specific voltage like 4.1V / cell) for balancing to start. Some people here state that the car can balance the pack at any SOC. I do not think it really looks like that.

My M3P pack always had ~4 mV imbalance even after 4-6 weeks between 55% and 10-30 before the charges. Imbalance was the same during long times at low SOC cycling.

Small imbalances do not matter and top balancing would modtmy be fone to maximize the energy.
The different capacity of cell bricks will cause the imbalanxe to increase at low SOC specially when newly balanced in top.

How long should it stay at 100% for the balance 1-2 hrs?

 

[AAKEE]

How long should it stay at 100% for the balance 1-2 hrs?

Theres two different things:

Calibrating the BMS:
Helping the BMS with the capacity estimate by showing the capacity (sleep @ high and low SOC)

Balancing the pack:
Discharging cells with higher Voltage (SOC) so the voltage spread is less.
Mainly performed at high SOC to maximize the energy on board.

The balancing at 100% does probably not change the capacity estimate noticable (or at all). A newly top balanced pack will have more imbalance at lower SOC, as the spread is reduced at high SOC and the capacity difference between bricks makes the voltage drop faster in small cells.

Time needed for a complete balancing depemds on the imbalance which we do not know without SMT pr something like that.

So, for the BMS calib we need enough time for the battery voltage to adopt to the Open Circuit Voltage. This takes a couple of hours to reach a approximate value.
The car needs to sleep for 1-2 or 3 hours until the voltage has reached the relaxed state the BMS needs.

 

[gottagofast]

Theres two different things:

Calibrating the BMS:
Helping the BMS with the capacity estimate by showing the capacity (sleep @ high and low SOC)

Balancing the pack:
Discharging cells with higher Voltage (SOC) so the voltage spread is less.
Mainly performed at high SOC to maximize the energy on board.

The balancing at 100% does probably not change the capacity estimate noticable (or at all). A newly top balanced pack will have more imbalance at lower SOC, as the spread is reduced at high SOC and the capacity difference between bricks makes the voltage drop faster in small cells.

Time needed for a complete balancing depemds on the imbalance which we do not know without SMT pr something like that.

So, for the BMS calib we need enough time for the battery voltage to adopt to the Open Circuit Voltage. This takes a couple of hours to reach a approximate value.
The car needs to sleep for 1-2 or 3 hours until the voltage has reached the relaxed state the BMS needs.

What % do I need to drop to get back my estimated range then since 100% doesn’t do anything for that

 

[AAKEE]

What % do I need to drop to get back my estimated range then since 100% doesn’t do anything for that

No, the cell balancing does probably not give any capacity back (looking at nominal full pack or energy graph calc).

Using 90-100% and letting the car sleep and also sleeping at low SOC can possibly help the BMS with the capacity estimate.
There is good information about BMS calibration here on TMC.

 

[gottagofast]

No, the cell balancing does probably not give any capacity back (looking at nominal full pack or energy graph calc).

Using 90-100% and letting the car sleep and also sleeping at low SOC can possibly help the BMS with the capacity estimate.
There is good information about BMS calibration here on TMC.

I’ll see if I can drop it to 15% tomorrow then leave it there for 3 hrs then I’ll do a charge to 100% for 3 hours that night

 

[dhrivnak]

Earlier we needed -> 90% (more precise a specific voltage like 4.1V / cell) for balancing to start. Some people here state that the car can balance the pack at any SOC. I do not think it really looks like that.

My M3P pack always had ~4 mV imbalance even after 4-6 weeks between 55% and 10-30 before the charges. Imbalance was the same during long times at low SOC cycling.

Small imbalances do not matter and top balancing would modtmy be fone to maximize the energy.
The different capacity of cell bricks will cause the imbalance to increase at low SOC specially when newly balanced in top.

From my Roadster days you needed to be above 82% to balance with 100% being better. Most lithium batteries have a fairly flat voltage curve from 20-80% so it is very challenging to balance as the voltage is nearly the same. But at high and low voltages you see the imbalance between cells. While somewhat true a pack can balance at any SOC for all practical purposes you need to be either high or low. Personally I think bottom balancing is best but that is hard to automate. So I use Tesla's recommendations to set at 80% and when needed, at least 6+ times a year I fully charge. At 127,000 miles and 293 at a full charge I seem to be be slightly better than average. So I will continue that way until I can get my hands on some real Tesla data.

 

[KenBlub]

I noticed I’ve been losing estimated range ever since I started daily charging to 50-55% and arriving home with 23-30%

2 months ago I was showing 145 miles of range at 50% charge, then it was 144 miles, now today it’s 143 miles, this wasn’t happening when I was charging to 70-80%

Car hasn’t been charged to 100% in a year, I might try it and see what happens and use my 100% charge over 3 days

Reading various comments on Reddit there seems to be a trend of people with low degradation charge their cars to 100% atleast a few times per year mostly for trips, but I haven’t been on any trips to charge to 100%, people that never charge to 100% seem to have more degradation

Good luck getting any useful information from Reddit. I find it to be a wasteland of useful and accurate information.

 

[Bouba]

Some thoughts after 10 years of me driving and living with EVs. They do pretty much everything we need them to do (save for fully support pull through towing). We still need better infrastructure, but that's coming. Purchase cost is still an issue. But I think there is really just one major issue that hasn't been dealt with or that doesn't have a clear solution yet.

The primary issues with EVs I think is concerns over how long the expensive batteries will last. If you think you will have to replace a $20K battery every 5 to 10 years, or if you really do, that's a huge problem for the long term prospects of EVs.

What do we know about the batteries at this point?

There is battery degradation, but it's reasonable. As long as you get a battery that is maybe 20%-30% more than you need, it will be sufficient long term.

Occasionally, a battery pack can fail, requiring a battery replacement. This is where the real problem is. Not degradation. But what are the frequency of failures, and how does that frequency change over time? Are the failures typically one cell that goes bad, or are all of the cells going bad just at different rates? These things we don't quite know (do we?).

What are acceptable long term maintenance costs?

We are off to a good start in that most EVs today come with 8 yr warranties on the battery systems. This gets us past paying off loans, after which we should be capable of affording expensive service if needed (or you can switch cars before the 8 yrs is up).

But I don't think it's good enough. It's still a major sore spot for EV value and adoption.

In general, I would think you would probably want a vehicle battery pack to last a confident 20 years. But thinking about this more deeply, we can come up with a metric. If a $20K pack should last 20 years, you could instead say long term maintenance costs on the battery systems should be less than or equal to maybe $1K per year. Is that reasonable? Or is that still too high? Should it be $500 per year? Coming up with this type of metric can help us figure out what solutions are acceptable for future EVs.

Achieving an acceptable service profile for EV batteries I think can go any number of ways.

1) Battery packs that are reliably lasting long enough to meet the service cost criteria.
2) Battery packs that are more modular and serviceable in way where battery costs will typically be smaller service or module replacements that cost a lot less than entire battery replacements. If a battery has a bad cell and is serviceable, you get that one cell or module replaced and get the vehicle back on the road, replacing only the parts that have actually failed. A lot of the cells in a bad pack are probably still fine.
3) Battery packs that are treated as separate products from the actual vehicle. This would open up options like being able to buy the car but lease the battery, or longer term service contracts for batteries (instead of a Tesla 8 yr 100K mile battery warranty, it could be a 20 yr unlimited mile service contract)

Furthermore, it would be a good idea to build battery architecture in a way where DIYers are generally sufficiently safe and capable of doing their own battery service. With some diagnostics assistance from software or readers, a DIYer should be able to determine what cell or module is bad, order a new one, disengage the contactors for the battery and the bad module, and replace that module themselves. This would put it on par with gas car repairs, and would probably make service visits cheaper too when having someone else do it.

I feel like we are headed more towards an option 1, but I would probably prefer option 2 or 3, as option 1 still comes with the bad luck lottery where some folks will have batteries that fail early.

Where do you think we need to be?

Here’s a poor analogy.....If, instead of an ice car, we compare a Tesla to an iPhone....when the battery goes, it cost what to replace ?...let’s say $40 ...and you drop it off and pick it up in one hour...all with very little stress. So if you multiply that by the difference in battery size with the car ....then in terms of cost, time and stress....the Tesla might be good value

 

[chinney]

Here’s a poor analogy.....If, instead of an ice car, we compare a Tesla to an iPhone....when the battery goes, it cost what to replace ?...let’s say $40 ...and you drop it off and pick it up in one hour...all with very little stress. So if you multiply that by the difference in battery size with the car ....then in terms of cost, time and stress....the Tesla might be good value

Mind you, I do hope that battery costs come down significantly and that the batteries themselves become less resource intensive. To be seen....

 

[Geir]

How to use the car/battery during the break-in period with a brand new car? Is there any advice to take into account regarding the charging procedure or using the car? Fossil cars must be driven carefully for the first 100km. Does this apply to electric cars?

 

[AAKEE]

How to use the car/battery during the break-in period with a brand new car? Is there any advice to take into account regarding the charging procedure or using the car? Fossil cars must be driven carefully for the first 100km. Does this apply to electric cars?

In general, batteries are cycled very gently during the manufacturing process until they perform like they should.
For the LG M-50 we have seen that they ”start low in capacity” and oncreases the capacity during the ~ first month. If this is a real low or a BMS off situation I do not know. Needs to be tested to know, but i think NMC has shown increase in capacity for gentle cycling when new.


I took delivery 1000km away for both my M3P 2021 and my current MSP 2023.
So both cars was driven highway for 1000 km with supercharging sessions to get home.
Both cars have had very low degradation.


MSP now at 21K km, one year since delivery and 2.3% degradation when comparing nominal full pack /full pack when new.

Looking at nominal full pack (97.1 yesterday) and nominal remaining (97.6) it actually are about where most packs top their numbers so from that, virtually no degradation… so using the Panasonic NCA full from start does not seem like a bad idea.

 

[Geir]

In general, batteries are cycled very gently during the manufacturing process until they perform like they should.
For the LG M-50 we have seen that they ”start low in capacity” and oncreases the capacity during the ~ first month. If this is a real low or a BMS off situation I do not know. Needs to be tested to know, but i think NMC has shown increase in capacity for gentle cycling when new.


I took delivery 1000km away for both my M3P 2021 and my current MSP 2023.
So both cars was driven highway for 1000 km with supercharging sessions to get home.
Both cars have had very low degradation.


MSP now at 21K km, one year since delivery and 2.3% degradation when comparing nominal full pack /full pack when new.

Looking at nominal full pack (97.1 yesterday) and nominal remaining (97.6) it actually are about where most packs top their numbers so from that, virtually no degradation… so using the Panasonic NCA full from start does not seem like a bad idea.

So I understand from what you say that it can be a plus for the NMC(A)batteri to get a gentle start.
Do both of your cars have NCA batteries?

 

[AAKEE]

So I understand from what you say that it can be a plus for the NMC(A)batteri to get a gentle start.

Not neccesarily gentle start but as it seems, the capacity grows initially, like they are not cycled to max capacity during the manufacturing process.

For Lithium polymer cells (not same chemistry) used for hobby purpose like RC Toys, the manufacturers at least before said to just use half the capacity (100-50% or so) for a couple of cycles a d not use full power. On the other hand hovering a RC Helicopter in a gentle way anyway used like 5-6C of power. This would be equivalent to 5-6 x 75-80 kW for a model 3 so ~ 430 kW, or about the absolute top a model 3 battery deliver.

I would say that however you use the EV battery, its ”nice and slow” enough.

If you start with low SOC cycles, like charge to 55% (or 60% for NMC) or less, the battery will reach the maximum possible capacity.
If you then continue to use low SOC, it will keep the capacity very well.
If you need to begin like I did both times for delivery, it will not matter.

Low SOC cycles gives the battery the possibility to recover capacity.

Do both of your cars have NCA batteries?

The M3P is sold, yes it had a Panasonic NCA (2170L cell).

The MSP has Panasonic 18650 NCA.

 

[gottagofast]

Low SOC cycles gives the battery the possibility to recover capacity.

Do you mean 20-30% or 5-10%

 

[AAKEE]

Do you mean 20-30% or 5-10%

Any below 55%.

 

[gottagofast]

Going by what Kyle says charging to 100% will restore the bms

https://twitter.com/x/status/1800258584380322132