First 500kW Tesla V4 Supercharger station under construction

[hemiarch]

It is all area under the curve and there are multiple ways to get to that. Typically in cylindrical cell vehicles (Rivian being the exception), you arrive low to maximize that. In pouch cell vehicles while there still is a penalty for arriving at a higher SoC, their design allows a bit more headroom 50-60-70%.

Yes. Absolutely agree with that. Whatever increases the integral the most is most important. If the engineers determine that’s the height of the peak, great. My mathematical intuition suggests otherwise in real world scenarios but that’s why there are smarter people than myself being paid to make these decisions.

Sponsored

 

[mongo]

It is all area under the curve and there are multiple ways to get to that. Typically in cylindrical cell vehicles (Rivian being the exception), you arrive low to maximize that. In pouch cell vehicles while there still is a penalty for arriving at a higher SoC, their design allows a bit more headroom 50-60-70%.

Those who always roll in with 30% vs 10% will have a slower time of it. Though the real penalty is leaving with more SOC than needed. Arriving high should mean less curve area addition needed and a shorter charge session.

 

[hemiarch]

So I’m curious, is this rationale something that’s factored in by the Tesla navigation software or is it purely distance/range numbers that drive what it recommends? Anyone know?
Anticipated charge speed that is.

 

[HaulingAss]

Not really. What I’m saying is that in real world terms of how people actually use the chargers, sustained higher speeds matter more than brief bursts of high speed.

But the original claim was that high peak speeds don't matter at all.

You might be a sociopath who pees in a bottle and charges for 10 minutes at a time but the reality is that most users are not.

How is that productive? I'm not a sociopath and I don't pee in a bottle. Peeing in a bottle is for people who charge for 45 minutes to an hour. I get a quick pee break every couple of hours.

My typical charging stop is right around 20 minutes, and that is only possible with those high peak charging speeds. Most of my miles are put on in the first half of that 20 minutes.

 

[hemiarch]

But the original claim was that high peak speeds don't matter at all.




How is that productive? I'm not a sociopath and I don't pee in a bottle. Peeing in a bottle is for people who charge for 45 minutes to an hour. I get a quick pee break every couple of hours.

My typical charging stop is right around 20 minutes, and that is only possible with those high peak charging speeds. Most of my miles are put on in the first half of that 20 minutes.

Ok. Duly noted. Reading back on my hastily written post, it sounded a little personal.
Obviously you’re not a sociopath. I was just saying that normal human behavior isn’t generally that efficient.
As far as the claim, I obviously understand it matters what the peak is. I just think it matters less than sustained performance in real world circumstances intuitively. I may of course be wrong but in my circumstance I care mostly about what can be achieved in about 25 minutes with less frequent stops because when I consider the time it takes to get on and off the freeway and unleash my kids and then get going again, 10-15 minutes is an incredibly inefficient stop.

 

[henchman24]

Yes. Absolutely agree with that. Whatever increases the integral the most is most important. If the engineers determine that’s the height of the peak, great. My mathematical intuition suggests otherwise in real world scenarios but that’s why there are smarter people than myself being paid to make these decisions.

And I'd add in comfort. Some people are not okay with pushing to under 5% or even 10% to maximize the it. Which I think some of that just comes from experience. In 2018, I wasn't super comfortable with it... but after years and years of it, I'm okay with leaving when Tesla says I'll arrive with ~5-6% (provided weather isn't crazy), burning that buffer off with speed, and arriving with 1-3%. I'm probably on the more extreme end of that though. If a person is not okay going below 20%, they will see almost no benefit to 500kW charging on the CT (only a minor amount if the amperage is an issue).

Ok. Duly noted. Reading back on my hastily written post, it sounded a little personal.
Obviously you’re not a sociopath. I was just saying that normal human behavior isn’t generally that efficient.
As far as the claim, I obviously understand it matters what the peak is. I just think it matters less than sustained performance in real world circumstances intuitively. I may of course be wrong but in my circumstance I care mostly about what can be achieved in about 25 minutes with less frequent stops because when I consider the time it takes to get on and off the freeway and unleash my kids and then get going again, 10-15 minutes is an incredibly inefficient stop.

As a rule of thumb with some safety net, in a Tesla you'll maximize 25 minutes of charging by arriving around 5%. With the current CT charge curve on a V3+ (V4 dispensers), you'll get around 5-65% vs 5-60% when the truck came out. If we estimate roughly on what I stated before on what a 500kW curve likely looks like. That same time frame is likely to be around 5-70% (all these numbers give or take a percent or so). Primarily with peak speed changes. IMO that is a substantial difference.

 

[henchman24]

So I’m curious, is this rationale something that’s factored in by the Tesla navigation software or is it purely distance/range numbers that drive what it recommends? Anyone know?
Anticipated charge speed that is.

Charge speed is factored in. You'll see it in real time if you navigate through an area with only V2s. When they fill up, you'll see the times dramatically shoot up. So if an 8 station V2 has 7 cars predicted on arrival, you'll see the charge time double. And each charge curve is anticipated by the software. On trips Tesla won't allow you to set charge station arrival percentage (only destination). In that, they target 10% arrival and more most stops they round the time up.

 

[hemiarch]

And I'd add in comfort. Some people are not okay with pushing to under 5% or even 10% to maximize the it. Which I think some of that just comes from experience. In 2018, I wasn't super comfortable with it... but after years and years of it, I'm okay with leaving when Tesla says I'll arrive with ~5-6% (provided weather isn't crazy), burning that buffer off with speed, and arriving with 1-3%. I'm probably on the more extreme end of that though. If a person is not okay going below 20%, they will see almost no benefit to 500kW charging on the CT (only a minor amount if the amperage is an issue).



As a rule of thumb with some safety net, in a Tesla you'll maximize 25 minutes of charging by arriving around 5%. With the current CT charge curve on a V3+ (V4 dispensers), you'll get around 5-65% vs 5-60% when the truck came out. If we estimate roughly on what I stated before on what a 500kW curve likely looks like. That same time frame is likely to be around 5-70% (all these numbers give or take a percent or so). Primarily with peak speed changes. IMO that is a substantial difference.

That definitely is a substantial difference. One other point to consider is that if changing the charging curve is software related versus building new superchargers l, one could produce an enormous improvement with just an OTA update as opposed to replacing thousands of superchargers everywhere. Both are obviously important over time but OTA updates are clearly more bang for the buck.

 

[henchman24]

That definitely is a substantial difference. One other point to consider is that if changing the charging curve is software related versus building new superchargers l, one could produce an enormous improvement with just an OTA update as opposed to replacing thousands of superchargers everywhere. Both are obviously important over time but OTA updates are clearly more bang for the buck.

The curve can be changed at any point by software. If you watch the temps in service mode though, you'll see that the cells are pushed pretty hard already in higher SoC. Not so much on the first stop or two, but if you hit 3-4 V3+ chargers in a row on a hot day, you'll see max cell temps at 40-60%. So they could potentially extend out a bit if they are okay with that wear, but after enough stops it would throttle back. Rivian actually has that issue (not on hot days even) where they allow you to hit max cell temps and once they are crested, they throttle back. Their 'improved' charging curve that made the news this summer just allowed the current to flow until max temp instead of limiting it earlier for more consistency. So for one stop they will do a lot better, but 2-3 stops in a row, they overheat (non-LFP ones).

As an early example of changing charge curve is to look at the early M3 charge curves. They have messed with that curve substantially over the years before the settled on today's around 2022 (and the 21+ cars have a worse curve, LG cells especially).

On the Model 3 note, I compare this V4 change here to that. If you look back you'll see the M3 in 2019 was a ~33/34 minute 10-80% vehicle and the swap to V3s cut off ~4 minutes. That was really with peak speeds too. They tweaked the curve a lot on those cars, and to this day those early Panasonic 3s charge really well compared to the ones today (about 27-28 minutes 10-80%)... but it shows the substantial difference peak can make.

https://www.motortrend.com/news/teslas-v3-supercharger-tested

IMO from watching my truck charge many times at public 800v and 400v V3+ stations, there is a little wiggle room below 30% and especially 23% for a bump up. The cells are plenty fine temp wise, but pin temps get hot (signal the amperage is a problem). I think we will see 325kW at 21-22% in the end. There is however, very little room for improvement >30%, any increase there is likely to create an inconsistent experience with temperature derating like Rivians.

 

[mongo]

V4 saves up to 2.25 minutes for a 44% charge.

 

[Gaximus]

Not very useful when speeds drop below 100kw at around 50%

So the speed before 50% isn’t useful to you? I rarely charge over 50% at a stop on long trips. I try to keep my range between 6% and 50%

 

[henchman24]

V4 saves up to 2.25 minutes for a 44% charge.

In the realm... like the early V3 days, we probably see tweaks for the next 12-18 months before they settle into a curve Tesla feels comfortable with.

 

[Urander]

 

[HaulingAss]

That definitely is a substantial difference. One other point to consider is that if changing the charging curve is software related versus building new superchargers l, one could produce an enormous improvement with just an OTA update as opposed to replacing thousands of superchargers everywhere. Both are obviously important over time but OTA updates are clearly more bang for the buck.

Speeding up the charge curve is just as much a matter of how much a manufacturer values battery longevity as it is the exact chemistry and construction of the battery (and of course the effectiveness of the thermal management system). All manufacturers make a value judgement as to how much additional battery degradation is worth a one minute shorter charge time. At some point they draw the line. My observations suggest many of them of them only look at the term of the battery warranty and actual monetary cost to the company to offer faster charging. Other manufacturers are more desperate to prove they have some EV advantage over Tesla, Tesla has little to prove. I think Tesla is not keen to push for fatter charge curves, even when they know it won't push up warranty expenses substantially, because they don't want millions of Tesla cars dying 2-4 years out of warranty. There is economic cost when too many die under warranty, but there is future reputational damage when too many die 2-4 years out of warranty.

Using first principles thinking, they likely value the battery, and their customers time, potential reputational damage down the road, and draw the line where it makes the most sense to them. Availability of a more ubiquitous charging network would lean towards a more conservative charge curve (because the motorist is unlikely to have to charge to a high state of charge just to feel comfortable). Tesla rewards shorter charger sessions and shorter legs between charges. This also maximizes fleet battery longevity. Tesla could easily fatten the charge curve with a few keystrokes, but at what future cost? Is it worth it? Every manufacturer will have their own way to parse the numbers depending upon what they value the most.

 

[henchman24]

Speeding up the charge curve is just as much a matter of how much a manufacturer values battery longevity as it is the exact chemistry and construction of the battery (and of course the effectiveness of the thermal management system). All manufacturers make a value judgement as to how much additional battery degradation is worth a one minute shorter charge time. At some point they draw the line. My observations suggest many of them of them only look at the term of the battery warranty and actual monetary cost to the company to offer faster charging. Other manufacturers are more desperate to prove they have some EV advantage over Tesla, Tesla has little to prove. I think Tesla is not keen to push for fatter charge curves, even when they know it won't push up warranty expenses substantially, because they don't want millions of Tesla cars dying 2-4 years out of warranty. There is economic cost when too many die under warranty, but there is future reputational damage when too many die 2-4 years out of warranty.

Using first principles thinking, they likely value the battery, and their customers time, potential reputational damage down the road, and draw the line where it makes the most sense to them. Availability of a more ubiquitous charging network would lean towards a more conservative charge curve (because the motorist is unlikely to have to charge to a high state of charge just to feel comfortable). Tesla rewards shorter charger sessions and shorter legs between charges. This also maximizes fleet battery longevity. Tesla could easily fatten the charge curve with a few keystrokes, but at what future cost? Is it worth it? Every manufacturer will have their own way to parse the numbers depending upon what they value the most.

Considering how the 2021 3/Y packs are holding up (not well), Tesla is now very wary of extending out the curve. The initial charging curve on the 2021 3/Y packs was basically the same as the 18-20 3s, with a different cell chemistry. Turns out that cell chemistry did not appreciate pushing 250kW into the mid 20s. They had some quick failures, and within ~3 months Tesla had nerfed the charging curve on those cars pretty dramatically. They figured out quickly those cells didn't like that, and ever since, whenever a new cell goes into a Tesla, they are very conservative on pushing the curve out... and they almost never push the cell temps high in the 20s and 30s anymore.

This is why a new Model 3 actually charges worse than a 2019. Still to this day, the failure rates on 2021 3/Y packs are far higher than any other year. There were some other problems with that pack, but the expensive lesson learned on fast charging them is one they are unlikely to repeat again. They'll go conservative and creep up to the curve they want.

Sponsored