petit_bateau
Active Member
Here's my simple model for Tesla's Battery Master Plan v.3
View attachment 591857
- The pilot plant is a proof-of-function prototype with a single line @ 10GWh/yr
- Tesla needs to build 9 additional Bty lines per year, each with 10GWh/yr capacity
- This gets us to 100GWh capacity in 2022, and ~3,000 by about 2029
- Note: if 3rd parties are unwilling/unable Tesla can build out 20 TWh/yr in 12 yrs
Now we get to the unstated consequence of Battery Day: Tesla MUST build a tertiary layer of production to build the machines that equip the Bty Cell factories.
Tesla MUST add 9 new such Bty lines (per chart above) to achieve the arithmetic growth required by their stated goals.
It might be practical to build a single such Meta-Factory, but given the scale required (potentially ~180 lines over 12 years), I think its more practical to build THREE such Meta-Factories (one each in China, Germany, and USA).
Each of these MTBTMs (Machines-that-build-the-Machine) would each have to be sized to produce 60 bty lines over 12 years, or about 5 each per year.
Think of these MTBTMs as Grohmann clones, but specialized for battery cells.
Does anybody know where (and from whom) Coca-cola Bottlers buy there production equipment? That's the scale of manufacturing we are discussing right now.
Cheers!
1. You are assuming a linear build-out of capacity. In my modelling I assume non-linear. The problem with my sort of modelling if you do it too crudely is that in 2030 you get 20mln cars/yr capacity and then in 2031 you get 30 mln cars/yr of capacity.Therefore I anticipate that at some point on the journey they will be observing a) competitor reactions, and b) FSD-impact on vehicle demand and c) storage penetration, and d) competition/regulatory concerns, and as a result may ease back from non-linear growth to linear or sub-linear growth so as to come off the S-curve in a bumpless transfer - at least that will be their aim. I think they will be making that transfer at about 2030 as the trajectory they discussed in battery day is much the same as the one I have been modelling.
2. This exactly why they were buying Grohman and the other acquisitions of manufacturing eqpt. Some of this new stuff is a different skillset (it is more akin to mining/paper industry processes, e.g. the Voiths, Metso, Valmet) and so some further acquisitions may be needed to bring the relevant core competencies in-house. Whether they do that through business acquisitions, or through personnel acquisitions I am unsure, most likely a mixture. They might as they have done in the past partner with a suite of companies then cherry pick as they hollow out their supply chain.
3. Vertical integration is a double-edged sword. Most obviously it creates downsides in bigger capital demands, and issues of controlling scale and scope to set against the evident upsides. However it also creates a problem that the supply chain that you wish to partner with today, is well aware that you will likely cannibalise them tomorrow. That in turn breeds a certain leeriness, and indeed I know companies that have turned down work with Tesla because they saw this exact risk several years ago (it is not just Tesla that looks forwards strategically) and chose not to put themselves at risk.
regards, dspp/pb