Megamirror or Windscoop?
I read that some astronomers sought thermal signatures for Dyson spheres in distant galaxies, and found none. I am not surprised, for I think that megastructures of the Dyson/Niven/Banks form are impractical.
Any large astroengineered structures would be heavy, fragile and expensive. They’d also be obsolete once built. They’d need constant upkeep and central control, but with long communication latencies due to lightspeed limits. For instance, in one of Bank’s Orbitals, the AI in the Hub will constantly be 1.5 seconds behind the time, yet it must busy-body in real time. And a ringworld is 16 light-minutes across. Orbitals have awful internets, and ringworlds can have no internet at all.
Orbitals, Ringworlds and Spheres all grab for maximum power, but must be so big that they sacrifice communications speed. How very 20th-century of them. Elsewhere I have discussed my “Wet Star” alternative, which matches power to habitat rather than the other way around. Who needs a star when you have a comet and a fusor? Also, small worlds have fast Internet. Who needs a quadrillion watts when you have a quadrillion apps?
Dyson originally intended a modular Sphere, made of zillions of space habitats, orbiting independently, between them obscuring the disk of their star. I worry about traffic control for such a mob, and the possibility of a debris cascade. A Dyson sphere would inevitably be swarming with junk. Nuts, bolts, paint chips, dropped tools, all flying at orbital speed. (Just like in LEO now!) This plus natural meteorites and micro-meteorites will turn Dyson-sphere space into an environment constantly erosive and occasionally explosive. And the larger the structure, the bigger the risk.
Which brings me to the question: megamirror or windscoop? Imagine two space habitats. One powers itself by a huge mirror; let’s say 100 km^2 in area. The other deploys a Bussard ramscoop to scoop up the solar wind, which it feeds to its fusion reactor.
The mega-mirror system is simpler; all it needs is the mirror and a heat engine at the focus, whereas the windscoop system needs the windscoop and a fusion reactor. But I think the megamirror is more vulnerable to micrometeorites. Even if they only dull its sheen, that would reduce efficiency. Whereas a windscoop is mostly EM fields and empty space. If a paint chip cuts a wire, then just send the robot to repair section 17-B.
I wonder, which has more power content? The sunlight passing through a given km^2, or the solar wind, once put through a fusor? On the one hand, that wind’s kind of thin; on the other hand, E = mc^2.