-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Gigafactory locations and products
- Thread starter Buckminster
- Start date
I changed my mind about lower volume car factories in locations like Australia, I now don't think it is possible...
Car factory requirements:-
- Energy
- Land
- Raw Materials
- Parts Supply Chain
- Engineering Talent
- Factory workforce
- High volume markets nearby..
Short term options:-
- Expand Austin
- Expand Berlin
- Expand Shanghai, or new China factory
- Redevelop Fremont
Redevelop Fremont is a game of Tetris, and that is why I can see Austin producing Model 3s before the major phase of Fremont redevelopment..
But a redeveloped Fremont ticks all of the boxes 1.-7. above.. It will have more floor levels and a optimised layout similar to Austin and Berlin, so it will be worth the hassle, time and money...
Countries / locations that iMO also are a good fit with 1.-7. above:-
- Mexico
- Brazil
- Japan
- Korea
- Italy
- France
- India
- UK
It is far easier to expand existing factories or expand in the US than open factories in new countries.
There are already other US car makers making cars in Mexico, so it may be one of the easier options..
Key assumption casting determines the peak rate of production...
A casting takes 2 minutes x 20 hours = 600 per day, by 340 days = 204.000 per year...
I'm deliberately rounding down here.. so the numbers are very conservative..
For Model 3/Y 2 6000 machines are required.
Cybertruck is 8000 series machine for the rear, I'll assume another 600 series machine for the (likely) front casting
We know Berlin has space for 8x6000 machines so peak production is 800,000 per year (phase 1) Model 3/Y
For Austin I'll assume 6 x 6000 series Model Y, and 8000 series + 6,000 series Cybertruck... (2023)
I'll also assume the production ramp tracks as follows, year 1,=30%, year 2=60% year 3=90%
So for 2022 Austin Model Y = 180,000, Berlin Model Y = 240,000 - Total = 420,000
With some likely expansion of Model Y production in Shanghai, Tesla should be close to 1.5M
Fremont will probably stop making Model Y in 2022, but I think the tents, stamping, and paintshop capacity and battery packs can be used to make more Model 3s up to the total capacity of the Model 3 body shop which is the limiting factor...
I also assume moderate amount of Cybertruck and possibly Semi production in 2022 also helps cancel out any drop of production in Fremont..
The point of these estimates isN'T a rigorous definition of exactly what will happen, more a rough idea on rate of production expansion.
When new factories will be required, where they may be located and what they may build...
Additional for 2023 (30% increased to 60%) Austin Model Y = 180,000, Berlin Model Y = 240,000 -Austin Cybertruck (30%) = 60,000 - Total = 480.000
We are short 250,000 units, Austin Semi may provide some, Fremont Model S/X, but the majority would need to come from Shanghai. probably more Model Y.
Already we can see that Tesla will need to plan for further expansion and start construction projects..
I think we will get 2 compact models as previously indicated:-
- Model 2C - China Compact - Shanghai
- Model 2E - Europe Compact - Berlin
- Shanghai - Model 2C (3) - Construction 2022
- Berlin Model 3 (2) - Construction 2022-2023
- Austin - Model 3 (2) - Construction 2022-2023
- Berlin Model 2E (3) - Construction 2023-2024
- Austin Model 2C and Model 2E (3) - Construction 2023-2024
If Shanghai Model 2C construction can be completed in 2022 we might get some Model 2C cars produced in 2023.
But I am mainly going to count Berlin/Austin Model 3 and Shanghai Model 2C in 2024 numbers
So far, I don't think we need a new factory site, but Phase 3 may be the last phase for Berlin and Shanghai,...
In both cases we would need to be sure the factories have enough land for phase 3.,,if not, a new factory location, in the same country is needed.
For Austin I have zero concerns about the amount of land available..
Last edited:
Watts_Up
Well-Known Member
Do you think that the Model 3 will be updated and will get a Casting chassis like the Model Y?
- If so, the Model 3 could then share the same assembly lane than the Model Y, both in Berlin and Austin.
- If not, I doubt that a Model 3 specific assembly lane will be created in Berlin and Austin. The reason been that the Casting makes the assembly line smaller.
Honestly, I doubt that the Model 3 will get a Casting, so the Moel 3 will continue to be build only in Fremont and Shanghai.
There is also only a remote possibility to get the Model 2 to be imported, or to be build in US, but really I doubt about it.
Additionaly, I am puzzle about the Model 2C (China) and Model 2E (Europe): Would they be different models?
I think I fact that there will be a Model 2 "Hatchback" and a Model Z "SUV".
Last edited:
Yes, that is what I think they will do at Berlin and Austin...
I'm just trying to apply logic to work out what they might do...
Mostly I'm trying to see how much of the expansion can happen within existing factory locations, the team to build Austin had to come from Fremont..
Factory expansion or new factories in Germany, China and the US, seems easier than starting in new countries..
I also think at some stage they need to shut Model 3 production in Fremont to do a major development, but can keep Model S/X and the Roadster production running.. That is the interesting part of the jigsaw, because they want to expand production, but Fremont needs a major development to become like the newer factories...
I'm factoring in redevelopment of Fremont, but it might not happen. Regardless Model 3 and Fremont wil also eventually move to castings, at a minimum that is a rebuild of the body shop.
Elon mentioned the ChInese team and German team both designing cars..
I think the Chinese design could be more of a Sedan (like Mode 3) and the German design might be more of a hatchback (like Model Y).
The German car may be slightly bigger and taller...
My speculation is based on Tesla making 20 Million cars by 2030, we need to break that up into numbers than make sense (round Millions)
- Model Y 3 Million
- Model 3 1.5 Million
- Cybertruck 0.5 million
- Model 2C - 5 Million
- Model 2E - 5 Million
Also Model 3. Model 2C and Model 2E are ideal Robotaxis...
There could also be other combinations of models that work...
My hunch is anything bigger than Model 3/Y would cost more and sell in lower numbers...
- Say Model 6 (Bigger Model 3) - 1 Million
- Say Model 7 (Bigger Model Y) - 1.5 Million
My other ideas are:-
- Smaller Cybertruck - 0.5 Million
- Cyber-van (same platform as smaller Cybertruck) - 1 Million ( vans are popular worldwide)
Another factor is between now and 2030, there isn't time to design a vast array of models.
Last edited:
Watts_Up
Well-Known Member
In Fremont, in a near by location, Tesla is making all the seats, and also a lot of smaller plastic molded pieces in Lathrop, CA, and batteries manufacturing in Nevada.
I suppose the same thing occurs in Shanghai?
Will also Austin and Berlin get a seat factory, and other stamping and miscellenous in house manufacturing parts, directly inside the new factories?
I suppose the same thing occurs in Shanghai?
Will also Austin and Berlin get a seat factory, and other stamping and miscellenous in house manufacturing parts, directly inside the new factories?
Yes, I expect that...
We now need to factor in the Robot, Fremont looks like an ideal location..
This is what I have so far attempting to model production in a spreadsheet:-
I've had to invent a Mexico factory to give any chance of closing Fremont for redevelopment at some future date.... (Not saying it is Mexico, I picked a random location that gives the best chance of redeveloping Fremont.)
The target of 50% compound annual growth is hard to chase down...
For Model 2E/2C there is some change of using smaller 4000 series casting machines for front and rear castings, these machines are smaller and cheaper but may have a higher run rate up to 1,000 castings per day per machine..
So I assumed in the primary locations for compact cars the factory would aim to produce 1 Million (1000K) cars per year...
If I don't make these kind of aggressive assumptions. the targeted 50% growth per year is even harder to chase down..
2021 | 2022 | 2023 | 2024 | |
| Shanghai 3+Y | 500,000 | 600,000 | 600,000 | 650,000 |
| Fremont 3+Y (500K) | 500,000 | 500,000 | 500,000 | 500,000 |
| Berlin Y (800K) | 240,000 | 480,000 | ||
| Berlin 3 (400K) | 120,000 | |||
| Berlin Model 2E (1000K) | 300,000 | |||
| Shanghai Model 2C (1000K) | 300,000 | 600,000 | ||
| Austin Model Y (600K) | 180,000 | 360,000 | ||
| Austin CyberTruck (200K) | 60,000 | 120,000 | ||
| Austin Model 2E | ||||
| Austin Model 2C | ||||
| Austin Model 3 (400K) | 120,000 | |||
| Fremont Model S+X | 40,000 | 100,000 | 100,000 | 100,000 |
| Fremont Robot | ||||
| Mexico Model 3 (800K) | 240,000 | |||
| Mexico Model 2C | ||||
| Mexico Model 2E | ||||
| Total | 1,040,000 | 1,200,000 | 1,980,000 | 3,590,000 |
| Target | 800,000 | 1,500,000 | 2,250,000 | 3,800,000 |
I've had to invent a Mexico factory to give any chance of closing Fremont for redevelopment at some future date.... (Not saying it is Mexico, I picked a random location that gives the best chance of redeveloping Fremont.)
The target of 50% compound annual growth is hard to chase down...
For Model 2E/2C there is some change of using smaller 4000 series casting machines for front and rear castings, these machines are smaller and cheaper but may have a higher run rate up to 1,000 castings per day per machine..
So I assumed in the primary locations for compact cars the factory would aim to produce 1 Million (1000K) cars per year...
If I don't make these kind of aggressive assumptions. the targeted 50% growth per year is even harder to chase down..
Bottom line....
My hypothetical number crunching exercise concluded 2 new factory sites in 2023 (Mexico, China) and 1 new factory site in 2024 (Europe).
I was able to redevelop Fremont - 2025-2027.
But obviously the assumptions are hard to pin down / verify.
What is interesting is that 50% compound growth is a very hard target.
My hypothetical number crunching exercise concluded 2 new factory sites in 2023 (Mexico, China) and 1 new factory site in 2024 (Europe).
I was able to redevelop Fremont - 2025-2027.
But obviously the assumptions are hard to pin down / verify.
What is interesting is that 50% compound growth is a very hard target.
petit_bateau
Active Member
I think the 50% + growth over the next few years is possible, indeed necessary if Tesla are to reach the 20m/yr target by 2030. When I try to solve for the various constraints it seems to me that Tesla will need to continue at >50% growth until 2024. Thereafter the rate of growth can taper off though the annual quantity added continues to increase until the annual add reaches approx 3m/yr in 2028. Obviously cell supply needs to keep pace with this. In addition to that one needs to continue build-out of the existing factories, as well as probably breaking ground on two new factories each year - my view is that this will be necessary as only so much can be achieved on any given set of workfaces in one location for a variety of reasons. Alongside this I think that in the period 2022-2025 they will need to introduce new platforms at a rate of min 1 / max 2 per year, plus of course managing the release of any model-derivatives that are based off the platforms. Only if they can achieve all this do they hit the stated target, but that is of course what we need to achieve or else humanity is in for a very rough ride.
Attachments
jeewee3000
Active Member
It has been mentioned that up until this point Tesla did not have the luxury of converting its Fremont Model 3 line to a newer version Model 3 (including front and rear castings and structural battery pack) since the Fremont Model 3 is/was responsible for the majority of Tesla's quarterly revenue.
Shutting down Fremont Model 3 would be a cash disaster. But after the China Model 3 ramp and Berlin and Austin spitting out Model Y's (therefore creating revenue/FCF), Tesla has plans to 'update' Fremont 3/Y lines to incorporate the newer, more efficient assembly methods.
I'm sorry I can't provide a source. This is of the top of my head but I'm certain I've heard this somewhere credible. (post Battery Day)
IIRC the main bottleneck was Roadrunner cells. Only when there are enough of them, the Model 3 in Fremont would switch over.
So it will happen, but it will take some time. (2023?)
I'm glad someone else attempted this exercise, I did recognise the need to flip my assumptions and try with more new factory locations earlier.
The main reason I tried to slow down the pace of addition of new factories is a skilled team is required to design, build, set up and commission new factories..
For example a team of people came Fremont to Austin to help get Austin built...
That is why I favor Mexico and a redevelopment of Fremont over Brazil..
Some of the team from Austin can move to Mexico to help do Mexico, then move back to Fremont to redevelop Fremont...
For Europe and China it is easy to see where the team to build the factory is going to come from...
It is the same with car design, we know the Chinese team is designing one car, the Germany team is designing one, and I can see the German team designing at least one more model.
I can see all Cyber designs (how ever many there are) being designed and initially built in Austin, and I see Fremont continuing to be a base for the development of new models... India might be a base for an eventual design of a ultra compact model.
So all new designs come from this 4-5 design centres and I would assume a new design takes 2 years.
Having done the exercise you know how complex it is...
My other thoughts on Mexico where a lot of workers at Fremont and Austin may have family links to Mexico and would be happy with a stint working in Mexico.
And also if Fremont is redeveloped at any stage, a factory in Mexico is a handy source of cars for the West Coast..
On the bot, I have had 2 ideas related to factory builds:-
On factory workers becoming engineers I can remember friends form school who were highly intelligent but also totally uninterested in doing school work. Their main interest was seeing how many rules they could break and not get caught...and they were very good at that..
What I am saying here is for a variety of reasons many smart people don't have formal education qualifications, some smart people could be doing factory work to earn a living, merely because that is how fate has played out. Dangle the carrot and see who wants to take it..
And also if Fremont is redeveloped at any stage, a factory in Mexico is a handy source of cars for the West Coast..
On the bot, I have had 2 ideas related to factory builds:-
- The bot starts on entry level jobs, and an attempt is made to move the factory workforce up one notch, workers are trained for higher level jobs...
- The next step is cadetships attempting to train the best and brightest factory workers as engineers.
On factory workers becoming engineers I can remember friends form school who were highly intelligent but also totally uninterested in doing school work. Their main interest was seeing how many rules they could break and not get caught...and they were very good at that..
What I am saying here is for a variety of reasons many smart people don't have formal education qualifications, some smart people could be doing factory work to earn a living, merely because that is how fate has played out. Dangle the carrot and see who wants to take it..
GigaTexas
Member
Please correct me if I'm wrong, won't GigaTexas also have a park open to the public?
It also has some amazing Crossfit locations within a short drive!
