Tesla's 85 kWh rating needs an asterisk (up to 81 kWh, with up to ~77 kWh usable)

[shred86]

I had to show ignored posts to understand the context of this resurrected thread, though I guess I have to say thank you to @vgrinshpun at this moment for their lack of ability to receive other data as they caused you to post this additional info in response. Wish people would read more carefully, eg bit on 3 data on CAN bus now padding real telemetry, plugged in power for testing, etc.
Not sure whether to repeat guidance on how to ignore noisy few now: makes for calmer reading, though we wouldn’t have gotten this gem if so.

Thanks for posting again Jason.

I’ve read this thread pretty closely but I’m still unable to make sense of it. You seem to understand it, so can you explain what I’m missing? You allude to a couple things that I’m not quite sure what they mean:

“bit on 3 data on CAN bus now padding real telemetry” - you mean the fact wk057 mentioned the energy buffer previously reported is now combined with nominal energy full pack and nominal energy remaining? If so, I’m not sure how this affects the discussion as it’s a comparison between two Tesla documents that don’t seem to match, not CAN bus data.

“plugged in power for testing” - not sure about this one...

“etc” - what else are we missing?

Still waiting for an explanation from anyone on how Ben Sullins test was flawed, because my understanding of the kWh trip meter is likely wrong if that’s the case.

wk057 even mentioned in one of his post that “there are a lot of outliers with the 100 after calibration (I've seen 100 packs calibrated to 103 kWh, as well as some at 96, both relatively new).” Is it unreasonable to think there’s some variation in the Model 3 packs as well which would explain the difference?

 

[wk057]

wk057 has, multiple times, called you out, saying, "put up or shut up", and you still post the same false data.

^ This.

--

Since no one realized why the video @vgrinshpun posted was fake, take a closer look at the man's trip meter in the video. I do not see a "Since Last Charge" screen. I see a "Trip A", which I can make show anything I want. So this proves absolutely nothing. Going further, the guy literally stops at a supercharger, cuts cameras, and claims "Oh, we didnt charge here, but..." ... lol. Check the time lapse of the dash and you'll find the discrepancy.

If you could actually get 75 kWh out of a 3 pack, I'd expect dozens or more independent examples all over this forum, YouTube, etc... not just one Tesla fanboy. I'd personally give kudos for Tesla for properly underspec'ing the pack if this were true... but it is not.

I specifically asked for a "Since Last Charge" screen. It's also possible to make this show false numbers, but takes a lot more effort to do so along with some knowledge of the way this is processed.

As for why the EPA/CARB/whatever numbers are BS, I think posters above pretty much covered that.

And what's outdated about anything I posted? That specific document was from like, 6 months ago.

wk057 even mentioned in one of his post that “there are a lot of outliers with the 100 after calibration (I've seen 100 packs calibrated to 103 kWh, as well as some at 96, both relatively new).” Is it unreasonable to think there’s some variation in the Model 3 packs as well which would explain the difference?

This is the 100 S/X, not the Model 3. The outliers exist because of the massive cell count (12% more cells than the 85). All packs have outliers, but they're magnified somewhat exponentially by cell count. The Model 3 has half the cells of the 100 S/X pack, and has far far fewer outliers as a result. Surprisingly, the highest capacity LR 3 pack in the data I have is 75 kWh, and lowest is 72.6 kWh (< 500 miles values). The percentage of vehicles outside of +/- 0.5 kWh of 74 kWh within 500 miles is under 1%. This is due to the lower cell count, and less potential variance between cells in the same batch as a result.

 

[vgrinshpun]

Still waiting for an explanation from anyone on how Ben Sullins test was flawed, because my understanding of the kWh trip meter is likely wrong if that’s the case.

The reason Ben Sullin's test did not produce definitive results is because the run included significant elevation change first going up, then down (in excess of 1,000 ft), so on a way down there was significant regeneration component, in effect charging the battery. So total kWh since the last charge included this additional energy.

On another hand, Ben went for lunch, so car was sitting in the bright sun, and that resulted in additional drain of the battery, including likely activation of the cabitn overheat protection during the bright sunny day. This energy loss was not accounted for in the energy meter.

So all in all the result was inconclusive for the purposes of our discussion. I have to acknowledge that before my intitial post on this I did not watch the whole video, just skipped to the end for the summary.

 

[vgrinshpun]

To set record straight on the "put up or shut up", I did provided testing documentation, complete with links. In his attempts to dismiss the testing data @wk057 made several claims that are wrong so it is his turn.

 

[vgrinshpun]

Since no one realized why the video @vgrinshpun posted was fake, take a closer look at the man's trip meter in the video. I do not see a "Since Last Charge" screen. I see a "Trip A", which I can make show anything I want. So this proves absolutely nothing. Going further, the guy literally stops at a supercharger, cuts cameras, and claims "Oh, we didnt charge here, but..." ... lol. Check the time lapse of the dash and you'll find the discrepancy.

If you could actually get 75 kWh out of a 3 pack, I'd expect dozens or more independent examples all over this forum, YouTube, etc... not just one Tesla fanboy. I'd personally give kudos for Tesla for properly underspec'ing the pack if this were true... but it is not.

I specifically asked for a "Since Last Charge" screen. It's also possible to make this show false numbers, but takes a lot more effort to do so along with some knowledge of the way this is processed.

Yes, this makes a lot of sense - as mentioned, anything that does not come from @wk057 research and does not conform to his claims is fwaud.

 

[shred86]

^ This.

--

Since no one realized why the video @vgrinshpun posted was fake, take a closer look at the man's trip meter in the video. I do not see a "Since Last Charge" screen. I see a "Trip A", which I can make show anything I want. So this proves absolutely nothing. Going further, the guy literally stops at a supercharger, cuts cameras, and claims "Oh, we didnt charge here, but..." ... lol. Check the time lapse of the dash and you'll find the discrepancy.

If you could actually get 75 kWh out of a 3 pack, I'd expect dozens or more independent examples all over this forum, YouTube, etc... not just one Tesla fanboy. I'd personally give kudos for Tesla for properly underspec'ing the pack if this were true... but it is not.

I specifically asked for a "Since Last Charge" screen. It's also possible to make this show false numbers, but takes a lot more effort to do so along with some knowledge of the way this is processed.

As for why the EPA/CARB/whatever numbers are BS, I think posters above pretty much covered that.

And what's outdated about anything I posted? That specific document was from like, 6 months ago.



This is the 100 S/X, not the Model 3. The outliers exist because of the massive cell count (12% more cells than the 85). All packs have outliers, but they're magnified somewhat exponentially by cell count. The Model 3 has half the cells of the 100 S/X pack, and has far far fewer outliers as a result. Surprisingly, the highest capacity LR 3 pack in the data I have is 75 kWh, and lowest is 72.6 kWh (< 500 miles values). The percentage of vehicles outside of +/- 0.5 kWh of 74 kWh within 500 miles is under 1%. This is due to the lower cell count, and less potential variance between cells in the same batch as a result.

Makes sense with the larger cell count - thanks for explaining!

So basically Ben Sullin’s video/test might be invalid because 1. He never showed the “Since Last Charge” screen and 2. he allegedly is lying about not charging. So really we’re just questioning his integrity - is he intentional lying? I don’t know the dude, but that would be pretty sad if that was the case.

The other point I guess is you’re saying the test results that Tesla did on a dyno that was submitted to the EPA was invalid as well? That’s the part that I’m honestly confused about. Is there something flawed with how they’re performing the test? Or is it just based on an assumption that Tesla is being dishonest and lying about their rest results?

The reason Ben Sullin's test did not produce definitive results is because the run included significant elevation change first going up, then down (in excess of 1,000 ft), so on a way down there was significant regeneration component, in effect charging the battery. So total kWh since the last charge included this additional energy.

On another hand, Ben went for lunch, so car was sitting in the bright sun, and that resulted in additional drain of the battery, including likely activation of the cabitn overheat protection during the bright sunny day. This energy loss was not accounted for in the energy meter.

So all in all the result was inconclusive for the purposes of our discussion. I have to acknowledge that before my intitial post on this I did not watch the whole video, just skipped to the end for the summary.

Interesting. The times I’ve watched the kWh in the since last charge trip meter while going downhill, I’ve seen it roll back (i.e. decrease) because I thought it was subtracting the energy being regenerated to show you the true kWh utilized for the trip/since last charge.

 

[diamond.g]

@vgrinshpun in the EPA filings, they are saying they are using a dyno, but they don't say how they are measuring pack capacity. Seems like they are seeing how far they can go till empty, then multiplying it by a wh/mi metric, but how are they getting said wh/mi metric to do the calc?

 

[vgrinshpun]

@vgrinshpun in the EPA filings, they are saying they are using a dyno, but they don't say how they are measuring pack capacity. Seems like they are seeing how far they can go till empty, then multiplying it by a wh/mi metric, but how are they getting said wh/mi metric to do the calc?

They are measuring battery voltage and current - see pp 4 and 5 of the Tesla EPA Certification Application (01/28/2019) document I linked up-thread, where the test instrumentation and setup are shown. Note Battery voltage and current measurement instrumentation - Hoiki 3390-10, and pictures showing the hook-up.

Also see pp 21, 23 and 25 - they show actual test data measured per the above for each of the segment: Energy in Wh and Distance.

 

[vgrinshpun]

Interesting. The times I’ve watched the kWh in the since last charge trip meter while going downhill, I’ve seen it roll back (i.e. decrease) because I thought it was subtracting the energy being regenerated to show you the true kWh utilized for the trip/since last charge.

I am not hundred percent sure on this. If your interpretation is correct, Ben Sullin's data are legit. @wk057 assertion that he doctored results is preposterous.

 

[wk057]

I am not hundred percent sure on this. If your interpretation is correct, Ben Sullin's data are legit. @wk057 assertion that he doctored results is preposterous.

If it were legit, he could have shown it. It's not, so he did not. Tons of clues in the video of this.

One. I just want one "Since Last Charge" photo that shows 74+kWh. If the pack is 78 kWh, this should be the simplest task ever with hundreds of examples.

Post one.

 

[ucmndd]

Interesting. The times I’ve watched the kWh in the since last charge trip meter while going downhill, I’ve seen it roll back (i.e. decrease) because I thought it was subtracting the energy being regenerated to show you the true kWh utilized for the trip/since last charge.

It most certainly does roll backwards. I start my morning drive every day with a ~3 mile steep descent - my kWh used since last charge not only rolls back but goes negative.

That said, I think it's perfectly likely that number, over a long trip with lots of elevation change and acceleration/deceleration/regeneration is far from the most accurate measure of how much energy was actually consumed and being off by a few kWh is a very reasonable expectation.

I find no reason to accuse the individual of "cooking the books" or otherwise deliberately attempting to misrepresent the test. That's just silly. For the same reason, I also believe that Tesla has zero incentive to misrepresent the actual capacity of their packs in their own internal documentation and words from the CEO spoken to shareholders.

 

[vgrinshpun]

It most certainly does roll backwards. I start my morning drive every day with a ~3 mile steep descent - my kWh used since last charge not only rolls back but goes negative.

That said, I think it's perfectly likely that number, over a long trip with lots of elevation change and acceleration/deceleration/regeneration is far from the most accurate measure of how much energy was actually consumed and being off by a few kWh is a very reasonable expectation.

I find no reason to accuse the individual of "cooking the books" or otherwise deliberately attempting to misrepresent the test. That's just silly. For the same reason, I also believe that Tesla has zero incentive to misrepresent the actual capacity of their packs in their own internal documentation and words from the CEO spoken to shareholders.

If regeneration is accounted the way you described, I actually think that "since last charge" number is accurate. In such case Ben's video proves that @wk057 research didn't produce accurate data.

 

[ran349]

If regeneration is accounted the way you described, I actually think that "since last charge" number is accurate. In such case Ben's video proves that @wk057 research didn't produce accurate data.

In an earlier post, wk057 said his data went as high as 75 kWh, so doctored video or not, it still falls within his data set.

 

[vgrinshpun]

In an earlier post, wk057 said his data went as high as 75 kWh, so doctored video or not, it still falls within his data set.

Not really. His claim is that TOTAL battery capacity is 74 kWh. Assuming 2 kWh buffer would yield battery usable capacity of 72 kwh. Ben's video was showing 75 kWh. As discussed earlier, this is likely conservative figure because they took a break for lunch. So any drain during this stoppage is not reflected in the above 75 kWh. If his cabin overheat protection kicked on during the lunch, there likely was additional significant drain that was not accounted in 75 kWh

 

[shred86]

It most certainly does roll backwards. I start my morning drive every day with a ~3 mile steep descent - my kWh used since last charge not only rolls back but goes negative.

That said, I think it's perfectly likely that number, over a long trip with lots of elevation change and acceleration/deceleration/regeneration is far from the most accurate measure of how much energy was actually consumed and being off by a few kWh is a very reasonable expectation.

Good to know I wasn't going crazy.

Yeah, agreed there's probably going to be a little bit of variation but a few kWh would seem a bit excessive. Regardless, this is why I'm personally more inclined to believe the test results that Tesla submitted to the EPA using instrumentation (Hioki 3390-10) that was measuring battery voltage and current directly.

I find no reason to accuse the individual of "cooking the books" or otherwise deliberately attempting to misrepresent the test. That's just silly. For the same reason, I also believe that Tesla has zero incentive to misrepresent the actual capacity of their packs in their own internal documentation and words from the CEO spoken to shareholders.

Couldn't agree with you more. I'm not sure what Ben Sullen's would have to gain about flat out lying about his test. I get it, he has a YouTube channel and it's all about getting likes and subscriptions, but I don't see what there is to gain in this case.

For the same reason you mentioned, I'm doubtful Tesla misrepresented their dyno test data that they submitted to the EPA. As for their internal documentation, it wouldn't be the first time they've advertised something that is rounded quite excessively if you ask me (Model S 85?). At least in this situation it would be rounded down, lol. As for Elon's comments at yesterday's shareholder event, he did mention "about 75 kWh".

Not really. His claim is that TOTAL battery capacity is 74 kWh. Assuming 2 kWh buffer would yield battery usable capacity of 72 kwh. Ben's video was showing 75 kWh. As discussed earlier, this is likely conservative figure because they took a break for lunch. So any drain during this stoppage is not reflected in the above 75 kWh. If his cabin overheat protection kicked on during the lunch, there likely was additional significant drain that was not accounted in 75 kWh

If we're assuming Tesla is using the buffer that's reported in the Model S/X (4 kWh) the same way in the Model 3 (which I realize is apparently no longer reported as a separate value but rolled into the nominal energy readings), then it's been proven you can sometimes "use" the entire buffer. It's not guaranteed, as there's several reports of vehicles shutting down with nominal energy remaining still reporting a positive value, but just something worth noting. That being said though, the trip meter only shows energy used in motion so I would agree, there's definitely some energy not accounted for.

 

[wk057]

If regeneration is accounted the way you described, I actually think that "since last charge" number is accurate. In such case Ben's video proves that @wk057 research didn't produce accurate data.

Wrong. Did you look at the video? I see "Trip A", not "Since Last Charge". Nice try though.

In an earlier post, wk057 said his data went as high as 75 kWh, so doctored video or not, it still falls within his data set.

Not really. His claim is that TOTAL battery capacity is 74 kWh. Assuming 2 kWh buffer would yield battery usable capacity of 72 kwh.

Aside from the caveat above that this was not actually a "Since Last Charge" reading, at least we're on the same page here.

Ben's video was showing 75 kWh. As discussed earlier, this is likely conservative figure because they took a break for lunch. So any drain during this stoppage is not reflected in the above 75 kWh. If his cabin overheat protection kicked on during the lunch, there likely was additional significant drain that was not accounted in 75 kWh

See above. I see "Trip A", not "Since Last Charge".

Next, any power consumed while the high voltage system is engaged is accounted for. This is only not true for around firmware v6.x and below, which would have no bearing on the Model 3. This data isn't accounted for properly in the energy GRAPH still, in all versions on all vehicles, but it is properly accounted for in the trip meters.

(Detailed technical explanation of the above: The BMS keeps a total lifetime counter of kWh charged and discharged. Since v7.0, the trip meters just snapshot the odometer and these two values when reset. To get the trip data, it just does the math and displays the difference. Since the battery updates the kWh values for both counters any time the HV system is engaged, the trip meters will always have updated information, in motion or not. This is why the power usage in the trip meters will still be updated even if you reboot the system while driving.)

HOWEVER... the accuracy of this is greatly exaggerated at lower power usage. The measurements Tesla makes of charge/discharge is very accurate above 10A or so. Below about 10A (which is 3-4kW) the accuracy can vary greatly, as poorly as +/- 100%, and worse when near zero. So basic vampire drain while sitting, or mild cabin conditioning while idle/off, can be read as way less or way more usage than it really is (S, X, and 3). This gets even more exaggerated when you have slight inclines/declines on a trip where the power consumption/regen ends up near zero often. The sum of these errors can come out to huge variances in the trip meters over a long period.


But really, none of this matters, because the video that was posted was clearly not a 100% to 0% drive with even remotely controlled conditions. "Since last charge" was not shown. Don't get me wrong, I believe the guy probably had good intentions when he made the video. He probably didn't figure anyone would care if he charged for 5 minutes at some point if needed (pretty obvious this happened), which is why he couldn't show "Since last charge" and instead just showed "Trip A", which can be across any number of charges.

The route was also pretty much the perfect route that would keep the vehicle in the area of poor power consumption measuring for a large period of the trip (can see this clearly in the time lapse portions).

None of that really matters anyway, since it's not a "Since Last Charge" screen showing the info. So doesn't matter if he faked it or not, this is not admissible as any kind of proof since, based on the "Trip A" alone, there is no way to say that this was a 100% to 0% trip without charging. (We'll even ignore that analysis of the video pretty much confirms this was not a 100% to 0% trip, for the sake of argument.)

So seriously. One photo. One Model 3 that shows "Since Last Charge" any number of miles driven, 74 kWh or more used. Heck, even 72 would be impressive (despite that this wouldn't help your claims). Just one guys. So far all I've got is a "Trip A" from a Tesla enthusiast video. I mean, I'll post my Trip A. It shows like 6,000 kWh. Yeah, one charge! lol. Come on guys, get real.

No EPA/CARB/whatever testing matters at all (again, ignoring the facts of why this is the case, for the sake of argument) if not a single real world vehicle can achieve the numbers claimed based on its own calculations.

One legit photo of a Model 3 showing the "Since Last Charge" screen that backs up these claims. Until then, everything said by the people arguing against me on this should be taken with a plate full of salt.

 

[wdolson]

I'm no expert on Ben Sullins, but I have seen a number of his videos. Enough to get an idea of his motivations and style. He advertises himself as a numbers guy. He spent many years in the tech world crunching numbers. He is very skilled at taking a data set and building a picture from it.

While he's also a Tesla fan of sorts, he has also never hesitated to criticize Tesla when he felt they did something wrong. He has frequently criticized Elon's management style (and I mostly agree with him there, Elon is brilliant, but a poor manager of people).

But number crunchers are not necessarily that skilled at experimental design. Even the best experimenter makes errors in design and has to re-run a test when they discover their mistakes. I doubt Ben Sullins was trying to create any kind of controlled lab experiment with the posted video.

To do the sort of test @wk057 wants to see would take a professional organization with laboratory equipment and a dedicated test track. There are people out there equipped to do that sort of thing, but there are not many, and they are investing their time and resources in projects for paying customers.

There are also variations in any manufactured product to consider. Each cell is going to vary a little in it's total Ah capacity. There can also be differences between what can be made in a limited production run under laboratory conditions and what can be mass produced. Problems found in testing can also change the final specs of the final product.

Tesla's initial Model 3 packs used for testing might have been closer to 78 KWh. The chemistry used in those cells may have been found to have some long term problems, so they went into production with another chemistry that was more reliable, but a couple percent lower capacity. Another possibility is the lab made 2170's capacity may have turned out to be impossible to consistently achieve in the real world mass production, so they had to scale back the rated capacity a bit.

The fact is that Tesla's current internal documents say the Model 3 LR pack is 74 KWh. There is no reason at all to lie to themselves about it. If they wanted to test and select the best rated cells off the production line (more effort than it's worth), they probably could make a small number of 78 KWh packs, but it would be an expensive and pointless exercise.

 

[ucmndd]

If they wanted to test and select the best rated cells off the production line (more effort than it's worth), they probably could make a small number of 78 KWh packs, but it would be an expensive and pointless exercise.

Unless they wanted to game the EPA test cycle?

 

[wk057]

To do the sort of test @wk057 wants to see would take a professional organization with laboratory equipment and a dedicated test track. There are people out there equipped to do that sort of thing, but there are not many, and they are investing their time and resources in projects for paying customers.

Well, I don't quite agree with this. I'm actually not asking for much... a "Since Last Charge" display on a 3 showing any numbers that can match the 78 kWh claim (so at a minimum, 74 kWh discharged) would go a long way towards validating that claim... but as it stands, such a thing doesn't (and really can't) exist.

As I noted above, I don't think the Youtuber had bad intentions when making the video, but the fact is the data is from "Trip A" and not "Since Last Charge"... any number of things could have happened to Trip A, including additional charging. Without the "Since Last Charge" shot, that video just has no place in this particular conversation, and that's something I think we should be able to agree on.

Tesla's initial Model 3 packs used for testing might have been closer to 78 KWh. The chemistry used in those cells may have been found to have some long term problems, so they went into production with another chemistry that was more reliable, but a couple percent lower capacity. Another possibility is the lab made 2170's capacity may have turned out to be impossible to consistently achieve in the real world mass production, so they had to scale back the rated capacity a bit.

The fact is that Tesla's current internal documents say the Model 3 LR pack is 74 KWh. There is no reason at all to lie to themselves about it. If they wanted to test and select the best rated cells off the production line (more effort than it's worth), they probably could make a small number of 78 KWh packs, but it would be an expensive and pointless exercise.

Actually, some of the earliest the release candidate vehicles used 18650-based packs (firmware for these packs is still in the model 3 software, actually, since some of these cars are probably still on the road), and didn't even use the Model 3 packs as we know them today. The one EPA test @vgrinshpun posted showed the VIN of an early release candidate car (release candidate 12, based on the last 6 of the VIN), so I don't think that should be relied upon for anything anyway even if it did have an early 2170 pack. If this were an independent test of a random vehicle purchased from Tesla without their knowing it was for testing, I'd put more weight on it. (I'm starting to think @vgrinshpun works for the EPA or something... would at least explain why he's so insistent about their data despite it clearly not being real world comparable.)

The other issue that's super obvious is the ridiculously poor charger efficiency suggested by that data. The Model 3 charger is well over 90% efficient, really around 92% or so real world. It's the most efficient charging setup yet, but that data puts it at like 86% or something. Even ignoring the other data, this alone should be enough to throw some red flags.

I don't believe there are any Model 3 packs with 78 kWh total capacity. The way the full pack energy is reported on the 3, though (with the 4 kWh buffer added to the total but lost linearly along the discharge curve, unlike S/X that partitions this), there are probably some 75 kWh packs out there... maybe a tiny fraction of 76 kWh packs. But there just aren't any 78 kWh packs. It might be possible to get that lucky on a car or two every few million cars, but I just don't see it happening enough to matter or be relevant to this conversation.

I'll stick with the original proof request: A "Since Last Charge" screen showing 74+kWh discharged. Again, this should be super simple, doesn't require any test equipment except a Model 3, a driver, a camera, and some non-mountain highway. Like I said, seems pretty weird to me that if the pack were in fact 78 kWh that no one will have posted a photo of this screen.

I own an early LR 3... and I've tried it. I can't even get 70 to show up, let alone 74, in the best conditions. I get the range (in the spring) without a problem, surprisingly, and the car still charges to ~312 or so miles at 100% after ~21k miles.

 

[llavalle]

I'll stick with the original proof request: A "Since Last Charge" screen showing 74+kWh discharged. Again, this should be super simple, doesn't require any test equipment except a Model 3, a driver, a camera, and some non-mountain highway. Like I said, seems pretty weird to me that if the pack were in fact 78 kWh that no one will have posted a photo of this screen..

Should be pretty straightforward.... Did a few of those tests on my 85D when I got it back in 2015... Was trying to get over 80kWh (it was a 85kWh after all right ). Closest I could get (without being stranded) was 74-75kWh since last charge... and extrapolating for the %left inthe battery, my estimate was 78 total usable.. which turned out to be pretty accurate.

Even had a weird math problem when I did my first road trip... and my math was adding up to around 75-76kWh usable

First road trip with the car - weird range behavior - 85D

Again, Model S.. not Model 3.. but yeah, you get the point... these don't compare to the EPA documents at all either.