theres a cock au vin joke in here somewhere.
The geostationary orbit is pretty high (about 22,000 miles) and inclined (to match the tilt of the earth’s rotational axis), so it spends very little time in the earth’s shadow. Most of the year a geostationary satellite is in sunlight 24 hours a day, and during the eclipse season it spends up to 70 minutes a day in the earth’s shadow. And assuming the satellite provides power to the region directly below, the eclipse happens around midnight when power demands are fairly low.
Of course, putting them in higher orbit means that you also have to transmit the power a longer distance, which amplifies many of the problems (your beam has to be better-colimated, your aim has to be more precise, etc.).
I thought I’d mention that one of my former colleagues in the European Space Agency’s Advanced Concepts Team is currently working on future concepts for space-based solar power systems.
No, it doesn’t. As the rest of your post demonstrates, no one even yet has anything approaching 80% efficiency, no matter how defined. It was an absurdly optimistic prediction to make back in the mid-70s, but possibly that’s what they were hoping would bring in interest and funding.
Space based solar power costs too much.
For 1 MW we would need at least 3333 kgs worth of panels. GEO launch costs are roughly $20,000/kg so we’d need to spend $66 Million just to get the damn things into space. That of course would only be a baseline value since we have yet to include the transmitter, scaffolding or anything else needed to hold everything together.
$66 Million for well less than 1 MW worth of power (due to transmission losses) – but let us be generous and give it a cost of $66,000/kW
What about geothermal which has the same advantages (Main costs are upfront, fuel costs are small)
Looks like you could build 13 1MW geothermal plants for the 1 MW solar plant.
Geothermal isn’t the best comparison, since only a few sites are suitable for conventional geothermal technologies (there are developing technologies which work almost anywhere, but I assume those aren’t what the DoE is talking about). You could, though, compare it to Earthbound solar power. Sure, those arrays will be significantly less efficient due to clouds and night and whatnot, and you’d have to pay for the land used, but even so, it costs a heck of a lot to launch anything into orbit.
I hear there’s a way to transmit power over distances where the power per area does not decrease as the distance squared. It’s almost as if the power were a scalar rather than a vector…
If you use laserbeams or microwave beams they spread very little over the distances involved.
But can you imagine what a laser or microwave that can process the power output of small power plant would weigh? At something like 20,000 dollars a pound to put into orbit, that adds up fast.
Like a 1920s-style Energy Ray…
Yeah, but they didn’t have orbital launch capability in the 1920s.
(someone had to give a punchline, or Ludovic’s head would have exploded)
True but the point was space based solar would be a waste of money. Why build a single 1 MW plant when you could build 13 1 MW plants down here? And as the DOE site states the cost for plants over 1MW drops allowing potentially a 26:1 ratio to exist.
What about these giant mirror farms we talked about not too long ago? It sounds like we could build a lot of those things dirt-side and we’d get more energy for the money.
The whole advantage of space-based is the idea that it is not weather or day/night dependent.
The demand for electricity fluctuates, but is very inelastic. We need NOW whatever we demand NOW; there’s no good way to store it. As a result, any power source which is going to provide a meaningful percentage of the total MUST be 100% full time reliable.
If we had reliable trans-global grids, we could mitigate the weather & day/night by shipping power from continent to continent. But that too has costs & losses. Space solevs teh day/night weather thing almost completely, and gains some efficiency by getting the collector above the atmosphere.
In any engineering or business project there are not black and whites. There are simply plusses & minusses to trade off. Right now space solar doesn’t pass financial muster by a sizable margin. Change your launch costs to $2000/KG and things would be different.
One of the driving assumptions of all Space fill-in-the-blank projects is that a large demand for orbital lift will work the usual economic magic of dropping the price dramatically as the volume rises. Which in this case is not out of the question, but is also far from a sure thing either.
No one would have predicted 50 mile daily commutes from the suburbs when looking at the primitive cars and mud tracks for roads commonplace in 1910. It’s the rest of the infrastructure that blossomed *around *cars which made suburban sprawl not only possible, but readily affordable for hundreds of millions of people.
Mildly on-topic but there was a NASA study on how to turn moondust into a mirrored surface. While it required transporting some substances to the surface of the moon, something like 99% of the building material would be moondust.
So an option would be to build mirror farms on the moon that all diverted light to a much smaller number of solar panels.
You would need to build at least two farms on opposite sides of the moon right on the edge of the transition from the light into the dark side of the moon. Most of the time you would only have line of sight to one farm, but it would spare you from having to figure out a way to store energy when the collectors couldn’t send power.
As far as I know 80% isn’t achievable by any form of solar power, as there are physical limits to what can be achieved. But it depends (for solar cell engineers) which efficiency you are talking about, a cell could be close to 100% efficient at capturing photons, but only convert 10% into useful electricity. I guess you mean overall efficiency, which as I said means that 80% is unlikely to ever be achieved.
The limit of efficiency of a single junction solar cell under standard (AM1.5) conditions is 33%.
This can be improved upon by building multilayer cells with different band gaps so that they can capture a wider range of photons.
Other methods of improvement include:
The use of hot carriers so that more than one electron-hole pair is produced by each incoming photon but this can cause device degredation.
Cooling, but this uses energy and may offset any increase in efficiency.
Surface treatment - anti-reflection coatings to trap photons more efficiently.
None of these will bring us anywhere near 80% and that is for silicon or silicon germanium or gallium arsenide devices. Organic solar cells currently work at about 7% efficiency (in the best cases), I don’t know if the upper limit has been determined, but suspect it can’t be much more than a traditional semi-conductor device.
Europe is seriously considering developing space-based solar power to increase its energy independence and reduce greenhouse gas emissions, the leader of the European Space Agency said this week.
"It will be up to Europe, ESA and its Member States to push the envelope of technology to solve one of the most pressing problems for people on Earth of this generation,"Josef Aschbacher, director general of the space agency, an intergovernmental organization of 22 member states.
Previously the space agency commissioned studies from consulting groups based in the United Kingdom and Germany to assess the costs and benefits of developing space-based solar power.
His twitter comment:
Here are the studies:
Ars Technica posters were overwhelmingly negative.
Thirty or forty years ago, I read several books by Gerard K. O’Neill, who described how orbital power stations could be built from material mined on the moon. He also described large cylindrical space stations that could house thousands of people. I was fascinated by the ideas.
A few years back, I saw a presentation on DE-STAR satellites. Although primarily intended as asteroid defences, they could be re-tasked for sending energy back to Earth. The modular design makes it much easier to start building them, and gradually build up capacity.
For North America, a solar farm system in the southwest could probably feed the power grid a lot cheaper - the difference from 30 years ago when orbital solar power was proposed is that the battery tech is good enough today to smooth out the day-night cycle. certainly cheaper that orbiting thousands of tons of material. Making twice the cells and charging batteries is cheaper than launching solar cells. You can buy a home-sized version of this from Tesla already.
For Europe - just put the solar cells in Africa or the Middle East and run the power cables to Italy or via Turkey.
As I understood, the power could be beamed to location from orbit using microwaves sufficiently weak that cows could safely graze under the antenna farm - just the antennas would be spaced every few yards apart within the area of a 10-mile circle.