A couple years ago I read a book[1] on space solar power. It was written before SpaceX really got going, and estimated a cost of 15 cents/kWh for a gigawatt-size solar power satellite in GEO.
Old designs from the 1970s were monolithic beasts, and would be horribly expensive even if launch were free. The new designs use a large number of small identical components (of seven or eight types), which self-assemble in orbit. That way you can mass-produce them.
I plugged SpaceX Starship launch costs into the book's estimate and got a total system cost of 5 cents/kWh, which is pretty good for steady clean power that doesn't need storage.
[1] The Case for Space Solar Power by John Mankins
For space-based solar to be efficient, ie for it to be better than just installing normal solar panels, the beamed energy would have to be at least as powerful as sunlight (1000watts per square meter). That is a very very high bar to overcome. Compared to the simplicity of solar+batteries, I don't see a future for beaming solar energy down from space. Maybe once every square meter of roof space has been dedicated to solar panels, once every city is essentially blanketed in panels, then it may be time to turn to space for more energy.
One might say that space-based solar would operate 24/7 and so is better than normal solar panels. That's isn't reality. Low-orbit satellites spend half their day in darkness too. To hit a receiver on the night side of earth a satellite would have to be at a very high orbit, reducing beam efficiency and increasing launch costs.
Not true. Photovoltaics are ~20% efficient, so you're looking at 200 W/m2 in full sunlight. Plus typical solar capacity factors are about 25% so you're down to 50 W/m2 average power generation, with a subjective penalty for intermittency.
RF rectification is ~90% efficient, so that same 1000 W/m2 gives ~4.5x the power density of daylight PV, plus a capacity factor of nearly 100% gives 18x the energy density of terrestrial PV - and its baseload power.
Mankins will pitch you 100 W/m2 as a useful RF energy density to be competitive with terrestrial solar. You may also be able to include photovoltaics in your receiving antenna and have it both ways.
If you replace microwaves with a laser, you can get much higher efficiencies out of a solar panel as solar panels really want to convert monochromatic light.
What really kills microwaves for power transfer is the minimum sizes of the transmitter/receiver thanks to diffraction. If your going from GEO to ground, I believe you need on the order of square kilometer sized arrays on both ends.
For ground to air applications, rectennas are too bulky and heavy to be practical. If you look at the last NASA power beaming challenge, all of the contestants went with optical power transfer because it was better in W/kg and in W/sq meter.
Disclaimer: I’m involved with free-space optical power transfer.
You're right about the size. The book I mentioned puts the cost of the ground antenna at 0.7 cents/kWh, so not terrible.
Lasers have definitely been considered for SPS. The main challenge is that clouds get in the way. (And you don't want the power density so high that it works as a weapon, but I imagine you can avoid that.)
Was the NASA challenge for SPS, or other applications?
The challenge was nominally for determining feasibility of a space elevator as there was a sister challenge for making high strength/weight cables (it had a bigger problem of the DOD snapping up the promising teams to work on armor every year).
If you keep the power down to the order of 10x sunlight, it doesn’t make a very good weapon.
Well you must be getting screwed over in the US. Québec hydro is 6 cents per kWh, at your house, and is profitable. State owned nuclear energy in Europe has comparable costs.
If only the transmitter and receiver for space solar is a very significant fraction of the cost per kWh shipped at your house with profits of nuclear or hydro it can't be viable.
13.3 cents (us)/kWh is the US national average for electricity cost, not any given region or source.
Canada's national average (this excludes the territories) is 10 cents (us)/kWh. Québec has the cheapest energy in Canada, primarily due to its proximity to hydroelectric sources, at 5.5 cents (us)/kWh. In parts of the US where hydroelectric is the dominant energy source, prices are comparable.
Including the territories Canada's average is 13 cents (us)/kWh.
The book's cost estimate is for a satellite in GEO. You get steady power for all but a few hours per year, and 5.4X as much energy from sunlight in 24 hours, compared to the same flat area of ground. According to the book you get about 50% power loss in transmission.
The satellite design uses a large area of mylar (or similar) reflectors, and a smaller area of solar cells collecting concentrated sunlight. Concentrated solar is costlier on Earth because the mirrors need to be a lot sturdier.
The ground collector covers a fairly large area but it's wire antenna and cheaper than solar panels.
The Falcon Heavy payload to GEO is 41% of its payload to LEO. I assumed a similar cost ratio for my estimate. Starship prices at scale aren't precisely known yet, but low enough that launch would be a relatively small portion of project cost anyway.
Large-scale grid storage is still quite expensive. Most "solar+storage" projects do not have enough storage to get through the night; if they did, it would cost significantly more than 5 cents/kWh.
Those numbers look very attractive. What was the expected lifetime of the system? Did it take increased deterioration into account from radiation/increased sunpower? What about cooling? (or are they just pricing currently available satellite panel technology?)
A big issue with a project like this are unforeseen costs (being a space-based project) and engineering costs (sometimes needs government stimulus). I'm also bothered sometimes those solutions are advertised as panaceas for climate problem; while currently we have mostly a policy problem (clean energy is almost on par with carbon energy, we're just missing key incentives for the new tech and phasing out carbon plants quicker).
But indeed looks like a promising avenue of exploration.
I think they figured only 20 years average lifespan, for the reason you mentioned. I don't remember about cooling.
They had a pretty detailed cost breakdown. The low cost is only once you have mass production and a large plant. They assume several smaller projects first, which would be quite expensive per kWh and include more R&D cost. Those would have to serve communities with very expensive power, like remote northern communities or military installations.
I wouldn't call it a climate panacea since it'll be quite some time before we can do it at scale. In the meantime we should phase out fossil energy as fast as we can with existing technology.
But it looks like in 15 or 20 years we'll have a thriving industry in space anyway, and it's possible that by then we'll be running up against the practical limits of wind/solar market penetration. SPS might play an important role then.
I'm not saying they are, I'm just describing what makes this SPS design affordable. But unlike your link, the design is to concentrate sunlight to solar panels in orbit, then beam power to Earth via microwaves.
Microwave based power transmission systems do not rely on the photovoltaic effect for the conversion of the energy in the beam into electricity. Instead you use something called a "rectifying antenna" or rectenna which can be 85-90% efficient[1]. If you put 1000 watts of power down on a square meter of a rectenna you could pull 850 to 900 watts out of it.
Most of the proposed space-based solar power systems (SSPS) posit a reasonably large rectennas (like maybe 10 - 20 acres or 4 - 8 hectares) sized fields.
The challenge is to make the field low enough density such that things flying through it aren't harmed by it. Fortunately energy density falls off with the square of the area so if you have a enough open space you can make it arbitrarily low power per square foot or meter.
Low orbit sats spend far less than half the day in darkness; it is only when they are in the penumbra of the earth that they are in darkness. What makes space-based power interesting is that such sats have a line to sunlight (and therefore power) when a portion of the planet below them does not, enabling them to provide power for a period of time after sunset and also before sunrise.
> Low orbit sats spend far less than half the day in darkness; it is only when they are in the penumbra of the earth that they are in darkness.
Yes, and over 24 hours, they are 12 hours in darkness: on a 90-minute orbit, 45 consecutive minutes are spent in Earth's shadow => half the day, they are in darkness.
On the upside, they get full insolation during those 12 hours, unlike ground-based solar. On the downside, at night, the satellite will be in Earth's shadow whenever it's above the ground station, so LEO solar power satellites can't supply power at night (unless they carry batteries).
> Yes, and over 24 hours, they are 12 hours in darkness: on a 90-minute orbit, 45 consecutive minutes are spent in Earth's shadow => half the day, they are in darkness.
FWIW I don’t think that’s correct. The effect isn’t significant in really low orbits but going by my hasty maths even at the ISS’ orbit of 350km a satellite spends - at worst - about 10 hours a day in darkness. I say at worst because the plane of the orbit also has an effect here. I make it that you can roughly calculate it as:
24 * ((asin (EarthRadius/OrbitRadius) *2) / (2 pi radians)).
For space-based solar to be efficient, ie for it to be better than just installing normal solar panels, the beamed energy would have to be at least as powerful as sunlight
1) False. You just have to find contexts where installing solar panels is undesirable or impractical. A forward military camp in the hills of Afghanistan?
2) It's quite easy to exceed the energy density of solar panels with a microwave rectenna to receive power. In fact, a lot of the old designs were dedicated to reducing the power density for safety reasons.
One might say that space-based solar would operate 24/7 and so is better than normal solar panels. That's isn't reality. Low-orbit satellites spend half their day in darkness too.
Zero cloud cover. Zero dust buildup. In terms of access to solar flux, there are a lot of advantages to being in orbit. Also, there's a "simple" way to get around the tyranny of the rocket equation and get stuff to geosynchronous orbit cheaply: mine the moon, manufacture the silicon solar panels there, and use lunar electromagnetic mass drivers to deliver bulk cargo to geosynchronous orbit. So I would agree with a lot of what you're saying about impracticality with that caveat: "short of having lunar industrial infrastructure."
But, given a major power that has the above, how is this so different from having fusion power?
EDIT: But if you carry forward this thinking a few steps, you get to...Oh NO!
So let's say that we don't get fusion power, but we do get to the point where lunar industrial infrastructure looks within reach. In that case, control over lunar resources will mean control over the most plentiful, clean, and convenient form of energy. Basically, more than half of the motivation behind major power wars in the last century and a half, has been control over resources, particularly energy. Having energy resources gives one military power which gives control over energy resources.
This dynamic would seem to set up the next major power conflict past the Taiwan issue. Space could well wind up being the Caucasus Mountains/Persian Gulf of the early to mid 21st century. A major power conflict over energy resources which fundamentally involves the power densities implied by space travel just seems like BAD NEWS.
Even worse. We first get the start of the above conflict. Then only afterwards does rapid wartime R&D finally yields military grade fusion power. Yup, we're living in our really nifty, really interesting Sci-fi future. "May you live in interesting times."
No nation could build this lunar mass driver alone. Maybe same thing for a Dyson swarm. So, all those are international projects, if not completely planetary projects. This reduces a lot risks of armed conflicts.
> It is now believed that a lunar mass driver several kilometers long, designed conservatively with present technology, should be able to deliver 600,000 tons a year to L-5, or more easily to L-2, at a cost of about $1 per pound, assuming only ten years of operation.
In this case, doing it internationally is just an option that I hope will be taken...
By the way most huge companies are already international in some sense, geographically or by employing many nationalities and origins. This also helps avoiding conflicts.
For transport back to Earth, possibly not. For solar power satellites, it has lots of metal and silicon. For rocket fuel, it has some water and lots of oxygen. For large space colonies, it's a convenient source of bulk shielding material.
That's often mentioned but there's a bit of a problem there. If you can get net power from fusing He3, then you can also get net power from the easier D-D reaction, and the waste product of D-D is....He3! Half directly, half as tritium which decays to He3 with a 12-year half-life.
Even though D-D emits neutrons, they're at an energy similar to fission neutrons, rather than the extremely energetic (and easiest) D-T reaction. It's almost certainly going to be cheaper to get your He3 from D-D fusion, and get energy in the bargain, rather than sifting through millions of tons of lunar dirt.
Fusion startup Helion, funded in part by YCombinator, is working on a hybrid D-D/D-He3 reactor, saying the combination will produce only 6% of its energy as neutron radiation, compared to 80% for D-T.
>Low-orbit satellites spend half their day in darkness too. To hit a receiver on the night side of earth a satellite would have to be at a very high orbit, reducing beam efficiency and increasing launch costs.
Pedantic, but for LEO there are dawn-dusk SSO orbits that ride the terminator [1] so they get continuous sunlight, you could power some peoples evenings depending on how far the grid spans into the dark side. Not a solution for getting power at 2am though.
Might be able to compete if the cost if a rectenna is cheaper/kwh than solar panels (and free as sunlight). Then you can use your same battery solution for night time.
Solar panels only recover around 20% of the sunlight that hits them under optimal conditions. Typical installations average to about 10-15% of their peak capacity, so they only harvest less than 4% of the ~1000W peak insolation if you average over a year. So if you manage to recover a larger fraction of the microwave beam you can get away with much less than 1kw/m^2. Plus you save on storage costs.
There are groups and conferences dedicated to Space-based solar power systems (SSPS)[1] That the Navy is going to fly this test vehicle is a pretty good endorsement of the concepts.
Has anybody looked at station-keeping issues (briefly searched, found nothing)?
A space based solar power array is a large solar sail. As it orbits, its angle of incidence to the solar wind will change, perturbing its orbit.
Lightsail 2 [1] showed the solar wind can be used to raise orbit. I wonder if such effects would be critical for space-based power systems, or just another design constraint.
Wouldn't such a system still require storage? Their nights might be a little shorter by virtue of being in GEO, but they'd still spend a chunk of each day in Earth's shadow.
Less than you think, because of the planet's axial tilt. You put your satellite in orbit over a fixed spot on the equator:
> satellites in geostationary orbit will spend some time in the shadow during what we call “eclipse seasons.” Each eclipse season lasts 44 days, during which the time that a satellite spends in eclipse (shadow) builds gradually from about a minute or two at the start of the season, to a maximum of 72 minutes at each equinox. It then gradually retreats over the next 22 days, at which point the solar arrays are again in the sunlight on a 24×7 basis.
So you do have a little bit of downtime, but so do coal, gas, and nuclear plants. The few hours of shadow per year still leaves a capacity factor of over 99%.
Old designs from the 1970s were monolithic beasts, and would be horribly expensive even if launch were free. The new designs use a large number of small identical components (of seven or eight types), which self-assemble in orbit. That way you can mass-produce them.
I plugged SpaceX Starship launch costs into the book's estimate and got a total system cost of 5 cents/kWh, which is pretty good for steady clean power that doesn't need storage.
[1] The Case for Space Solar Power by John Mankins