Column There's a fine line between madness and magic in technology. Unless you're talking about wireless power transmission, where woo outweighs watts every single time.
It all started with Tesla – Nikola, not the car company – who got fixated on the idea at the start of the 20th century and built a giant tower to test it out. He spent all the money and never made it work, although his disciples still cling to the wreckage. Not that anyone has learned.
Since then there's been a steady stream of startups promising to push power through the ether to distant gizmos. None of them can make it work either – the physics is just about good enough for small-scale demonstrations for the crowdfunding video, but bonkers thereafter. Think about supplying a town's water by high-pressure jets squirted through the air – and that's much more sensible than the radio equivalent.
There is one respectable area: satellite power stations. Europe, America, and Japan have been steadily refining the idea for decades – another demonstrator is due to be flown in 2023.
Space is a great place for solar power, with a vacuum-clear line-of-sight to the Sun. The physics isn't so bad for a single big fixed microwave link: think waterfall. With the world gagging for clean renewables, a permanent supply from the sky is too good to miss. But will it fly?
There are two main sides to that question, economic and technical. The European Space Agency stuck its finger in the exosphere [PDF] a while back and said space power is possibly competitive with terrestrial renewables at various scales, with a sweet spot around 150 gigawatts. That's between a third and a half of the average European power consumption.
It's a pleasant thought, not just because of how wild it would drive the 5G wingnuts, who go batty at a few milliwatts. Yet safety isn't a showstopper; you can design antenna farms on the ground for power densities overall that fall within current safety guidelines.
Heat, now that's a problem. 150GW deliverable power is a beast. Do the sums on conversion efficiency for gathering solar electricity and turning it into radio waves, and you get around 20 per cent. That means 600GW of heat to get rid of, and forget about cooling towers. The only way to get rid of heat in a closed system in a vacuum is to radiate it away, and that'll need some heatsink. You thought Alder Lake was toasty?
Then there's the question of where to park your smoking sputnik. To be any good, a powersat has to point its panels at the Sun and its antennas at the ground. In medium or low earth orbit, the satellite scoots around the sky, meaning it has to swivel one of the two. Have you ever tried designing a rotating joint that handles gigawatts of power for 20 years or so? No, you have not. Many have tried. It's not happening.
So, let's pop Hotsat One into geostationary orbit. It stays in one place in the sky, which makes things a lot simpler. It's in almost permanent sunlight – let's not worry about the 90-odd equinoctial days with eclipses of up to 80 minutes. If you design your panels so that they're a giant mosaic of small solar cells, transmitters and antennas integrated onto all surfaces, it doesn't even matter that much that the orientation to the Sun is always changing.
But if all your surfaces are active, what about the heat? Moreover, it turns out that cold is just as much a problem, as the panels edge-on to the Sun get very cold, much colder than the electronics likes. The massive thermal swings are not good for reliability, and if you want to make something challenging to fix, put it into space thousands of kilometres away.
Then there's the scale. With about 10 times the sunlight available in space compared to the ground with its weather, atmosphere and nightfall, you don't need as much solar array space up there as down here. You still need many square kilometres of tiles that must be lofted, assembled, and moved to the right orbit. And if you think the radio astronomers are unhappy about Starlink's weedy little two-way radios, wait until those sensitive dishes get the full output of a power station pointed right at them. That's a whole new field of electromagnetic compatibility, right there.
Even if those and an asteroid's worth of other problems are fixable, it'll take many decades of expensive, painful work to get close. Meanwhile, the actual business of building solar plants on the ground where we can visit with a van, and creating better storage and transmission technologies we can deploy with a digger, carries on. We're getting very good at that, and better every year.
The bottom line is that the Law of Commercial Wireless Power holds in orbit just as well as in Omaha: you can build a technology demonstrator that's good enough to extract more funding, but you'll never make it useful. It's even worse than fusion. If you're still tempted, bear one last thought in mind: any idea that's so crazy even Elon Musk steers clear is on the express hyperloop for Wibble City. Train's a-comin'! Woo-Woo! ®