Slashdot Mirror
How Space-Based Solar Power Plants Could Be Built By Robots On the Moon (blastingnews.com)
MarkWhittington writes: The concept of space based solar power has been around for decades. The late Gerard K. O'Neill proposed building them as a way to finance space colonies in the 1970s. Recently Popular Science reported on a modern approach to building space based solar energy stations. Instead of relying on massive, orbiting space colonies filled with construction workers to put the plants together, why not automate the entire process?
> "The problem with regular solar power is that the sun isn't always up." (from the article)
This problem exists on the Moon too. It makes sense to get raw materials from the Moon, but not to put your factory there. It takes about 900 MJ to produce a square meter of silicon solar panel, and their output is about 245 W/m^2 in space. So they make back the energy to copy themselves in 3.67 million seconds, or 42.5 days. Typical working life against radiation damage is 15 years, so the panel can copy itself 128 times in orbit away from the Moon, but only 64 times on the surface, where sunlight is available 50% of the time.
Space Station era space solar panels had a power output of 55W/kg, so a square meter has a mass of about 4.5 kg. Kinetic energy of escape from the Moon is 2.83 MJ/kg, so launching the materials for the solar panel require 12.75 MJ/m^2. The panel in orbit can make back that energy in 14.5 hours, so the extra energy to launch the materials is small compared to the 7.5 years of extra output you get.
Automation was nowhere near as good in the 1970's as it is today, so by all means use automated factories. But put them in high orbit so they get full-time sunlight to operate with. The Moon and Near Earth Asteroids serve as sources of raw materials to feed the factories. The reason you want both is the various asteroid types have different compositions than the Lunar surface and each other. So you get a wider range of materials to work with. In particular, some asteroid types are nearly pure iron-nickel alloy, and others have lots of carbon and water. Those are not easily obtained from the Moon, and any mining engineer will tell you to go for the highest grade ore, because it's less work to extract the product.
Actually, the ratio is 7:1 in space vs an average location on Earth. 24 vs 4-5 hours/day of usable sunlight, and 36% brighter Sun above the atmosphere. The economics of space power then boil down to if you can provide power from space for less than 7 times as much as the same solar system on the ground, space makes sense. Otherwise it doesn't. Per your arguments against:
* Launch costs - The point of using local energy and materials in space is to avoid those massive launch costs. Orbital mining has mass return ratios of hundreds to thousands to 1 (depends on where you mine, and how), so the amount you need to launch from Earth is greatly reduced.
* Expensive maintenance - Communications satellites typically last 15 years with zero maintenance (though they do carry spare hardware). They consist of solar panels, and microwave transmitters. Solar power satellites have the same parts, just way way bigger. So maintenance should be minimal, and what there is can be automated, since the SPS has lots of copies of the same items.
* Expensive transmission systems - Klystrons and Gyrotrons are pretty simple devices. If you can make solar panels in space, you can make those too. You will need thousands, so you would automate the production.
* Large ground-based stations - Solar farms on the ground need that too, so that cost is a wash.
* Beam weapons - The power beam can't be focused smaller than a few km, so the beam intensity is less than or equal to sunlight. The focusing is determined by the wavelength, size of the transmitter antenna, and distance from space to ground. I wouldn't recommend standing in the beam, but I wouldn't recommend being inside a coal plant furnace or a nuclear reactor either.
* Putting 7 times as many panels on the ground - This is the correct answer today. Launch costs would have to come down a lot, or mining and production in space would have to be well developed and efficient for space power to make economic sense. Those don't exist yet, but that is not an argument to stop research. It's just an argument to not build space power plants *today*.
* Self-replication - this is very difficult, but not required. Automated machine tools today can make parts for more automated machine tools. They don't make *all* the parts, just the metal ones. Mostly automated machinery that can make most of the parts in space is sufficient. The remainder of the hard-to-make parts are sent from Earth, and humans on-site or by remote control do the tasks that automation can't handle. You are correct that this works just as well on Earth. A starter set of machines that can mostly copy itself and make parts for other machines is called a "Seed Factory". Working on that concept is my day job. See https://en.wikibooks.org/wiki/... for a path that starts on Earth and uses the seed factory idea to expand into space.
* Moon vs asteroids - The various types of asteroids (metallic, carbonaceous, etc.) are different compositions from each other and from the Moon. Depending what raw materials you need, you will likely want to mine both. Asteroids don't stand still. Even if they have an easy to reach orbit, they are not always in the right place in that orbit. So your departure windows are limited. The Moon has a more limited range of elements available, but it's always nearby, and has a low enough orbit velocity you can mechanically throw cargo into orbit. The right answer will depend on a detailed assessment of actual needs, which as far as I know, nobody has done using up-to-date information.