Yea why not put the next 10 gf right there
That's pretty much what I expect them to do. When you think about the challenge of doubling capacity every 18 months for about 15 years, what you need are a series of Gigafactory sites where you can expand at a constant rate for 15 to 20 years.
So basically they need enough land in Sparks that they can just continue to roll out 10 to 20 GWh of capacity every year until EVs gain about 50% share of the new car market. This is also preferable to doubling the number of Gigafactory sites every 18 months. Under that model, management spends far too much time negotiating new sites. So it is better to have one site where you can keep expanding for 15 years than 5 sites that you can only expand for about 3 years each. Moreover, this is efficient from a hiring point of view. Experienced employees are more productive. As you expand for 15 year in the same location, productivity continues to increase, labor cost per GWh keeps going down. By comparison, if you have 5 smaller sites that can only expand for 3 years, as productivity increases, you have to shrink head count after reaching full capacity. This is not as attractive for the best tallent, and job preservation works against productivity gains. Basically productivity and jobs are in competition. But when there is a really long expansion jobs and productivity can both increase and there is lots of advancement opportunity to motivate the best performers.
So how does Tesla accomplish exponential growth if a single site like Sparks can only support linear growth albeit for 15 to 20 years? Basically if Tesla launches one new Gigafactory site each year for 16 years, and if each site grows linearly 250GWh capacity over a 15 year ramp up, then that would ultimately lead to 4 TWh capacity with logistic growth by 2045. That, along with some density gains along the way, would be about enough capacity for the new car market. The halfway point is reached in 2030, about 2TWh capacity plus density gains, enough to power half of the new car market. This growth is not quit exponential leading up to 2030; it is just quadratic. But that is close enough to expontial and the longer trajectory to 2045 must be sigmoidal anyway. That is, as EV market share expands beyond 50% the rate of growth must slow down. Otherwise, there would likely be excess capacity leading to a short-term glut.
I worked out this trajectory in a model awhile back, but was reluctant to share it. My concern was that it may well tip off too much of Tesla's unpublished long-term plans. Additionally there was also the personal risk that it would be misunderstood and laughed off. But for myself, I was satisfied to see that there was a manageble path Tesla could follow to supply half the new car market by 2030. Moreover, cash flow from this operation is sufficient to be self-funding given Tesla's cash on hand and other capital commitments from Panasonic. It's one thing to discribe such a model; it's another thing to publish such a model.
So I am still reluctant to publish the model. The key insight behind the model, however, is that sites like Sparks need to be large enough that they can sustain linear growth for at least 15 years. The fact that Tesla has chosen a site that gives them an option on 10x the initial land is consistent with my hypothesis that sites must be large enough for 15 years of linear growth. Further confirmation of my theory and model would be the announcement of another Gigafactory location to begin development within 24 months of the Sparks site, and this second site must also have the option for sufficient land to develop over 15 years. Actually, they should announce this within 12 months, but I can envision special reasons why they may wait longer for the second site. After that, new site announcements should arrive about 12 months apart. Of course, there are other ways this could be scaled which could throw off this prediction. But basically, I'm thinking that they've got the scale about right: each site should roll out 50GWh of capacity every 3 years and repeat this for 15 or more years. So keep my predictions in mind. They may seem crazy right now, but if I understand correctly the magnitude of this transformation of the auto industry and Tesla is truly up for the challenge, then these predictions are just based on the simplest way to accomplish this transformation. If anyone can think of a smarter way to scale to 2TWh by 2030, please let me know.