Questions after 1 Month

[LukeT]

Re: battery health. I've been playing around with lithium cells running home made electric bikes, a home made electric motor cycle and a home made electric boat for around 10 to 15 years now. I've had loads of battery failures (none spectacular, barring some nickel metal hydride cells that blew up right at the start of my experiments with electric power).

I've learned few things about battery life. The biggest single factor is the quality of manufacture, everything else pales into insignificance, as manufactured quality determines calendar life, and calendar life can easily be the determining factor as to how long cells may last in some applications, especially EVs, where cycle life is really no big deal.

The next biggest factor is depth of charge and discharge. Cycling between a nominal 0% capacity (lower voltage cut-off typically ~3.0 to 3.2 V) and 100% (nominal high voltage cut-off ~4.2 V) will shorten life dramatically, maybe down to just a few hundred charge-discharge cycles, with a loss of capacity with each charge-discharge cycle. Cycling between 10% and 90% massively increases cycle life, maybe by as much as a factor of 10 over the 0% to 100% case. Cycling between 20% and 80% further increases cycle life, but not by much, maybe a 10% to 30% improvement over the 10% to 90% case.

Putting cycle life into perspective, for a car that uses ~275 Wh/mile on average, that has a 75 kWh battery pack and is driven around 10,000 miles a year, then that's roughly equivalent to around 37 full charge-discharge cycles a year, or maybe 46 10% to 90% charge-discharge cycles a year.

The cells I've been using in my home made stuff are nowhere near as good as Tesla's 2170 cells, and yet they will easily give around 1,000 cycles when run over a 10% to 90% SoC (State of Charge) range. I'd be very surprised if Tesla 2170's can't manage at least 2,000 cycles under such a charge-discharge pattern, and suspect they may be capable of maybe 3,000 cycles when used like this.


1,000 cycles, over a 10% to 90% charge-discharge depth, is equivalent to around 21 years at 10,000 miles per year. The cells will die from old age before they die from any cycle life limitation. At a life of 2,000 cycles, that age is doubled, to 42 years at 10,000 miles per year, or 420,000 miles.

In general I'd not be the slightest bit worried about charging to 90% all the time, as the chances are the cells will die of old age long before cycle life starts to have any real impact. The only proviso to that is that if the car is to be stored for a long time (several months) then I'd set the maximum charge to 75% and leave the car plugged in. It will periodically recharge to 75% every few days, to allow for vampire drain, and the 75% SoC point is about optimum for long term storage.

@Jeremy Harris would you infer from your experience that the fastest dc charging, say from a Supercharger, is materially worse for battery life/health than 3/7/17kw ac?

Should using them a lot be considered sub-optimal in the way frequent >>90% is?

 

[CertLive]

To turn heated seats straight off press slightly longer with your finger and let go it goes from 1 to off then.

 

[Glan gluaisne]

@Jeremy Harris would you infer from your experience that the fastest dc charging, say from a Supercharger, is materially worse for battery life/health than 3/7/17kw ac?

Should using them a lot be considered sub-optimal in the way frequent >>90% is?

Rapid charging from a supercharger isn't really that rapid in terms of battery C rate. Even older lithium chemistry cells, without active temperature control, will happily charge at 2C, and 2C on a Model 3 LR or P equates to roughly a 150 kW charge rate.

I've charged smaller packs at 4C without seeing any significant problems, other than the cells getting a bit warm and the charge tailing off after about ten minutes or so as the cell voltage rises. That equates to about 300 kW for a Model 3 LR or P.

The Model 3 will discharge at peaks of well over 4C. Just had a quick look through the Teslamate data for mine and it seems I fairly frequently hit around 315 kW to 320 kW peak discharge on trips, although the average is a great deal lower.

 

[CertLive]

But am I right in saying 150kw split up over 1000,s of cells in your car charging at once is really not all that big of a deal when charged at optimum temperatures which the car attempts to achieve where possible if not it would slow it down? I think a lot of people see it as baking your battery's with such high volts. Even 250kw tapers off quick not just because if gives most people the miles they need to move on fast but again its not over-volting each individual cell no?

 

[Jason71]

I'm actually not completely convinced that it does use the mains supply for preheating. .

I am using Octopus go and have charging scheduled to start at 12:30am to take advantage of the cheaper rate, and it usually finishes charging before 4:30am unless i've done lots of miles the day before.

Do you have a dedicated charger? if so the charger logs should tell you something.
If you are using UMC then it cannot supply enough current for a full preheat so it will use that and the battery

 

[Glan gluaisne]

But am I right in saying 150kw split up over 1000,s of cells in your car charging at once is really not all that big of a deal when charged at optimum temperatures which the car attempts to achieve where possible if not it would slow it down? I think a lot of people see it as baking your battery's with such high volts. Even 250kw tapers off quick not just because if gives most people the miles they need to move on fast but again its not over-volting each individual cell no?

Yes, it's why using the C rate is key. This is the charge/discharge rate relative to the cell capacity, or pack capacity, and is the same no matter how many cells there are in the pack.

Cells are typically rated for a maximum discharge C rate, and may or may not be rated for a maximum charge C rate. We don't know the specs for the 2170 Tesla cells, but this is the spec for a fairly big brand 2170 cell that's probably not as good as a Tesla cell: SAMSUNG-INR21700-40T-Datasheet.pdf

From that the cell capacity is given as 4 Ah nominal, with a max discharge rate of 8.75C (35 A) without temperature control, or 11.25C (45 A) with temperature control. Max charge rate isn't given, just the rated charge rate of 1.5C (6 A). In practice, the max charge rate tends to be similar to the max discharge rate, although the main challenge isn't then temperature control, it's ensuring that no cell in the pack exceeds the allowable max set terminal voltage (probably around 4.1 V or so for 90% SoC). The pack charge rate has to be reduced as soon as the first cell group in the pack start to approach the charge cut-off voltage.

This latter point is the reason that rapid charging slows down the charge rate pretty dramatically as the SoC rises. The battery management system cannot easily bleed off excess charge from the highest voltage cells in the pack at high charge rates, as the power dissipation is just too high. Cell balancing needs a lower charge current, just to keep dissipation within the BMS to a reasonable level.

 

[tsh2]

Do you have a dedicated charger? if so the charger logs should tell you something.
If you are using UMC then it cannot supply enough current for a full preheat so it will use that and the battery

An afternoon top-up charge (30A), where I switched on pre-heat half way through. With an ambient temperature of about 5C, and the car starting cold (I just moved to the charging point).

 

[RichieC]

No. It uses shore power any time it’s connected, irrespective of the charge state shown in the app.

The difference is when the app says “charging” then that takes priority over battery warming. it seems to decide range is your priority, not a warm battery. Which is logical.

i did a test a couple of weeks ago while it was charging ( only on the TMC @ 2kW). I turned on Climate but the power use (total of battery + mains) was only 2kW - enough to run the heaters but not drain the battery - even after 10 minutes.
As soon as I hit "Stop Charging" the power use ramped up to 10kW ( 8kW from the battery, 2kW from the mains) to switch to prioritise warming the battery & cabin

I think this could actually be a bug in the model 3 software.

The issue seems to be, if you have charging scheduled and completed (stopped charging). It then no longer seems to use shore power for preheating - it uses the battery. If you restart the charging or remove the schedule, it then works as expected!

So the schedule option seems to schedule not only charging, but the use of all mains power.

 

[Joelly]

I think this could actually be a bug in the model 3 software.

The issue seems to be, if you have charging scheduled and completed (stopped charging). It then no longer seems to use shore power for preheating - it uses the battery. If you restart the charging or remove the schedule, it then works as expected!

So the schedule option seems to schedule not only charging, but the use of all mains power.

In my view, if I’ve told the car to stop charging, in my head I have told it to stop drawing power (stop spending money) in that case I wouldn’t want the car to draw power (money). Therefore I see it acting logically. If you want to start drawing her power then set a schedule.
A good option would be preheat with onshore power Y/N or something similar. Then we all win.

Finally, is there any advantage to pre heating either onshore power v battery? The only thing I can think of is less range in the car.

 

[MrBadger]

If you leave car plugged in after charging is complete, it will still periodically draw power. Its only for a very short period, but still a power draw.
10A charge and topups.

 

[MrBadger]

I'm not convinced think that Tesla think that supercharging is benign otherwise why impose ever increasing charge limits after certain models have exceeded a particular supercharging kWh threshold ?

 

[Glan gluaisne]

I'm not convinced think that Tesla think that supercharging is benign otherwise why impose ever increasing charge limits after certain models have exceeded a particular supercharging kWh threshold ?

I think you're right about it not being benign, but what we don't know is how the BMS works, and what impact rapid DC charging has on pack cell balance. I suspect that the cell balance issue is key to the limits that Tesla impose, as it seems that some cell groups in a pack are more likely to get out of balance following repeated rapid charges than would be the case when AC charging at a slower rate.

The BMS seems to periodically go into cell balancing mode at the end of some AC charges, probably doing some general housekeeping and maintaining the same SoC across all cell groups. It doesn't seem to do this on every AC charge, AFAICS, as I've had a few where the charge current hasn't tapered towards the end of the charge, and that charge current tapering is probably a good indication that the BMS is in balancing mode, I think.

 

[tsh2]

I suspect that the cell balance issue is key to the limits that Tesla impose, as it seems that some cell groups in a pack are more likely to get out of balance following repeated rapid charges than would be the case when AC charging at a slower rate.

I think this is a good point. It probably makes sense to plan a slow AC charge to ~95% in advance of any trip where you plan to rapid charge to any significant extent - it's over a month since I last charged to over 90% (and I think it would be hard for the BMS to balance at all over this time even if the algorithm wanted to).

 

[Gatsojon]

I have nothing to add here except to note it is a fascinating thread and makes me realise I am the bottom of a pretty steep learning curve

 

[Drmouse]

I'm actually not completely convinced that it does use the mains supply for preheating

Anecdotally, I can say I'm pretty sure it uses the mains (at least in part) to heat the car when plugged in.

My reasoning is fairly simple. I have a smart meter for Octopus Go. If I watch the instant consumption, it jumps from around 200W to around 7kW when I turn on the heat, and drops back to 200W soon after stopping it. So, it's definitely drawing the max power available from the mains to heat the car, but may also use the battery if it's using more than 7kW.

 

[UrbanSplash]

Anyone know what these stickers were for on the wheels? (On the rim of the alloy, under the aero covers)

 

[Adopado]

Anecdotally, I can say I'm pretty sure it uses the mains (at least in part) to heat the car when plugged in.

My reasoning is fairly simple. I have a smart meter for Octopus Go. If I watch the instant consumption, it jumps from around 200W to around 7kW when I turn on the heat, and drops back to 200W soon after stopping it. So, it's definitely drawing the max power available from the mains to heat the car, but may also use the battery if it's using more than 7kW.

On the Zappi charge point you can see the charging rate on a small screen ... there is no doubt, the car takes the power from the mains, not the battery if it's plugged in. The only reason it wouldn't is if the charge point is effectively "off". Perhaps some charge points have their own timer that would switch it completely off after a set time .. in that instance then no further power would be available until reset.

 

[CertLive]

Yes, it's why using the C rate is key. This is the charge/discharge rate relative to the cell capacity, or pack capacity, and is the same no matter how many cells there are in the pack.

Cells are typically rated for a maximum discharge C rate, and may or may not be rated for a maximum charge C rate. We don't know the specs for the 2170 Tesla cells, but this is the spec for a fairly big brand 2170 cell that's probably not as good as a Tesla cell: SAMSUNG-INR21700-40T-Datasheet.pdf

From that the cell capacity is given as 4 Ah nominal, with a max discharge rate of 8.75C (35 A) without temperature control, or 11.25C (45 A) with temperature control. Max charge rate isn't given, just the rated charge rate of 1.5C (6 A). In practice, the max charge rate tends to be similar to the max discharge rate, although the main challenge isn't then temperature control, it's ensuring that no cell in the pack exceeds the allowable max set terminal voltage (probably around 4.1 V or so for 90% SoC). The pack charge rate has to be reduced as soon as the first cell group in the pack start to approach the charge cut-off voltage.

This latter point is the reason that rapid charging slows down the charge rate pretty dramatically as the SoC rises. The battery management system cannot easily bleed off excess charge from the highest voltage cells in the pack at high charge rates, as the power dissipation is just too high. Cell balancing needs a lower charge current, just to keep dissipation within the BMS to a reasonable level.

In your view what would be the maximum theoretical C rate of the SR+ 54KW battery pack? Its at 175kw now via update which is still blisteringly fast. I ask this because its something that was never stated on the website, more of V3 is 250KW but is that not achievable on the smaller pack? I was just never informed my car was charging speed limited compared to a more expensive model.

 

[Keeper]

View attachment 503429 Anyone know what these stickers were for on the wheels? (On the rim of the alloy, under the aero covers)

Yes, these are placed on the wheels at manufacture to ensure the car is driven respectfully slowly. When peeled off they allow the car to attain maximum speed. Technology - its marvellous!

 

[Glan gluaisne]

In your view what would be the maximum theoretical C rate of the SR+ 54KW battery pack? Its at 175kw now via update which is still blisteringly fast. I ask this because its something that was never stated on the website, more of V3 is 250KW but is that not achievable on the smaller pack? I was just never informed my car was charging speed limited compared to a more expensive model.

I honestly don't know, as Tesla don't publish any details about their 2170 cells. All we can go on is the sort of performance that other big brand 2170 cells makers are publishing, and then guess that the Tesla cells will most probably be a bit better.

In the case of the Samsung cell data linked to earlier, then they don't quote a max charge C rate, but it's reasonable to assume that it is probably around the same as the max discharge C rate, at least for short periods (say 70% SoC or so), so perhaps over 10C. 175 kW with a 54 kW battery pack is only 3.24 C, which isn't that high. Maximum discharge rate for the SR+ is 250 kW, which is of 4.63C, again not really very high.

As mentioned earlier in this thread, I doubt that the cell max C rate is the limitation on supercharging speeds, it's most probably related to the battery management system and specifically concerns over cell groups getting out of balance as a consequence of several rapid charges in succession. I would guess that Tesla are deliberately being a bit conservative, both because they don't want to see battery failures, and because they are still gathering data about how the battery pack and management system perform in the field. It may be that they change things in future in the light of the data they are collecting from the cars sold so far. It seems that they have done this in the past, which tends to support the view that they start off with conservative settings and then maybe change them in the light of experience.