Wow, I step away from this thread for a month, and tftf is still going on about $5B as if we are supposed to be impressed by the bigness of the number.
Thank you for this, I hadn't really considered the cost of CapEx amortized over the span of its assumed life. I mean it makes sense of course.
You know the other aspect of this, consider the most cut throat business rocket launching. There is an inconceivable amount of new cars produced every year (currently around 90M) and we are worried about Tesla getting boxed out of the market before they can hit .5% of the share? Consider instead Rocket Launches. There are only what? Like 100 some odd launches every year? And SpaceX did 6. Where is the outcry that SpaceX is *only* at 70M per launch. I mean, you better watch out because the competition is just going to come in there and wreck their world, right? Yet, somehow, somehow... the ancient dinosaurs of the industry Boeing and Lockheed with their launches exceeding 250M somehow still are able to get some amount of business. Why is that? Why would anyone willingly pay 3x the cost for what amounts to the same product? Because there is more market (demand) than there is people able to meet that demand (supply).
So here we have an EV company with a supply, but their supply amounts to a fraction of the potential demand. Why do people buy the BMW i8? Why do people buy the Porsche 918 Spider? Why do people buy any number of other inferior Plug-in type vehicles? Because there is more demand than supply (or people have some amount of brand loyalty and/or other draw toward that product causing a perceived added value... but mostly its a supply issue). So even if this magical competition comes out of nowhere and by 2020 there is enough supply for 12 equal companies to each produce 500k cars and all of them at a better price than Tesla, worst case, Tesla becomes like Boeing and Lockheed with their rocket launches still finding business at 3x cost because now instead of the EV market being at .5% of the total demand, the market is at 6.5% of the total annual demand. Tesla is still going to find business. It won't be until we start to see *maybe* 50% market supply of the underlying demand that we really start to look at having trouble which means some 45M new EVs produced each year.
It had been mentioned before that in order for Tesla to have an issue LG (or some other company - companies) would need to not just be building a gigafactory at 35GWh annual, but multiple factories at 350GWh annual (supply for approx 5M cars). I would counter that is still being generous and the number is likely closer to 3.15TWh before Tesla would have an issue finding market for their cars and margins would become "razor thin" and would have to cut costs further than they have.
This is all of course assuming that they also have no market for stationary storage. That means a drop of 50% of the worlds power plants could be converted to storage. According to EIA.gov there are 19,243 generators with a nameplate capacity of at least 1MW. In 2012 global installed capacity was 5,550,000,000 kWs. Yet the total annual generation was 21,532,000,000,000 kWhs. So lets look at the minimum average hourly generation to come to a true baseload demand if the curve was flat. That amounts to 2,457,990,868 kW per hour. So looking at the difference between the two is where they come to a reduction of 50% of the worlds power plants.
The question then becomes how much storage would one need? Well lets ignore the irregularities of renewables and assume the are all consistent generation sources such as hydro, coal, natural gas, or nuclear. So we need to support the potential for double the demand at certain time periods (5.5TW). Let's then also assume that storage only needs to match hourly at a 1-1 ratio of this peak power draw, so a removal of half the worlds power production would mean we need at a *minimum* 2.5TWh of storage. If the cost benefit of undercutting *all* forms of peak plants is at 850$/kWh it won't matter if Tesla is undercut on their current pack price of 250$/kWh advertised. Why? because the total addressable market is vastly larger than anyone can currently produce for. It would take 71 years of full gigafactory production to meet that demand (that's 2,500 GWh demand / 35GWh production per year). So even if, again, 2020 rolls around, and we have 350GWh of annual production combined by all this non-existent competition, it would *still* take 7 years of full production to meet this minimum supply requirement.
Realistically though, the 2.5TWh supply need is *vastly* under scoped, because you need account for multiple hours of over draw, and then multiple hours of under draw to replenish the store. Based on a 2012 curve (to try to remove renewables from the equation) this amounts to around a 15% +/- on either side, so lets assume for half the day you are running 15% too hot, and half the day running too low. You would then need to produce enough storage to hold out for at least 12 hours of running too hot for the day, before getting to the other side of the curve. Back to global numbers, the daily average is 58,991,780,821 kWh, if we are assuming that we only have 2,457,990,868 kW installed capacity, and dealing with just a 15% over/under, this amounts to 33,920,273,978 during the peak draw. Subtracting out the baseload generation of 29,495,890,416, amounts to 4,424,383,562 kWh of required storage to ensure enough draw during the 12 peak hours. This has now nearly doubled the required storage needs to 4.4TWh. And this is only with a *very* gentle 15% curve. As we go more into the extreme with Solar using 2020 projections, the over/under rises to around 36%, which increases the storage needs to 10,618,520,544 kWh or 10.6TWh.
The point is, we are not trying to compete battery vs battery... that doesn't matter. Otherwise how in the world did LG land a contract costing them around 1,000$/kWh with an order of 1GWh if Tesla is already able to take orders for 250$/kWh? By TFTF's logic, LG should be going out of business today, and not be landing more contracts that cost 4x the "going rate". So what are we *actually* competing against? The cost of other power generations. Coal, Natural Gas, Nuclear... etc. And I have heard numbers ranging from 500$/kWh to 850$/kWh before you have basically undercut all those plants completely. The variance is likely trying to take into account that difference between the need for say, 10TWh of global storage needs vs something lower like 2 or 4TWh. What is clear, the market is vast, and the supply... the supply is *very* low... and will remain low for the foreseeable future.
Again, even with the assumption that all the "competition" comes out of nowhere to build an vast amount of factories about to produce 350GWh a year, it won't matter what price anyone sets, as long as they are lower than 500$/kWh, Tesla will easily sell anything they can make, because the demand will be there to support the higher price. If someone comes and sells then at 1$/kWh, it won't matter, because they just won't have the supply.
