I broke down in another article why I have a hunch that the Tesla Cybertruck is going to include a battery pack that is costing Tesla around $108/kWh at the pack level (or better!), but I decided to keep running the math to calculate what it would mean for other Tesla products. To be clear, there has been no indication from the company that there has been any breakthroughs, nor that Tesla would use this technology in these other products. This is my conjecture on what may be afoot, especially if these prices are anywhere close to correct.
Editor’s update, digging through the archives, I found this article from June 2018: $100/kWh Tesla Battery Cells This Year, $100/kWh Tesla Battery Packs In 2020. In that light, $100/kWh by end of 2021 sounds very reasonable. Also, I don’t think Tesla hit $100/kWh cells in 2018, so I would expect the 2020 forecast to be delayed/pushed back to 2021.
I am going to be begging Tesla and CleanTechnica to send me to the Battery Investor Day so I can try to find out more. If my analysis is even close to right, for all of the Cybertruck’s divisive design, the reveal event could be the Trojan Horse that signaled the end of huge portions of the fossil fuel era.
We don’t know the size of the battery on the Tesla Semi, but we do know it goes a mile using less than two kWh of battery. Therefore, if we assume the 300 mile range Semi has a battery pack of 600 kWh, the cost for that battery pack would be $64,956. While this is still obviously a big cost, with the base price of $150,000, it leaves over $85,000 for cab design and profits.
The 500 mile range Semi would have a battery pack around 1000 kWh, at a cost of $108,260, leaving the $200,000 model to have nearly $90,000 for cab design and profits.
There is no confirmation that this technology will be in the Tesla Semi, but I’ve felt like the delay in deployment had to do with Tesla swapping battery tech. If I’m correct, this would be a really good reason for the delay.
The Tesla Powerwall is currently a 13.5 kWh system and costs $6,500. Using the $190 per kWh at the pack level cost that has been mentioned in the past, now imagine the Powerwall costing the same amount, but containing a 24 kWh battery in it instead.
Where I live, if I was on time-of-day-based energy charges, I would save 10 cents per kWh. A current Powerwall pulling extra power in at night and using it during the day would take me 13.7 years to pay off. A 24 kWh battery at the same price would take me 7.42 years. (Note: this doesn’t include the installation costs.)
In other areas — like California, where the peak rates are significantly larger — it’s even faster. Let’s look at SCE, since it publishes its rates. The smallest difference in any of SCE’s plans is the 19 cent per kWh difference in its TOU-D-4-9PM program. This only applies 5 days a week, however. So, how does the payoff time here look?
Even not assuming the benefits from the weekend, this would give the owner a payoff period under 6 years.
To be fair to this argument, it probably would not fill and empty every day for maximum savings. However, the value proposition of a Powerwall is also that it backs up your house during power outages, meaning that even if the payback period is 10 years, it becomes more compelling.
Megapack & Energy Storage
I saved what I’m the most excited about until the end. We don’t know the cost of the Megapack, but we do know it contains 3000 kWh of battery backup. Does the Megapack include these new batteries? I don’t know if it does yet, but I strongly believe that it will, and here is where the game really changes.
Let’s take a look at the Hornsdale Power Reserve. The Tesla battery cost about $50 million (or maybe $91 million, I’ve seen it cited as both) to construct and has 129 megawatt-hours of energy storage (or 129,000 kWh). With $190 per kWh packs, the cost for just the packs for Hornsdale would have been $24,510,000. At $108.26, we can get 226,399 kWh (or 226.4 megawatt-hours) for the same expense.
What would this mean? This article from Greentech Media indicates that California decided to not build a 262 megawatt gas peaker plant, instead opting for a 195 MW portfolio of batteries. The EIA estimates [pdf] that a new natural gas power plant’s overnight cost (or the cost to build it if you could build the plant without incurring any interest costs) is $999/kW to construct. That would mean California’s 262 MW gas peaker plant would have cost a minimum of $261,738,000 to construct, never mind how much more it would cost to operate.
Since we don’t have the capacity sizing of the portfolio (the article noted that part of it would be 400 MWh), it’s hard to make a straight-up comparison as to the pricing Tesla could get here, but if we assume that Tesla can start installing 226 MWh of capacity for every $100 million spent, this builds in an additional $9 million in profit above the highest price I’ve seen Hornsdale quoted. Therefore, Tesla could install about 590 MWh of storage for the same price as that peaker plant.
But here’s the thing, battery backup doesn’t just replace peakers. It stabilizes the grid at all times. It increases reliability, while smoothing out peaks and lulls in demand and production. Batteries boot instantly. And, Tesla states:
“Using Megapack, Tesla can deploy an emissions-free 250 MW, 1 GWh power plant in less than three months on a three-acre footprint — four times faster than a traditional fossil fuel power plant of that size. Megapack can also be DC-connected directly to solar, creating seamless renewable energy plants.”
And there are additional savings. The Hornsdale battery was estimated to save about $40 million for the Australian market in the first year of operation. At that rate, it will pay for itself within three years. Hornsdale has certain market factors that are working in its favor, but if we saw half the savings from the hypothetical 590 MWh storage plant in my example above (noting that it would be 4.57 times larger than Hornsdale), this new battery plant would pay for itself within 3 years, earning $91.4 million a year. If it only makes a quarter of the profit of the Hornsdale battery, it still manages to pay for itself within 6 years.
If this is true, NextEra’s CEO, Jim Robo, would have been exactly accurate when he stated last year that “post 2020, there may never be another peaker built in the US.”
It also makes this statement from Elon Musk on the Q3 2019 Tesla earnings call really stand out:
“I think there is generally a lack of understanding or appreciation for the growth of Tesla Energy, as Kunal was talking about. In the long term, I expect Tesla Energy to be of the same or roughly the same size as Tesla’s automotive sector or business. This is the most underappreciated group.”
…and, a short time later…
“But it would be difficult for me to overstate the degree to which, I think, Tesla Energy is going to be a major part of Tesla’s activity in the future.”
I’ll admit I didn’t understand the significance of this at the time, but if my numbers are close to right, I sure do now.
But Wait, There’s More
According to this article from GE last year about the threat energy storage poses to peaker plants, the US will need about 20 GW of peaking power capacity to be installed over the next 10 years, with 60% being installed between 2023 and 2027. If Tesla is the clear market leader, I see no reason why it wouldn’t capture a huge segment of this market, generating nearly $1 billion per quarter from US sales alone.
The world is a much bigger market, though. It isn’t too hard to see Tesla Energy generating far more revenue from this, with potentially significant margins.
I have so many more questions on this now than I did, but I’m really excited to find out more. If I’m close to right, we are about to see a major shift in costs. If Tesla controls the market for batteries priced this low, it will be unstoppable in both the automotive and energy storage business. If all manufacturers’ battery prices are dropping this quickly, the market for EVs and grid-based battery backup is going to absolutely explode in the next few years, which is also good news for Tesla.
A lot of questions remain. What happened to potentially drive costs down like this? Where will Tesla make these batteries? How does it expect to bring capacity up to the levels needed to supply the world with gigawatts of batteries? Is this technology already in use anywhere? Is the Megapack a test bed for these batteries?
Finally, one really important point — I’ve been challenged already on the battery costs when I was discussing this with a different person, one who looked at my numbers and said that the idea that Tesla was paying $190 at the pack level for its batteries at any recent time was far too expensive. I concede that, with how protective the industry is, this could be a high number.
But here’s the thing — even if it the pack number is too high, if we are maintaining a similar margin for the base Cybertruck as we are on the Model 3 Standard Range Plus, it would either mean that pack costs have come down even further than my numbers are anticipating, or pack weights have come down significantly. If it turns out Tesla has figured out even lower-cost batteries than this, everything gets even better. If it turns out that pack weights have come down significantly, that just helps efficiency for its vehicle line up further.
All of the options here are outstanding, and I’m really, really excited to see what the future holds.
— Elon Musk (@elonmusk) November 23, 2019