| Submit your comments on this article |
| -Signs, Portents, and the Weather- |
| Elon's Terrible, Horrible, No Good, Very Bad Battery Math |
| 2023-04-13 |
| [Substack] In 2007, I interviewed Vaclav Smil by email. I asked the Canadian polymath and prolific author a simple question: why are so many people so easily duped when it comes to discussions about energy and power? He replied: "There has never been such a depth of scientific illiteracy and basic innumeracy as we see today. Without any physical, chemical, and biological fundamentals, and with equally poor understanding of basic economic forces, it is no wonder that people will believe anything." I am reusing that quote from Smil (who is one of my favorite writers on energy and power) because it’s germane to a report published on Wednesday by Tesla Inc. called "Master Plan Part 3: Sustainable Energy for All of Earth." The 41-page document is the latest in a shelf-full of studies I’ve endured over the past decade or so that have been produced by academics who work at expensive universities like Stanford, Princeton, and Cal-Berkeley. The studies are packed with elaborate graphics, complicated spreadsheets, and Dallas-size assumptions. And all of them make almost identical claims about how the U.S., or even the entire world, can be powered solely with wind, solar, and batteries, with maybe a lagniappe of nuclear and hydropower on the side. All of them downplay the obvious problems, including the ridiculous amount of land that will be needed, the raging backlash in rural America against the encroachment of big wind and solar projects, the difficulty of building high-voltage transmission, and the need for massive amounts of mining, metals, and magnets to make their schemes work. In all of those facets, the Tesla paper is familiar. And like the others, it doesn’t contain the word "transformer" even though distribution transformers (and large transformers) are critical pieces of hardware and are in desperately short supply. (See my last piece, "Untransformed." -- snip -- But the number that jumped off the screen when I read the Master Plan was this one: 240. That’s the number of terawatt-hours (TWh) of battery storage the authors of the report say will be needed to make the jump to weather-dependent renewables. I won’t quibble over the number even though it sounds awfully low. (Global electricity production in 2021 was more than 28,000 TWh.) Nevertheless, it only takes a minute to understand why that 240 TWh of storage is so gobsmackingly silly. First some basics. Recall that 1 TWh is equal to 1,000 gigawatt-hours (GWh) and that one Gigafactory can manufacture about 50 GWh of batteries per year. Tesla currently has 5 Gigafactories. Thus, Tesla’s current battery storage output is, in rough terms, 250 GWh per year. Now recall that the Master Plan requires 240 TWh, which is 240,000 GWh. Therefore, as can be seen in the graphic directly above, producing 240,000 GWh of battery capacity would require the output of all of Tesla’s existing Gigafactories for the next 960 years. But let’s give Elon Musk and his people the benefit of the doubt and assume Tesla can expand its output ten-fold, to 50 Gigafactories. Even with that mega-expansion, manufacturing 240,000 GWh of storage would still take 96 years. Of course, the difference will be made up in unicorn farts. |
| Posted by:M. Murcek |
| #8 An interesting point, ruprecht. I did not know this, but apparently Tesla has dropped both ultrasonic and radar sensors, and gone all in on optical methods. Welcome to Tesla Vision! Lidar gives you a nice clean 3D point cloud but starts to fall down in dusty or rainy environments. Optical methods and computer vision tricks can use information like color for object discrimination and tracking. If I were me, I'd be trying some sensor fusion from both lidar and visual. But maybe Elon tried that already. |
| Posted by: SteveS 2023-04-13 23:29 |
| #7 That is a really good theory but it has one flaw. The autonomous vehicles part. Musk has insisted on using visual systems for his autonomous Teslas instead of Lider which would seem much better for outer space usage. Maybe there is something about visual that makes it better on Mars with the dust and all, if so I can't see it. |
| Posted by: ruprecht 2023-04-13 21:29 |
| #6 My personal theory, based on no evidence whatsoever, is that Elon cares fookall about batteries here on earth. He is going to Mars and to do that he needs rockets, autonomous vehicles and batteries. Hence, his investment in these technologies. |
| Posted by: SteveS 2023-04-13 17:46 |
| #5 Projecting the future based on current tech is a fools shell game. Musk may find a way to produce far more out of his gigafactories, or they may find the original numbers they were hoping to reach were in reality far lower than they thought. |
| Posted by: ruprecht 2023-04-13 16:45 |
| #4 Someone did a fantastic layman's summary of the laws of thermodynamics: You can't win. You can't break even. You can't get out of the game. |
| Posted by: M. Murcek 2023-04-13 13:59 |
| #3 Also left out of the calculations is a thing called "transform efficiency", which is basically, every time you store, transform, transmit, or use energy there is a loss. Always more than 20%, sometimes north of 80%. You can't get as much power out of a battery as you put into it; you can't get as much out of a transformer as you put in. You notice how these things get warm in use? That heat is the power that's being lost in the transform. It's called thermodynamics. And there's no getting away from it. (It's why they keep trying to develop superconductors.) |
| Posted by: ed in texas 2023-04-13 13:57 |
| #2 There has never been such a measurable depth of scientific illiteracy and basic innumeracy... Now, everybody counts. /pun |
| Posted by: Skidmark 2023-04-13 08:08 |
| #1 Cried a City Hall clerk from Nahant, "Buy a house and a car? Why, you can't! Por favor, don't be pissed -- For mordida, I list You to share an electric Trabant." |
| Posted by: Spereting Cleregum7438 2023-04-13 00:41 |