Model Y - Gigafactory Texas Production

[ZenRockGarden]

It is possible the current rendition of the structural battery pack has a different chemistry. If, for instance, they went iron, that would explain a lesser range with the same number of cells. It would also be more readily available as materials go, and cheaper to make. It would also negate much of the weight gain from that structural pack.

Just a thought.

Given the visual counts in the low 800's for the cells in the 4680 pack seen yesterday vs the 960 cell original design drawings, I'm thinking it's the simplest explanation we came up with for the SR+ EPA cert.

4680 cells are pretty much as intended, with NMC chemistry give or take a little colbalt.

Current pack configuration of 828 cells at 70 KWH giving a range of 270 miles and allowing them to make ~15% more cars from their currently limited count of cells.

They ramp up on that setup, maybe sending a lot of them to their increasingly large fleet commitments.

Later, when 4680 production and the two plants are ready, they roll a 960 cell LR version of the pack into all MYLR's, ending up with a 270 mile entry point, and a 340-350 mile top model.

and yes, I am making all this up - but it does fit the available data and might explain Teslas rather weird silence on the topic since it means although "customers" are getting 4680 packs, it's not really the general public just yet.

 

[iamnesty]

Might get more info during the earnings call on 4/20. I’m sure there will be a question regarding Austin and ramp up. Regardless, I’ll be happy to get my MYP from any factory seeing as they have improved it with a new 12v battery and nvidia chip.

 

[terranx]

Later, when 4680 production and the two plants are ready, they roll a 960 cell LR version of the pack into all MYLR's, ending up with a 270 mile entry point, and a 340-350 top model.

Not sure your math there quite works out.

If we assume the 828 cell pack is 70 kWh, an 82 kWh battery (i.e matching the current 2170 battery) would require 969 cells. Probably can call 960 margin of error at that point. So I don't see how they'll get to 340-350, especially since based on those EPA docs, the MYSR is not really any lighter than the LR/P.

 

[ZenRockGarden]

Not sure your math there quite works out.

If we assume the 828 cell pack is 70 kWh, an 82 kWh battery (i.e matching the current 2170 battery) would require 969 cells. Probably can call 960 margin of error at that point. So I don't see how they'll get to 340-350, especially since based on those EPA docs, the MYSR is not really any lighter than the LR/P.

I'm rounding and assuming a lot there - but the rough numbers make reasonable sense.

 

[terranx]

I'm rounding and assuming a lot there - but the rough numbers make reasonable sense.

Given the numbers, I think most likely scenario is that when they do make the "full" 4680 battery pack, it'll have a very similar capacity to the current model. Since they apparently plan to supply cars from both Fremont and Texas to different parts of the country, it makes sense that they'll have similar specs overall.

 

[ZenRockGarden]

Given the numbers, I think most likely scenario is that when they do make the "full" 4680 battery pack, it'll have a very similar capacity to the current model. Since they apparently plan to supply cars from both Fremont and Texas to different parts of the country, it makes sense that they'll have similar specs overall.

Manufacturing efficiency is a huge goal for them, so if you have one "structural pack" design, and just two variants with different cell counts inside, you can keep almost everything else in the car and assembly line identical, which has a lot of charm. I'm also guessing they've not yet squeezed everything they can out of the new 4680 cell hence my guess at future MYLR with a small range bump.

 

[Yperformance]

How about the dry cathode issues that baglione mentioned. Reduces the cost saving by about 10% Elon said 15%.

Maybe Elon should do AMA to answer all the rumors around Giga Texas as he’s doing one on twitter to clear some air raised by twitter employees after he bought 9% twitter shares.

Totally agree!!!!!!!

 

[Ghostchili]

After reading too many posts in this forum, I've formed my own speculation.
1)Texas will build an entry level Y (Including supplying rental company orders), Semi, and Cyber Truck production.
2)MYP will remain 2170 batteries for at least another year for several possible reasons like 4860 supply and cooling batteries with a higher current motor . The S and X are still 18650 batteries after all.
3)If and when the 4680's go in the MYP, it will be the same time at both US factories.
4)The 4680 and Structural pack/frame are cheaper but may not be better.
5)April 7th will leave more questions than answers.

Take the above with a grain of salt.

I’m still pretty close

 

[Yperformance]

Remind me again how many million 4680 batteries do we need just to complete MYP orders per year , let say 950 per battery pack.

 

[dhanson865]

The whole point of an EPA certification is that you MUST have produced the thing they certify.

Every once in a while someone will mention how EPA tests cars and I'll post this:

EPA doesn't test the car in any way, it's all self reported numbers provided by Tesla.

• EPA numbers are self reported and only a small number of car models each year are chosen for random testing. In 2017 that percentage was near 9.58% (0.0958 if I got the math right).
  
  
• For example in 2017 the EPA tested 0 Tesla cars. Same for 2018 as well. Same for 2021 (since I just checked while editing this post).

See Data on Cars used for Testing Fuel Economy | US EPA for the raw data.

Go ahead and download the list for 2021 and 2022 and tell us which cars have been tested by the EPA (I'll give you a clue I've looked at the list for 2021 and all the Tesla models were self reported). That's marked in the "Test Originator" column.

 

[ZenRockGarden]

Every once in a while someone will mention how EPA tests cars and I'll post this:

EPA doesn't test the car in any way, it's all self reported numbers provided by Tesla.

• EPA numbers are self reported and only a small number of car models each year are chosen for random testing. In 2017 that percentage was near 9.58% (0.0958 if I got the math right).
  
  
• For example in 2017 the EPA tested 0 Tesla cars. Same for 2018 as well. Same for 2021 (since I just checked while editing this post).

See Data on Cars used for Testing Fuel Economy | US EPA for the raw data.

Go ahead and download the list for 2021 and 2022 and tell us which cars have been tested by the EPA (I'll give you a clue I've looked at the list for 2021 and all the Tesla models were self reported). That's marked in the "Test Originator" column.

Do we have any examples of the EPA certifying a completely hypothetical non-existent car based on manufacturer self report?

 

[advocate8]

Maybe Elon should do AMA to answer all the rumors around Giga Texas as he’s doing one on twitter to clear some air raised by twitter employees after he bought 9% twitter shares.

But....Tweets are more important than cars. Soooo vital. Should we go to 142 characters?
Should we let people spread deadly lies about Covid and incite riots?
Vote yes or no

 

[advocate8]

I'm rounding and assuming a lot there - but the rough numbers make reasonable sense.

The scenario makes sense, and they will get a little more out of each cell eventually. But what’s hurting here is the idea that it seems inevitable a 4680 MYLR is going to be heavier than current one. quick calculations say 960 cells will yield about 105 lbs more than the 279-mile range version.

 

[ZenRockGarden]

The scenario makes sense, and they will get a little more out of each cell eventually. But what’s hurting here is the idea that it seems inevitable a 4680 MYLR is going to be heavier than current one. quick calculations say 960 cells will yield about 105 lbs more than the 279-mile range version.

I think the MYAWD came in like 70 or 80 pounds under the MYLR, so my feeling is a future 4680 MYLR pack is roughly equal to today's if all our assumptions hold true. I know everyone was looking for 400 pounds savings, that just wasn't gonna happen, although I could see Tesla refining the structural pack over time to shave more off it. You always gotta start conservative till you get experience. I am reminded of the space shuttle that had a white external tank till they found out the paint wasn't needed and could be skipped to save a surprising amount of launch mass

 

[dhanson865]

The whole point of an EPA certification is that you MUST have produced the thing they certify.

Do we have any examples of the EPA certifying a completely hypothetical non-existent car based on manufacturer self report?

Your claim the first time was that EPA certifies cars that have been produced. Produced doesn't equal sold. You can make a car (prototypes, alpha, beta, release candidate, whatever you want to call it), document it, get ready to sell it and not have started selling it yet. Then it could later be released or canceled.

I'm not going to dig through the EPA records for examples of that happening, but you are welcome to if you don't understand the concept.

The point is that the EPA doesn't come in and test cars, cars are tested and the data is sent to the EPA. And just because the data is sent doesn't mean the car is currently for sale.

 

[ZenRockGarden]

In order to be tested, the cars have to _exist_. I was countering the original assertion that an EPA cert didn't mean anything had actually been produced.

 

[dhanson865]

In order to be tested, the cars have to _exist_. I was countering the original assertion that an EPA cert didn't mean anything had actually been produced.

They can literally make 1 car (not cars plural) test it, send the data in with the intention of making more later. That 1 car could be hand built, outside of the normal assembly process, does not have to be built using the assembly line, and so on.

So using the EPA to prove production has or hasn't happened isn't productive unless you want to be very very specific on what you think has happened, and waste time on disclaimers about the various possibilities.

You replied to someone that said EPA isn't proof of production. Literally that is wrong, there is an implied word they should have included to be technically correct that that is "mass" as in "mass production". You just don't know how many someone made by looking at the EPA data.

You were correct if all we are doing is being pedantic. I'm just trying to clarify what the EPA is and isn't telling us about Tesla cars.

If you want data on quantity, crash test data helps because it has to have VINs and random selection from a pool of normally produced cars. But that's only like tens of cars, not hundreds, not thousands. You could get a new model EPA listed and crash tested with a pretty small pool of cars produced compared to the size of the Austin Gigafactory.

edit: To be certified for sale, every new model sold in the U.S. must be crash-tested internally to ensure minimum federal safety standards are met. But a publicly available rating isn’t required.

Some Cars Will Never Be Crash-Tested, Leaving Consumers in the Dark

Consumer Reports explains why some cars will never be crash-tested. Big SUVs, luxury cars, and specialty vehicles often lack publicly available crash-test scores, leaving consumers in the dark without this information from NHTSA or IIHS.

www.consumerreports.org

 

[MP3Mike]

EPA doesn't test the car in any way, it's all self reported numbers provided by Tesla.

• EPA numbers are self reported and only a small number of car models each year are chosen for random testing. In 2017 that percentage was near 9.58% (0.0958 if I got the math right).

You say that the "EPA doesn't test the car in any way", and then you immediately contradict yourself and say they do sometimes... Also, Tesla had to ask the EPA to retest the Model S, without leaving the keys in it overnight, to get the 400 mile range estimate:

EPA denies claims over Tesla Model S range test, Elon Musk responds

According to Musk, the keys were left in the car and the door open that sapped some of the range, but the EPA says that's not the case.

www.cnet.com

So while a lot of times the data is self reported, in some cases the EPA does the testing themselves.

 

[Andrew Smith]

If MIT MY is SR+ then where are those cars going? In the drone footage, they were being transported somewhere. There’s no option to buy them online.

Since Elon mentioned about robotaxi (he’s been saying for few years now) that makes me think that SR+ might be for robotaxi… and they might even use these for boring company too. There was a news that at one of the boring tunnel city, it’s fully autonomous.

Tesla might hire few folks, give them these SR+ and start working on city by city autonomous robotaxi. Similar to Waymo I guess that’s fully autonomous in SF
This will also ensure if there’s any issues with 4680 they’ll fix it and won’t leak out in the public.

 

[Curious Nomad]

How do we know that the new MIT Model Y has 828 cells? I am seeing reference to this in many posts but dont recall that being stated by Tesla. Was someone able to actually count all 828 cells? Or is this just a guess? Seriously, not trying to be difficult but wondering where this number came from?