V3 Supercharging Profiles for Model 3

[Zoomit]

Also, where does it say on Tesla.com that the SR charges slower? I know it's a matter of physics, but I didn't learn that the SR has a slower supercharging rate until I started browsing this forum. I cannot find a single place on the Tesla website that says the average/peak charge rates per configuration.

"Will all Tesla vehicles be able to receive a 250kW charge?
All Tesla models will benefit from the elimination of power sharing in V3s architecture. The peak rate each vehicle achieves will vary with size and age of battery pack, state of charge and ambient temperature conditions."

Ref: Supercharging

 

[darth_vad3r]

Wow! 240 Kw through a 4416 cell pack! That amounts to a peak power of almost 54 watts delivered to each cell. That’s over 12 amps per cell more than double the original peak charging current allowed by Tesla for the 2170. That means a peak power of about 1.5 watts dissipated in each cell and almost 7 kw total in the pack for a few minutes, the equivalent heating power of about 7 toasters. Pretty impressive cooling system.

240,000 W ÷ 4,416 cells = 54 W per cell
Um Fonguy your math is not correct.
We are charging the battery which I hope means at least most of the energy (240kW) is going into the chamicals in the cell and so not going to heat. I'd hope that 90% or more is going into the calls for use later. That leaves:
240,000 X 0.1 = 24,000 ÷ 4416 = 5.4W per cell.

Here:
Comparing the Battery with other Power Sources – Battery University

It says the charge efficiency is 99%. so that would be 0.54W per cell.

Huh? Fonguy said 54W delivered to each cell same as you, and 1.5W dissipated. You said 5.4W (10%) or 0.54W (1%). Seems same ballpark to me. 1.5W is using 2.78% waste heat. Do we know how efficient the Tesla Supercharger is in charging the pack?

 

[darth_vad3r]

So the smaller battery charges each cell at a higher rate and therefore must decrease the charge rate in less time to protect the battery..

No, for Teslas I think they want to stress the cells the same no matter the model range, so they limit the charge power for the smaller packs. e.g. 102-105kW “2C” limit on the ~50kWh SR+ at V2 chargers instead of higher for the LR.

At uncorked V3 if the LR peaks at X kW I would expect the SR to peak at 31X/46 kW (as long as X is less than 250 KW charger max and the LR rate is not charger-limit-constrained).

 

[darth_vad3r]

This might already be common knowledge, but someone over on Reddit took a snapshot of the specs on the back of a V3 supercharging station in Fremont: https://i.redd.it/urce8ra3mi331.jpg

Hmm max voltage of 500V and current of 350A only equals 175kW max, not 250..

The 200A has a higher temperature range. Is there some way they can combine two separate outputs of 350 and 200A (275kW at 500VDC)? I guess no and it’s just two different operating temps based on two current levels?

100% duty cycle rating could be lower than short term bursts. 175kW is 70% of the ‘advertised’ 250kW peak. At home on L1 or L2 we run at 80% of rated max amps and could easily max out at 110% for a couple minutes without melting anything or tripping a breaker.

 

[Feathermerchan]

No, for Teslas I think they want to stress the cells the same no matter the model range, so they limit the charge power for the smaller packs. e.g. 102-105kW “2C” limit on the ~50kWh SR+ at V2 chargers instead of higher for the LR.

At uncorked V3 if the LR peaks at X kW I would expect the SR to peak at 31X/46 kW (as long as X is less than 250 KW charger max and the LR rate is not charger-limit-constrained).

The 2170 cells are about 5,000 mAh (5.0 Ah) so 2C would be 10A. That should be no big deal for any Lithium battery.
For the long range pack with 46 cells per module, that's 460A charging. That's only about 176 kW. The LR pack is capable of a much higher charge rate.
The SR pack is estimated to have about 36 cells per module. Charging at 2C (10A) would be 360A or about 138 kW.
Is that what is being seen?

 

[craigcabrey]

The 2170 cells are about 5,000 mAh (5.0 Ah) so 2C would be 10A. That should be no big deal for any Lithium battery.
For the long range pack with 46 cells per module, that's 460A charging. That's only about 176 kW. The LR pack is capable of a much higher charge rate.
The SR pack is estimated to have about 36 cells per module. Charging at 2C (10A) would be 360A or about 138 kW.
Is that what is being seen?

You’re assuming pack voltage, which isn’t the case for each module. So power won’t be nearly that high.

 

[Big Earl]

@Zoomit here is a charging session for @JuiceBx's mid-range using software 2019.20.1

Time SOC Power
0 10 2
0.5 11 120
1 13 120
1.5 14 120
2 16 120
2.5 17 120
3 19 120
3.5 21 120
4 22 120
4.5 24 120
5 25 120
5.5 27 120
6 29 120
6.5 30 120
7 32 120
7.5 33 120
8 35 120
8.5 36 120
9 38 120
9.5 40 120
10 41 120
10.5 43 120
11 44 120
11.5 46 120
12 48 120
12.5 49 120
13 51 118
13.5 52 114
14 54 109
14.5 55 105
15 57 102
15.5 58 98
16 60 94
16.5 61 91
17 62 88
17.5 63 85
18 65 82
18.5 66 80
19 67 78
19.5 68 75
20 69 74
20.5 70 71
21 71 70
21.5 72 68
22 73 65
22.5 73 63
23 74 60
23.5 75 58
24 76 56
24.5 77 54
25 77 52
25.5 78 50
26 79 48
26.5 80 47
27 80 45

27 minutes to go from 10% to 80%, peak power of 120 kW sustained all the way up to 50%.

And here it is without time.

 

[darth_vad3r]

The 2170 cells are about 5,000 mAh (5.0 Ah) so 2C would be 10A. That should be no big deal for any Lithium battery.
For the long range pack with 46 cells per module, that's 460A charging. That's only about 176 kW. The LR pack is capable of a much higher charge rate.
The SR pack is estimated to have about 36 cells per module. Charging at 2C (10A) would be 360A or about 138 kW.
Is that what is being seen?

46-cell bricks in LR, 31-cell bricks in SR.

Not sure that 5Ah number is a reliable source or not?

176kW / 460A = 382.6V only ... 382.6 / 96 = 3.985V per cell (or per brick). Where are you deriving these numbers from?

Anyways, the burst charge rate is probably up to ~4C, and then it tapers quickly. 2C is sustainable for a longer time. V2 was already charging SR at ~2C peak for a sustained amount of time.

 

[darth_vad3r]

You’re assuming pack voltage, which isn’t the case for each module. So power won’t be nearly that high.

The supercharger output supplies power to charge all the modules in the pack simultaneously. Using pack voltage is a suitable way to calculate SC power. The on-board BMS, etc are responsible for distributing the total power from the SC to the individual modules / bricks.

E.g.
(a) Charging 1 pack a 100kW (say 400V x 250A), is the same thing as saying
(b) you are charging 4 modules at 25kW each (say 100V x 250A), which is the same thing as saying
(c) you are charging 1 brick at 1.04kW (4.167V x 250A) (assuming a 4 x 24 config, adjust accordingly for the LR with different sized modules, but you get the idea), which is the same thing as saying
(d) you are charging 1 cell at 33.6W (4.167V x 8A).

*Assuming 31-cell brick x 24-brick module x 4-module pack SR config (aka 31p96s, 31-parallel 96-series)

 

[Zoomit]

The 2170 cells are about 5,000 mAh (5.0 Ah) so 2C would be 10A. That should be no big deal for any Lithium battery.
For the long range pack with 46 cells per module, that's 460A charging. That's only about 176 kW. The LR pack is capable of a much higher charge rate.
The SR pack is estimated to have about 36 cells per module. Charging at 2C (10A) would be 360A or about 138 kW.
Is that what is being seen?

46-cell bricks in LR, 31-cell bricks in SR.

Not sure that 5Ah number is a reliable source or not?

176kW / 460A = 382.6V only ... 382.6 / 96 = 3.985V per cell (or per brick). Where are you deriving these numbers from?

Anyways, the burst charge rate is probably up to ~4C, and then it tapers quickly. 2C is sustainable for a longer time. V2 was already charging SR at ~2C peak for a sustained amount of time.

SR has 31 cells per "brick" [Tesla's term]. There are 24 bricks per module and 4 modules but those modules are just for packaging and don't have much significance from a cell electrical configuration perspective. The 96 bricks are in series with 31 cells in parallel for SR/+ and 46 cells in parallel for LR.

Each cell puts out about just under 5Ah. Jack Richard measured one at 4.458Ah. The Tesla EPA submissions had LR pack totals of 222.81, 221.81, and 225.56Ah; an MR pack total of 181.8Ah and SR pack total of 155.6Ah. Dividing those by 46, 37 and 31 respectively yields: 4.84, 4.82, 4.90, 4.91, 5.02Ah per cell. Obviously those are all higher than Jack Richard's result but that testing was presumably done under more controlled conditions. Assuming 4.9Ah is probably reasonable.

The highest C-rate occurs when pack voltage is low and amperage highest. The SR pack is known to charge at 105kW from low SoC. 360V is a reasonable charger voltage when low. That's 292A [105000/360] or 9.4A per cell or 1.9C [9.4/4.9] at low SoC. The amperage and c-rate decreases as the pack is charged.

If we ratio the LR 250kW charge rate down for the SR, the SR may charge at 168kW. To do that, the charger voltage is probably 380V at low SoC. That's 443A total or 14.3A/cell or 2.9C. This has not been seen publicly, just speculated.

The LR pack has been shown to charge at 250kW. Again assuming a 380V charger voltage at low SoC, that yields 658A or 14.3A/cell or 2.9C.

 

[Zoomit]

@Zoomit here is a charging session for @JuiceBx's mid-range using software 2019.20.1

27 minutes to go from 10% to 80%, peak power of 120 kW sustained all the way up to 50%.

Thank you @Big Earl and @JuiceBx! Those graphs seem to have more precision than the table of data but nevertheless I plotted that session over the previous MR sessions seen in this thread: 150kW Supercharging for Model 3

Fundamentally, it looks like 19.20.1 MR charging profile has not been increased like the LR profile. This is disappointing and an ominous sign for SR/+ charging on 19.20.1. The MR curve does appear to have improved however. The rapid rise at low SoC is new as is the later taper. It will certainly yield a time savings but not as much as I hoped.

I'm also reconsidering my optimistic predictions for both V2 and V3. Under ideal thermal conditions, it might just be a linear taper from ~20% to 100%, without the dip between 20 and 50%. That might mean MR V2 charging at ~148kW up to 38% before it picks up the taper. We'll have to see but now I'm pretty confident 19.20.1 does not have the "final" MR charging profile.

 

[Big Earl]

Thank you @Big Earl and @JuiceBx! Those graphs seem to have more precision than the table of data but nevertheless I plotted that session over the previous MR sessions seen in this thread: 150kW Supercharging for Model 3

Fundamentally, it looks like 19.20.1 MR charging profile has not been increased like the LR profile. This is disappointing and an ominous sign for SR/+ charging on 19.20.1. The MR curve does appear to have improved however. The rapid rise at low SoC is new as is the later taper. It will certainly yield a time savings but not as much as I hoped.

I'm also reconsidering my optimistic predictions for both V2 and V3. Under ideal thermal conditions, it might just be a linear taper from ~20% to 100%, without the dip between 20 and 50%. That might mean MR V2 charging at ~148kW up to 38% before it picks up the taper. We'll have to see but now I'm pretty confident 19.20.1 does not have the "final" MR charging profile.

View attachment 419367

The time based graph was made from data points every 30 seconds. The SOC based graph was made with data points every 1%, so there are more of them. I didn’t put the latter data set into the code block; sorry.

 

[darth_vad3r]

SR has 31 cells per "brick" [Tesla's term]. There are 24 bricks per module and 4 modules but those modules are just for packaging and don't have much significance from a cell electrical configuration perspective. The 96 bricks are in series with 31 cells in parallel for SR/+ and 46 cells in parallel for LR.

Each cell puts out about just under 5Ah. Jack Richard measured one at 4.458Ah. The Tesla EPA submissions had LR pack totals of 222.81, 221.81, and 225.56Ah; an MR pack total of 181.8Ah and SR pack total of 155.6Ah. Dividing those by 46, 37 and 31 respectively yields: 4.84, 4.82, 4.90, 4.91, 5.02Ah per cell. Obviously those are all higher than Jack Richard's result but that testing was presumably done under more controlled conditions. Assuming 4.9Ah is probably reasonable.

The highest C-rate occurs when pack voltage is low and amperage highest. The SR pack is known to charge at 105kW from low SoC. 360V is a reasonable charger voltage when low. That's 292A [105000/360] or 9.4A per cell or 1.9C [9.4/4.9] at low SoC. The amperage and c-rate decreases as the pack is charged.

If we ratio the LR 250kW charge rate down for the SR, the SR may charge at 168kW. To do that, the charger voltage is probably 380V at low SoC. That's 443A total or 14.3A/cell or 2.9C. This has not been seen publicly, just speculated.

The LR pack has been shown to charge at 250kW. Again assuming a 380V charger voltage at low SoC, that yields 658A or 14.3A/cell or 2.9C.

Thanks, so I should have said “burst charge rate is probably up to ~3C, and then it tapers quickly. 2C is sustainable for a longer time. V2 was already charging SR at ~2C peak for a sustained amount of time.” Perhaps ~4C charging is in the future. V4 anyone?

I didn’t do any math for my 4C guess, I should have ... LR was supercharger/stall-constrained vs SR at V2 still.
102-105kW on 31p bricks = 151-156kW on 46p bricks.

So LR will likely see the biggest performance increase from V2 to V3 by a few % compared to the MR and SR.

BTW, I vote to rechristen the SR+ to the “SR”, IMO it is the original design, just with longer than originally ‘advertised’ range. The SR should be called the “SR-“, especially since the “-“ is different per country even

 

[Feathermerchan]

If the charge limit is 2C than there should be no difference between LR, SR, whatever range. As long as the charger can provide 2C charging. Charge time will be the same.

 

[darth_vad3r]

If the charge limit is 2C than there should be no difference between LR, SR, whatever range. As long as the charger can provide 2C charging. Charge time will be the same.

Basically, ya. It’s possible the MR and SR could be better at cooling the battery due to less cells so they may be able to sustain higher rates for longer before heating issues take over. That may factor in a smidge.

The new charge limit for V3 seems to be ~3C. ~2C seemed to be the old limit for V2 for SR and MR (for LR, 2C was over the V2 stall limit).

 

[Big Earl]

Thank you @Big Earl and @JuiceBx! Those graphs seem to have more precision than the table of data but nevertheless I plotted that session over the previous MR sessions seen in this thread: 150kW Supercharging for Model 3

Fundamentally, it looks like 19.20.1 MR charging profile has not been increased like the LR profile. This is disappointing and an ominous sign for SR/+ charging on 19.20.1. The MR curve does appear to have improved however. The rapid rise at low SoC is new as is the later taper. It will certainly yield a time savings but not as much as I hoped.

I'm also reconsidering my optimistic predictions for both V2 and V3. Under ideal thermal conditions, it might just be a linear taper from ~20% to 100%, without the dip between 20 and 50%. That might mean MR V2 charging at ~148kW up to 38% before it picks up the taper. We'll have to see but now I'm pretty confident 19.20.1 does not have the "final" MR charging profile.

View attachment 419367

New development ... it jumps straight to 120 kW, even at 6%.

Live reporting via text message.

 

[Zoomit]

While not V3 Supercharging, a French Model 3 Performance owner posted a few charge sessions on reddit from a 120kW SC and "350kW" CCS charger. I added them below as they reinforce what to expect on 350kW EA chargers in the US (assuming CCS1 adapter is sold) and the 120kW V2 Superchargers. The rapid rise is seen on all sessions and the late taper is very consistent, remarkably so above 65% SoC. Curiously, both 350kW IONITY sessions had a dip at 60% SoC that is unique from the other examples.

 

[israndy]

Well after experiencing the Superfast 25-minute V3 Supercharging previously uploaded I headed to my home Supercharger. I arrived preconditioned with almost the same range and was quoted a 30-minute charge to 80%

So I am not sure what the advantage is of having the V3 Supercharger. I got 620 Miles per Hour at my local Supercharger for the first time where I had previously only ever seen a straight 500 MPH. That's not that much below the 900+ MPH V3.

That did get me thinking however that likely the whole point of the V3 is not for the Model 3, although it's COOL to tell people you can charge at 1000 MPH, and certainly better tech than the S/X currently have. Its really going to be for the Pickup and the new Roadster. Put 4 of them together and charge a Semi. That's who this tech is for.

I was actually more thrilled at the update to the software controlling charging at all the V2 stations out there. Time for another road trip!

 

[Zoomit]

Well after experiencing the Superfast 25-minute V3 Supercharging previously uploaded I headed to my home Supercharger. I arrived preconditioned with almost the same range and was quoted a 30-minute charge to 80%

So I am not sure what the advantage is of having the V3 Supercharger. I got 620 Miles per Hour at my local Supercharger for the first time where I had previously only ever seen a straight 500 MPH. That's not that much below the 900+ MPH V3.

That did get me thinking however that likely the whole point of the V3 is not for the Model 3, although it's COOL to tell people you can charge at 1000 MPH, and certainly better tech than the S/X currently have. Its really going to be for the Pickup and the new Roadster. Put 4 of them together and charge a Semi. That's who this tech is for.

I was actually more thrilled at the update to the software controlling charging at all the V2 stations out there. Time for another road trip!

Great to see a 150 kW session with 19.20.1. Thanks for the video. The previous firmware had the mi/hr indication limited to 500 but the actual charging power represented the same 620-630 mi/hr at just under 150 kW. The indication limit has since been removed in 19.20.1.

The Model 3 LR does have lower recharge time from a V3 stall when starting at low SoC but you're right that future larger battery cars will rely on it more. The dedicated power for each V3 stall is very helpful for any car if the station is busy though.

I noticed you turned HVAC off around 50%, right at the taper beginning. You might have seen a few kW higher during the initial level part but that's hard to say definitively. Once again, this shows a very fast ramp-up and a taper consistent with the other examples.

 

[Feathermerchan]

Remember that when DC fast charging, the CAR is controlling the charger. So if two cars have the same version software and the chargers have similar capacity, and the same size battery, the charge curves should be similar.
Also remember what I said above about applying a charge current to a battery cell. The voltage always rises immediately. The higher the charge rate, the more the cell voltage rises. At some point, it reaches 4.2V and the charge current must be reduced to prevent battery failure. So you can have an large amount of charge current available but you can never charge above a certain rate.
The graphs above show how the shape changes with additional available charge current. The area under a charge curve more or less shows how much charge (kWh) has been added to the battery. You can see by the shape that adding more and more available charge power results in declining returns as far as overall charge speed. Notice that there is going to be almost no effect above 250 kW. Also notice that you can only make use of the full 250 kW rate from about 13-16% charge. Where as you can charge at 125 kW from about 8% to 55%.
I haven't made a lot of long trips but it seems like I have hit the SC with around 18 to 20%. At that I would expect to enter the chart at around 210A and see an immediate drop from there.
LiIon batteries with some chemistry tweaks could charge faster but would give up life or some other valuable characteristic.

In modeling (planes, helicopters, drones) we have LiPo batteries you can charge at 5C and that helps some compared to 2C but not that much and it does seem to shorten the life of the cells. And the cells with lower internal resistance (helps to charge faster, discharge faster and run cooler) are heavier. Probably due to additional conductive material.