No they don't. Use a distributed database like the bitcoin blockchain. Instead of storing transaction data like bitcoin does, store torrent information and comments. The bitcoin network distributes blocks peer-to-peer, just like bittorrent. When you want to look for a particular torrent, you can access your local copy, or do a lookup across the network from your peers.
> Solar power is worthless, it costs too much and provides power only when the sun shines.
A solar plant in Chile would disagree. It's half PV and half solar-thermal, and provides round-the-clock power. The thermal part heats a working fluid via curved trough mirrors. The hot fluid is stored in a tank until night, when it's used to boil water and turn a turbine. As far as cost, the world is installing 65 GW of solar this year because it's cost effective, not because it costs too much.
> why couldn't Argentina pay their bondholders in Bitcoin,
Bitcoin's total money supply (number of coins in existence x market price per coin) is $7.1 billion as of today. Not all of the money supply is available for circulation. Some of it is lost forever because the owners lost the private key that controls it. Some is being held as long term savings, Argentina's national debt was $330 billion as of last year. Any attempt to buy enough bitcoins to make a debt payment on a national scale would drive the exchange rate crazy, because there simply isn't enough of it, and the exchange rate is set by supply and demand on the bitcoin exchanges.
To put it another way, total monthly volume on all the bitcoin exchanges together is about 1 million BTC ($460 million). The interest alone on Argentina's debt is three times higher, to say nothing of paying off maturing notes. Bitcoin's just too small for a significant nation to use. Zug, Switzerland, yes, they are small enough.
Nobody who understands power grids proposes using *only* solar and nothing else for power. Rational grids use a mix of power sources. Since power demand is higher during the day, solar can supply "peaking" power, while other sources can fill in on "baseload" (the part of demand that's there all the time).
I wish people would stop using this stupid argument. Solar works fine when connected to a grid with other power sources, because *demand* goes down at night too. Guess when peak demand for power is in Las Vegas in the summer? That's why they built a 400 MW solar plant about 10 miles away, it produces the most power exactly when it's needed to run air-conditioners.
It's not one key. It's 20,000 different keys, one for each block he generated in the early days, each with 50 BTC. Given how careful he was in creating bitcoin in the first place, it would be highly unlikely he kept all his keys in one place that could be lost, unless he "lost" them on purpose.
> So.. the orbits of several comets were disturbed by something large moving through our solar system at some point in the past.
Actually, by repeatedly disturbing the orbits of these Scattered Disk objects, so their orbital parameters are opposite the proposed planet. In that location, the planet doesn't disturb them any more, so they stay put. A single pass of a rogue planet would not change the orbits of *anything* that much, and once that one pass was done with, the Scattered Disk Objects would drift away from their anti-alignment.
We see other examples of this "gravity shepherding". For example, Pluto and the Plutinos stay in a 3:2 resonance orbit with Neptune, because Neptune's gravity keeps them there.
The categories of "asteroid" and "comet" are accidents of history, and how they were first discovered. In reality, most every object beyond the "frost line" has lots of water ice and other compounds we consider frozen gases (ammonia, methane, nitrogen, carbon dioxide, etc.). The frost line is the distance from the Sun where water can remain frozen, and thus could condense out of the Solar Nebula. Closer than that and you get dry objects. It happens to be at 2.8 AU, which is the middle of the Asteroid Belt.
If any of these icy bodies from the outer Solar System get too close to the Sun, they become a comet when the ices evaporate. So ThatAblaze wasn't really wrong calling these Scattered Disk objects "comets", any more than it's wrong to call Halley's Comet not a comet in it's current frozen state. He's wrong about everything else, though.
Musk knows about rockets, but I know way more about solar power satellites than he does, because I have worked on the concept off and on since 1980, much of it at Boeing, who did the original studies way back then. He's also involved with Solar City, which makes ground-mounted solar panels. Space solar power is a competing concept, so he has an economic incentive to bad-mouth it.
Of course we looked at the energy conversion chain. You have to remember that the same panel in space starts out with 7 times as much energy to work with. That's because there is no weather, no night, and no atmospheric absorption in space. Transmission to the ground loses some of that gain, but using abundant solar power to manufacture the panels in space means you can make lots more of them to start with. Not just electric solar power, but also concentrated thermal power, to process asteroid and lunar raw materials.
The other reason Musk may be against solar power satellites is they require an industrial base in orbit. If we are mining and manufacturing in space, we have less need to launch from Earth, and that means his SpaceX company has a smaller market.
To start with you would fling raw materials into orbit with an electric centrifuge. This is a rotating arm made of high strength materials like carbon or basalt fiber, and an electric motor and solar array to power it. Since there is no air drag on the Moon, you can bring the centrifuge up to speed gradually. A linear accelerator would be a better option for larger mass throughput (million tons per year), but using a series of coils rather than rails. They have no mechanical contact, so more durable. Accelerations of 300 g's were demonstrated in the late 1970's, which would require a 500 meter accelerator on the Moon.
The coilgun fires about twice a second, so to reach 1 million tons a year you need 16 kg "pellets" of sintered rock, and 50 MW of power, which isn't a starter project.
The radiation belts have high radiation levels, Earth-Moon L2 (behind the Moon) isn't that high. You are gathering and processing large amounts of lunar and asteroid rock. Some of it can be used as radiation shielding, both for electronics, and the humans who will be there too. Automated factories does not mean 100% automated, any more than factories on Earth have no human workers.
Carbonaceous type Near Earth Asteroids are *up to* 20% water or carbon compounds, but not 20% water plus 20% carbon. This has to do with where they originally formed, and their thermal environment since then. The water is in the form of "hydrated minerals" such as Serpentine. Water as a liquid or ice can't survive in a vacuum this near the Sun. The carbon compounds (which are similar to asphalt or oil tars) break down at about 200-300C, which is the same range as the hydrated minerals give off their water. If they formed closer to the Sun, the water and carbon never condensed, it was too hot. If they formed much farther from the Sun, water as water, and other ices, could condense. If it later changes orbit to come near the Earth, what you get is comet, and the ices evaporate. Even in the right temperature zone, you tend to end up with 80% mineral components, because that's the abundance of the mineral oxides they formed from.
Night, weather, and atmospheric absorption at low sun angles means the *average* output of a panel on the ground is much lower than the reference output (clear sky and Sun directly overhead). Weather also makes the output unpredictable.
Space solar power would be available at full intensity 24 hours a day (except for 1% of the time the satellite is in the Earth's shadow). It would supply "baseload power" (always on) rather than "peaking power".
> the advantage in space is less than 7 even for poor-efficiency installs like the 15% CF on my garage roof in Toronto.
You are neglecting that sunlight in space is 36% more intense than at standard sea-level conditions. In cities like Beijing, the ratio is higher due to local pollution.
> you think you can launch an entire industry to the moon for less money
Well, near-lunar orbit, where you get full-time sunlight, but not an entire industry. What you launch is a starter set of automated machines (a "Seed Factory" - http://en.wikibooks.org/wiki/S... ). These machines make parts for more machines, and bootstrap up to an entire industry. At first you have to supplement the machines with items brought from Earth, but as your industrial capability grows, the need for that decreases.
> Gradually dying all the time. That goes directly into the CF of a SPS.
Solar panels on the ground lose output power too, about 0.5% per year. How fast they degrade in space depends on their orbit and details of their design.
> A panel in space will deliver the same amount of power to the grid over its lifetime as a panel on the ground.
This isn't correct. The one in space has 24 hour sunlight, and it's 36% more intense. Ground-based solar cells degrade slower, but the rest of the panel besides the cells (front sheet, frame, back sheet, wiring, mounting) is exposed to the weather, and can fail before the cell output is low enough to require replacement.
I suspect you are trolling, but in case you are not, we have the Moon and 14,000 Near Earth Asteroids discovered so far ( http://neo.jpl.nasa.gov/stats/ ). Solar energy is available nearly everywhere in open space.
My report I linked to in my previous post discusses that we only use 13% of the planet's surface, and very little in the vertical direction. I literally said the same thing you did:
"In a literal sense we are only scratching the surface of our own planet" (next to last paragraph of section 2.2)
My math is only off by 1%. The Earth's equator is tilted with respect to our orbit around the Sun, and so are the orbits of synchronous satellites. They only cross the Earth's shadow during "eclipse seasons" around the equinoxes. The rest of the time they miss the shadow.
There are very limited locations at the poles that have 24-hour sunlight, and because of the low sun angle (1.5 degrees or less), each panel casts a long shadow, which means your area fill is low. You can't produce enough power that way to satisfy Earth's needs.
The standard method is a microwave beam aimed at a large ground receiver. Antenna elements and diodes convert the beam to DC electricity. It's about 75% efficient, and the beam is always on, so you collect more energy per day than the same area covered with solar panels.
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.
> "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.
Slashdot Mirror
You may be interested in my space elevator class notes and slides:
https://en.wikibooks.org/wiki/...
https://imgur.com/a/cCTY5
No they don't. Use a distributed database like the bitcoin blockchain. Instead of storing transaction data like bitcoin does, store torrent information and comments. The bitcoin network distributes blocks peer-to-peer, just like bittorrent. When you want to look for a particular torrent, you can access your local copy, or do a lookup across the network from your peers.
> Solar power is worthless, it costs too much and provides power only when the sun shines.
A solar plant in Chile would disagree. It's half PV and half solar-thermal, and provides round-the-clock power. The thermal part heats a working fluid via curved trough mirrors. The hot fluid is stored in a tank until night, when it's used to boil water and turn a turbine. As far as cost, the world is installing 65 GW of solar this year because it's cost effective, not because it costs too much.
> why couldn't Argentina pay their bondholders in Bitcoin,
Bitcoin's total money supply (number of coins in existence x market price per coin) is $7.1 billion as of today. Not all of the money supply is available for circulation. Some of it is lost forever because the owners lost the private key that controls it. Some is being held as long term savings, Argentina's national debt was $330 billion as of last year. Any attempt to buy enough bitcoins to make a debt payment on a national scale would drive the exchange rate crazy, because there simply isn't enough of it, and the exchange rate is set by supply and demand on the bitcoin exchanges.
To put it another way, total monthly volume on all the bitcoin exchanges together is about 1 million BTC ($460 million). The interest alone on Argentina's debt is three times higher, to say nothing of paying off maturing notes. Bitcoin's just too small for a significant nation to use. Zug, Switzerland, yes, they are small enough.
Nobody who understands power grids proposes using *only* solar and nothing else for power. Rational grids use a mix of power sources. Since power demand is higher during the day, solar can supply "peaking" power, while other sources can fill in on "baseload" (the part of demand that's there all the time).
I wish people would stop using this stupid argument. Solar works fine when connected to a grid with other power sources, because *demand* goes down at night too. Guess when peak demand for power is in Las Vegas in the summer? That's why they built a 400 MW solar plant about 10 miles away, it produces the most power exactly when it's needed to run air-conditioners.
It's not one key. It's 20,000 different keys, one for each block he generated in the early days, each with 50 BTC. Given how careful he was in creating bitcoin in the first place, it would be highly unlikely he kept all his keys in one place that could be lost, unless he "lost" them on purpose.
For technical reasons, the coins in Block 0 (zero) can't be spent (moved).
> Do you give your work/time away for free?
Yes, I do. See https://en.wikibooks.org/wiki/... and check the edits on any page. For that matter, look at all of Wikipedia.
> So.. the orbits of several comets were disturbed by something large moving through our solar system at some point in the past.
Actually, by repeatedly disturbing the orbits of these Scattered Disk objects, so their orbital parameters are opposite the proposed planet. In that location, the planet doesn't disturb them any more, so they stay put. A single pass of a rogue planet would not change the orbits of *anything* that much, and once that one pass was done with, the Scattered Disk Objects would drift away from their anti-alignment.
We see other examples of this "gravity shepherding". For example, Pluto and the Plutinos stay in a 3:2 resonance orbit with Neptune, because Neptune's gravity keeps them there.
The categories of "asteroid" and "comet" are accidents of history, and how they were first discovered. In reality, most every object beyond the "frost line" has lots of water ice and other compounds we consider frozen gases (ammonia, methane, nitrogen, carbon dioxide, etc.). The frost line is the distance from the Sun where water can remain frozen, and thus could condense out of the Solar Nebula. Closer than that and you get dry objects. It happens to be at 2.8 AU, which is the middle of the Asteroid Belt.
If any of these icy bodies from the outer Solar System get too close to the Sun, they become a comet when the ices evaporate. So ThatAblaze wasn't really wrong calling these Scattered Disk objects "comets", any more than it's wrong to call Halley's Comet not a comet in it's current frozen state. He's wrong about everything else, though.
Musk knows about rockets, but I know way more about solar power satellites than he does, because I have worked on the concept off and on since 1980, much of it at Boeing, who did the original studies way back then. He's also involved with Solar City, which makes ground-mounted solar panels. Space solar power is a competing concept, so he has an economic incentive to bad-mouth it.
Of course we looked at the energy conversion chain. You have to remember that the same panel in space starts out with 7 times as much energy to work with. That's because there is no weather, no night, and no atmospheric absorption in space. Transmission to the ground loses some of that gain, but using abundant solar power to manufacture the panels in space means you can make lots more of them to start with. Not just electric solar power, but also concentrated thermal power, to process asteroid and lunar raw materials.
The other reason Musk may be against solar power satellites is they require an industrial base in orbit. If we are mining and manufacturing in space, we have less need to launch from Earth, and that means his SpaceX company has a smaller market.
To start with you would fling raw materials into orbit with an electric centrifuge. This is a rotating arm made of high strength materials like carbon or basalt fiber, and an electric motor and solar array to power it. Since there is no air drag on the Moon, you can bring the centrifuge up to speed gradually. A linear accelerator would be a better option for larger mass throughput (million tons per year), but using a series of coils rather than rails. They have no mechanical contact, so more durable. Accelerations of 300 g's were demonstrated in the late 1970's, which would require a 500 meter accelerator on the Moon.
The coilgun fires about twice a second, so to reach 1 million tons a year you need 16 kg "pellets" of sintered rock, and 50 MW of power, which isn't a starter project.
The radiation belts have high radiation levels, Earth-Moon L2 (behind the Moon) isn't that high. You are gathering and processing large amounts of lunar and asteroid rock. Some of it can be used as radiation shielding, both for electronics, and the humans who will be there too. Automated factories does not mean 100% automated, any more than factories on Earth have no human workers.
Carbonaceous type Near Earth Asteroids are *up to* 20% water or carbon compounds, but not 20% water plus 20% carbon. This has to do with where they originally formed, and their thermal environment since then. The water is in the form of "hydrated minerals" such as Serpentine. Water as a liquid or ice can't survive in a vacuum this near the Sun. The carbon compounds (which are similar to asphalt or oil tars) break down at about 200-300C, which is the same range as the hydrated minerals give off their water. If they formed closer to the Sun, the water and carbon never condensed, it was too hot. If they formed much farther from the Sun, water as water, and other ices, could condense. If it later changes orbit to come near the Earth, what you get is comet, and the ices evaporate. Even in the right temperature zone, you tend to end up with 80% mineral components, because that's the abundance of the mineral oxides they formed from.
Night, weather, and atmospheric absorption at low sun angles means the *average* output of a panel on the ground is much lower than the reference output (clear sky and Sun directly overhead). Weather also makes the output unpredictable.
Space solar power would be available at full intensity 24 hours a day (except for 1% of the time the satellite is in the Earth's shadow). It would supply "baseload power" (always on) rather than "peaking power".
> the advantage in space is less than 7 even for poor-efficiency installs like the 15% CF on my garage roof in Toronto.
You are neglecting that sunlight in space is 36% more intense than at standard sea-level conditions. In cities like Beijing, the ratio is higher due to local pollution.
> you think you can launch an entire industry to the moon for less money
Well, near-lunar orbit, where you get full-time sunlight, but not an entire industry. What you launch is a starter set of automated machines (a "Seed Factory" - http://en.wikibooks.org/wiki/S... ). These machines make parts for more machines, and bootstrap up to an entire industry. At first you have to supplement the machines with items brought from Earth, but as your industrial capability grows, the need for that decreases.
> Gradually dying all the time. That goes directly into the CF of a SPS.
Solar panels on the ground lose output power too, about 0.5% per year. How fast they degrade in space depends on their orbit and details of their design.
> A panel in space will deliver the same amount of power to the grid over its lifetime as a panel on the ground.
This isn't correct. The one in space has 24 hour sunlight, and it's 36% more intense. Ground-based solar cells degrade slower, but the rest of the panel besides the cells (front sheet, frame, back sheet, wiring, mounting) is exposed to the weather, and can fail before the cell output is low enough to require replacement.
I suspect you are trolling, but in case you are not, we have the Moon and 14,000 Near Earth Asteroids discovered so far ( http://neo.jpl.nasa.gov/stats/ ). Solar energy is available nearly everywhere in open space.
> you wouldn't put solar panels in the "average location".
Tell that to Germany and the U.K., who are installing lots of solar panels.
> You don't even have the beginnings of a plausible concept, let alone anything working.
I have a Wikibook partially written ( http://en.wikibooks.org/wiki/S... ), and we have an R&D location under development. What have you got?
My report I linked to in my previous post discusses that we only use 13% of the planet's surface, and very little in the vertical direction. I literally said the same thing you did:
"In a literal sense we are only scratching the surface of our own planet" (next to last paragraph of section 2.2)
My math is only off by 1%. The Earth's equator is tilted with respect to our orbit around the Sun, and so are the orbits of synchronous satellites. They only cross the Earth's shadow during "eclipse seasons" around the equinoxes. The rest of the time they miss the shadow.
There are very limited locations at the poles that have 24-hour sunlight, and because of the low sun angle (1.5 degrees or less), each panel casts a long shadow, which means your area fill is low. You can't produce enough power that way to satisfy Earth's needs.
The standard method is a microwave beam aimed at a large ground receiver. Antenna elements and diodes convert the beam to DC electricity. It's about 75% efficient, and the beam is always on, so you collect more energy per day than the same area covered with solar panels.
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.
> "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.