Germany is going to have to figure out lower cost energy solutions, continue to export wealth in exchange for energy, or turn to nuclear fission as that provides reliable and low CO2 energy. There's other options but they are not pleasant.

There's another option: grid-scale storage.

Historically, the only practical grid-scale storage was pumped hydro. But that has changed. The Tesla Megapack, and similar products from other countries, are truly practical grid-scale batteries. They are expensive... but they pay for themselves quickly.

Tesla was able to make about 4 GWh of Megapacks in 2022. They now have two new factories operating that, when fully ramped, are expected to produce 40 GWh per year each. And if they can get enough lithium iron phosphate battery cells, I'll bet they will scale out their Megapack production even further in the coming years.

Germany can start buying grid-scale batteries, and using those to run their grid at night.

Tony Seba has been predicting for years that wind power and solar power would become so cheap that everyone will switch to them nearly exclusively and back them up with batteries. His think tank ("RethinkX") calculated that the whole USA could switch to solar/wind/battery (SWB) power by overbuilding production by 4x and having enough battery backup for two to four days.

So, I expect that if Germany follows that plan (buy lots of solar cells, put them on roofs, build them onto parking lots, etc. plus get two to four days of battery storage) they could run on 100% renewable power. IMHO they shouldn't have turned off their nuclear power plants but in the long run they should be able to make things work without them.

New technologies are adopted on an S-curve. I think the S-curve on grid-scale storage is about to hit the exciting part and will do something crazy like doubling every year worldwide.

Note: there are other technologies that could be used for grid-scale storage. In particular I'm hoping that Ambri's liquid metal batteries will work out, and flow batteries seem like a good idea.

SpaceX probably could not run ten launches in one day, in this century.

This prediction will age exactly as well as if 100 years ago someone had said "Commercial airplane flights will not have ten flights in one day, in the 20th Century."

it would take over a week to prepare an emergency payload mission

I'm predicting that SpaceX will have their own space station, with a warehouse full of supplies, and a fuel depot. So launching supplies will be routine for them. And maybe it would just be "you don't need to launch your supplies, we'll loan you some we already have in orbit."

Finally, Elon Musk is not the patron saint of marooned spacemen. He still has to make up for the billions lost buying Twitter.

LOL. He doesn't have to "make up" anything; he's still the second-richest person in the world, and in a year or so he will reclaim the #1 spot as the richest person in the world. Twitter isn't losing money anymore, which was the only thing he really worried about; it wouldn't have hurt him but he might have had to sell Tesla stock and he wants to hold on to it.

tens of thousands of satellites into LEO to provide commercial internet services

a) Developing a new technology from scratch (reusable rockets, which nobody else has managed to make) is expensive. I don't mind that they are going to make money from Starlink.

b) Starlink can be used for good things, like education in areas that could really use it.

SpaceX Connects Schools In Brazil's Amazon Region To Starlink Internet

In fact, you could argue it's irresponsible.

You could, if you were narrow-minded.

...without using Google, name the first astronauts to go into space with the space shuttle. Now name the first astronauts to board the ISS. We need more people who are of note, not just scientists, on these missions.

I think you are missing the bigger picture.

SpaceX is making space flight routine. They can deliver more tonnage to space than anyone else, including governments. If their current plans prosper they will have the ability to put more tonnage into space than everyone else put together.

https://www.spacexstats.xyz/#payloads-upmass-per-year

The Space Shuttle was putting professional astronauts into space. These were fine people but they spent years working for NASA and training. What excites me is the thought of SpaceX sending the first welders into space, the first miners into space, heck, the first cooks into space. We are going to have more people in space and the need for every one of them to be a highly trained mission specialist will taper off quickly. "people who are of note"? I just want lots of people.

What I'm really excited by is the thought of actually colonizing places outside of Earth. Not sending a single-use expendable mission to put flags and footprints on the Moon, but build infrastructure and use that infrastructure. Build a "lunar cycler", a ship designed to go from Earth orbit to Moon orbit and back without ever landing anywhere and without any pieces falling off. Build a fuel and supplies depot in orbit. Use the depot to top off the food, oxygen, food, etc. on the lunar cycler. And then what do you have? You have the ability to go to the Moon any time you like, for an extremely low incremental cost.

Now add some kind of lander. Maybe the lander is just another SpaceX Starship. The lunar cycler hauls the lander to Moon orbit, and it lands. Heck, the cycler brings one or two hot spares, so that if there's an emergency, a rescue craft and rescue crew is right there.

After visiting the Moon becomes routine, use the lunar cycler to bring over supplies to build a moon base.

Reusable spacecraft, reusable infrastructure. Once you build it, everything gets easier and faster.

"Once you get to orbit you are halfway to anywhere in the solar system." -- Robert Heinlein

P.S. While I very much enjoyed Andy Weir's novel The Martian I felt that it was implausible. I figured (even back when that first came out) that by the time we are sending missions to Mars, SpaceX would be launching huge amounts of stuff. So the plot point where they desperately needed to send supplies to Mars and wondered where they could find a rocket didn't work for me. I figure if those events actually happened they could just call up SpaceX, and SpaceX would say "sure, we'll cancel one of the ten flights planned for this Tuesday and you can use one of those Starships to put those supplies into space."

The Martian shows space flight as being very much like it was in the 2010's, except somehow NASA managed to build a "Mars cycler" and some Mars landers. How they could do that without reusable spacecraft like Starship is beyond me.

I don't much care for [Elon Musk] either.

Overall, I'm a fan. He's not perfect, but he's on my list of people I think I would enjoy meeting in real life.

I think it's clear that he has exceptional ability in founding and running tech companies. IMHO his best ability is to clearly see the best way to solve a problem and then insist on solving it that way. For example, while people have talked forever about reusable rockets, he insisted on making rockets with a path to reusability and now SpaceX has made it happen. (IMHO the Space Shuttle system was not truly reusable, because it took man-centuries of labor to prep for another flight, and major parts were not reusable.)

Elon Musk is not a perfect person. The lowest thing I'm personally aware of was him calling a guy he had never met a "pedophile". Even his friends said "Uh Elon you better walk that back" and he refused. I've heard some crazy stories (for example, that he fired a Tesla guy for no good reason) but I assume that at least some of these crazy stories are lies made up by people who don't like Musk.

But he is who he is. The good he is doing massively outweighs any bad he has done.

You know you can plug a car into a regular old socket with a cable, right? There's no transformer in the socket on the wall. Even if you have a fancy "charge point", it's just a socket, microcontroller and contacter.

I'm not an electrical engineer, but my understanding is that there has to be a transformer somewhere between the charging circuit and the mains power.

I cannot plug my car into mains power. I can plug in a special charging adapter, which has a box full of electronics, and I'll bet you that there's a transformer inside the box. If there is a car that can plug directly into mains, I'll bet you there's a transformer inside the car.

Charge [robotaxis] when they return to the human-manned depot for the regular cleaning that they will surely need?

That will happen to some extent but I predict that in practice inductive charging will be used so that robotaxis can go charge themselves. We'll see in a few years.

the solution to congestion is not more cars, it's good alternatives to cars

We'll see. I think robotaxis will be an excellent solution... for the lowest prices you can agree to rideshare, and then a robotaxi can take you and at least one other person at the same time.

The big win of any kind of taxi vs. mass transit (a bus, a train, etc.) is that it can take you exactly from where you are to exactly where you want to go. With electric robotaxis, objections based on pollution and resource consumption mostly vanish.

Or if you're parked, use a cable.

Oh wow, thanks for sharing that, so helpful. Yeah, I use a cable just about every day when I park my BEV at home. I'm aware of the option.

much more efficient and compatible will all the existing charging infra.

Did you even read the article I provided, which says that inductive charging is in the same ballpark for efficiency as using a cable?

Gruzen dispelled the idea that wireless charging is inherently much less efficient than using a wire. "When we talk about our system, we talk about them as being 90-92 percent efficient," he said. "But that's end to end; that's from the grid all the way to the battery of your car. And if you were to look at the equivalent for plug-in charging, it tends to be in the mid- to high-80s; [for] the best in the market, it's been about 94-94.5 percent. So we're right in the sweet spot."

Inductive charging can be installed anywhere there's AC power to run it. The charging coil, plus the other coil in the car, combined form a transformer, which incidentally provides isolation (needed in any kind of charger). So it's actually kind of elegant.

And what is your solution for robotaxis... how do they "use a cable"? I provided a link to the video of Tesla's tentacle snake robot, but that has moving parts, while an inductive charging coil just sits there. Seems like a much more scalable solution for robotaxis.

And make no mistake, robotaxis will reshape transportation. Once they start actually working, people will use them, a lot. There will be a lot of them and they will need a solution for charging. The ideal solution would be simple, scalable, and available everywhere robotaxis are in service. IMHO that looks like inductive charging pads, but if you have a better idea please share it.

The idea of cars driving on a highway, at highway speeds, charging as they go from the road... it's solving the problem in the hardest possible way.

A much simpler way: put an inductive charging coil in a parking space, and let it charge a car that is parked in that space.

I now expect inductive parking spaces to be the preferred future way for charging robotaxis. I know Tesla experimented with a tentacle snake charger... the inductive charger is a no-moving-parts solution.

Plus, if this article can be believed, we may actually end up with a single world-wide standard for inductive EV charging, which would be pretty great.

https://arstechnica.com/cars/2022/09/whats-the-state-of-wireless-ev-charging/

Solar PV panels are more efficient when they're cooler.

This is true, as far as it goes.

But where I live, on most winter days, I really do get poor generation from my solar panels. Winter days are shorter, and where I live we have very cloudy weather in winter.

If it snows, and you don't clear the panels, you will get zero power.

But on very cold days with no clouds in the sky, I have seen pretty good generation for the hours we do have sunlight.

I'm looking forward to seeing how much power my roof can make when days are the longest, around the Summer Solstice in June (also known as "the longest day of the year").

But multiple times in winter my roof made a mere 1 kWh for a whole day, which you might as well round down to zero.

Tony Seba and his think tank "RethinkX" say we can get by with nothing but renewables plus batteries, but they do recommend a higher mix of wind power to solar for the northern latitudes.

P.S. If a building is designed to have both solar on the roof and a heat pump system for heating/cooling, the heat pump can be set up to cool the panels by harvesting heat from them. When sunlight hits black solar panels they warm up a bit. This both provides an easy source of some heat and chills the panels to make them more efficient.

I think this implies the use of, specifically, a ground-source heat pump. You would harvest the heat from the solar panels simply by pumping some of the water from the ground loop through a chiller plate on the back of the solar panel, before the water goes to the heat pump.

https://www.greenmatch.co.uk/blog/2015/07/combining-heat-pumps-and-solar-panel-heating

They're just killing it off to move to a higher-margin truck production.

GM lost money on the Bolt. They are killing it in hopes of selling something else that will make money.

It's that simple.

Note that for years people have been saying "Tesla had better watch out... the competition is coming." Well, GM, one of the biggest legacy car makers, couldn't make a profit on the Bolt even after making it for years. Where's that competition?

None of them make their own batteries - even Tesla who use Panasonic and CATL.

Tesla has their own battery factory in Nevada. This is a partnership deal with Panasonic so you aren't completely wrong but it's not fair to say that Tesla doesn't make their own batteries.

And then, for the 4680, Tesla absolutely makes their own batteries. They bought several battery companies to get their technology, and they are working out the problems now. They had hoped the 4680 would be problem-free by now, but it took longer than they had hoped. But you can absolutely buy a car today with Tesla USA-made 4680 cells in it.

https://electrek.co/2023/04/21/tesla-update-4680-battery-cell/

When the Cybertruck ships (probably in Q4 of this year) it will only be available with 4680 cells.

Every time Tesla reported good sales BYD said "that's cute, wake us when you put another zero or two behind that number".

Tesla sells more BEVs than BYD does.

BYD sells more "new energy vehicles" (NEVs) because they have a hybrid, i.e. a combustion engine car that has a small battery and an electric motor. When you put BYD's BEVs against Tesla's BEVs, Tesla is the larger number.

https://insideevs.com/news/631227/top-electric-car-oems-sales-2022q4/

Muddying the waters further, Tesla exports lots of cars from their China "gigafactory". As Tesla's Germany gigafactory ramps up, Tesla will export many cars from Germany instead of China, and will be able to sell more BEVs in China. Tesla is increasing their production faster than any other auto company, so presumably this means that BYD won't outsell Tesla anytime soon, even in China.

Worldwide, Tesla will make at least 1.8 million BEVs in 2023. In 2022, Tesla made 1.3 million BEVs.

BYD made 0.9 million BEVs in 2023.

https://insideevs.com/news/651548/byd-plugin-car-sales-january2023/

I believe your comparison of Tesla vs. BYD is not correct. Your "put another zero" implies that BYD is a factor of 10 larger than Tesla, when BYD is smaller. Even if you add in "PHEVs" BEV is less than 2x as large as Tesla and not growing as fast.

Tesla is now larger than Audi and growing rapidly. By 2027 or so Tesla will be a top 3 auto maker, and by 2030 I expect Tesla will be the largest auto maker in the world.

I don't really understand what's gone wrong with Toyota.

Here's what I know.

First of all, Mr. Toyoda didn't like EVs. It's not a coincidence that he stepped down and then suddenly Toyota is talking about EVs.

https://www.teslarati.com/toyota-ceo-akio-toyoda-steps-down/

Second of all, the Japanese government has been pushing hydrogen, and it might have seemed like a good bet to a Japanese company. I don't really get it... you look at the efficiency numbers for hydrogen vs. batteries, you look at electricity being everywhere and hydrogen infrastructure being non-existent, and I personally would never bet on hydrogen over BEVs.

https://thediplomat.com/2023/01/japan-looks-to-promote-a-hydrogen-society/

Finally, Toyota is heavily invested in combustion and would really prefer to keep making combustion cars. They literally went to the amazingly extreme effort of developing a hydrogen combustion car, which IMHO combines the worst of all worlds. Less efficient than a fuel cell! You will still need oil changes! You can only refuel it at a hydrogen filling station and there aren't many!

https://www.youtube.com/watch?v=vJjKwSF9gT8

IMHO the future is clearly BEVs. If we didn't have Tesla we would still have all the Chinese companies. Toyota was stupid not to figure this out and get on the BEV train as early as possible... but the other legacy car companies were nearly as bad.

The rule of thumb for California used to be that solar panels would pay for themselves in 3 to 5 years, which IMHO is "quickly" in the world of infrastructure. But the net metering program in California was just changed to make it a much better deal for the power companies and a much worse deal for the customers, making the new payback period more like 9 years. It's up to you whether 9 years is "quickly".

https://www.reuters.com/world/us/us-home-solar-installers-brace-slowdown-california-reform-looms-2023-04-10/

IMHO if you want to make money, you would invest in the stock market (Tesla stock is a particular good buy right now IMHO). But there are nice side-effects from having solar power, and it will eventually pay for itself.

Tony Seba has said that he wants to see the power grid of the future be like the Internet. He wants to see people be able to buy and sell power in a spot market.

IMHO a great step forward would be to make the power companies charge separately for "grid connection" and for power delivery. I think it is fair for people who use the grid as their battery to pay a grid connection fee to help the power company maintain their infrastructure.

Also, grid-scale batteries will change everything. Power companies have trouble managing the solar power from distributed sources (like the roofs of customers' homes) because the power arrives when the sun is good, not when the power company needs it. If the power company had enough batteries, they would be happier about accepting all that power.

Which form of energy supply is easy to defend against aggressors?

How about a decentralized mesh network?

My own house now has solar panels and backup batteries; it produces quite a lot of power. To completely deprive me of power you would have to go to my house and sabotage it. That wouldn't exactly be hard, but if all the houses in my neighborhood also had solar power with storage batteries, it would add up to a lot of work to take them all out.

If a substantial fraction of houses were net electricity generators, and most of large industrial buildings and large parking lots also had solar... and let's imagine a bunch of Tesla Megapack installations scattered around the grid to stabilize it and keep it reliable... it would become difficult to get the whole thing to fail.

I actually really like the idea that all the parking lots and parking garages and industrial buildings with huge flat roofs should all install solar panels. They will pay for themselves, and in places like California they will pay for themselves quickly.

Note that France passed a law requiring parking lots to have solar panels.

https://cleantechnica.com/2023/02/09/new-law-50-solar-power-over-parking-lots-in-france/

Rhode Island is considering a similar requirement:

https://electrek.co/2023/04/07/solar-new-parking-lots-buildings/

P.S. From time to time I wonder what would happen if an enemy lit off an EMP bomb above where I live, or if we had another Carrington Event.

I wonder if you could design solar panels with a Faraday cage adequate to protect against EMP that nonetheless let the sunlight through to get the power. If it were possible, I wonder how much more expensive it would be.