Hacker News new | past | comments | ask | show | jobs | submit login
China wants to put a solar farm in space by 2025 (engadget.com)
36 points by bcOpus 65 days ago | hide | past | web | favorite | 59 comments

A year ago I read the book The Case for Space Solar Power, which goes into the engineering in detail.

The old SPS designs from the '70s were monolithic beasts that would have been ridiculously expensive. Current designs would use a large number of identical parts, of about eight types, which would self-assemble in space. That way they can be mass produced in factories.

The basic design was PV with microwave transmission, which has already been tested over distances of ten kilometers or so. (Iirc it was between a couple Hawaiian islands.) The receiving station sends a sort of homing signal for a phased array transmitter, which makes it difficult to redirect to somewhere without the homing signal and keep a reasonably focused beam, though you could switch between multiple receiving stations (which are fairly cheap). The station would be in GEO.

The book did some detailed cost analysis and found that with pre-SpaceX launch costs, the SPS could provide power at 15 cents/kWh retail.

Elon Musk has said that at scale the BFR could ultimately get to a $50/lb launch cost, so for grins I plugged that number into the book's calculation and got 4 cents/kWh.

Since this type of solar would work through the night, with little or no need for storage, it could be pretty competitive.

book: https://www.amazon.com/Case-Space-Solar-Power-ebook/dp/B00HN...

It would be competitive, if there were no nuclear power or maintenance costs or manufacturing costs. These are not counted in launch costs.

You have to launch replacement panels pretty often due to trash collisions or put them in a less than convenient orbit.

The space solar module with a good enough battery and emitter would cost about as much as half a radiotelescope... Because it is one.

And consider we have problems covering sizable area with wires and cheap panels which are easy to maintain.

The costs I quoted include everything, including manufacturing and maintenance. The book breaks down all these costs. In the 4 cents/kWh scenario, launch isn't the dominant cost anymore, and dropping that further wouldn't make much difference.

You don't put a battery on a solar power satellite. To whatever extent you need battery, you put it on the ground, but you don't need much because a satellite in geosynchronous is in full sunlight almost all the time.

You want to put it in synchronous? How do you keep it lit up all the time?

You do understand you need many more panel fields as opposed to receiving stations that way and half of them is wasted anyway?

I thought the idea is to put them in solar Lagrange or solar synchronous orbit and bend or reflect the maser to stations. 24 hour lit, though a bit trashy if Lsun is used.

The math is suspicious anyway. Wouldn't it mean that the panels would cost 25% of this on ground?

Due to axial tilt, for most of the year geosynchronous satellites are in full sunlight 24/7. Twice a year, around each equinox, there's a period of 44 days during which the satellite goes into shadow briefly each day, for a time ranging from 2 minutes, to 72 minutes at the maximum.


This compares pretty favorably to capacity factors of other power sources.

Illumination per square meter averages quite a bit higher than ground solar, and except for the brief times in shadow is 30% higher than ground solar at noon.

The book evaluated other locations including Lagrange points. Everything had pros and cons; the main disadvantage of Lagrange points was the greater distance, requiring bigger antennas. I think more delta-v in deployment was a factor too, though I forget by how much.

A doomsday weapon in space.

As the article mentions, the station will beam the collected energy down to Earth. Whether in visible light, microwave or any other spectral band, it will need two things for everyday operations:

- ability to point the beam at an arbitrary point on Earth, to track the collector ground installation, and

- ability to focus the beam to proper size

The second capability is crucial; if it's fixed focus - and one with sufficiently large footprint at that - and fixed orbit, it would be safe.

However having either adjustable focus, or adjustable orbit[1] makes this a proper weapon, due to the possibility of pointing multi-megawatt beam of energy at any point on the ground, with little to no countermeasures available. Whether "rogue hackers", or actual owners of the station, somebody would have the capability to destroy, without warning, large swaths of infrastructure on the ground.

I sure hope I'm very wrong about this - that somebody have figured out a way of inherently neutralizing the danger. But for now I treat the project as a doomsday weapon in space.


[1] the later is a given to an extent, due to the need for orbital stationkeeping, and only possible to limit via provision of weak engines and/or limited fuel loadout

It's a pretty shitty weapon because many countries have the capability to launch a bomb to the satellites and the satellites are way more expensive than the bomb.

Depending on how close it is, you could destroy or at least damage the sattelite without even needing to launch anything. As a weapon, it would be downright silly to threaten a major world power with it. Of course if your goal is just to terrorize countries and groups without a space program, and assuming (big assumption granted) that other countries don’t interfere, then it would be a fairly exotic way of terrifying people. VERY expensive though, and diverting it from power generation to “kill” mode would be even more expensive.

All told, a country that could use this as a weapon doesn’t need to use this as a weapon.

It takes a good while to detect preparations for an attack with this kind of weapon & take the decision ("Mr President, the station has slightly shifted around, should we nuke it now?"), and it takes long time for any physical weapon to reach it.

A decent weapon would be space-burst version of ICBM. For intercontinental ground attack, the flight time is typically 20 minutes, but the most reasonable spot for the station is a geostationary orbit - some 36'000 km above equator. With a sensible speed on the order of 8km/s, it would take our missile around 1h25m to reach geostationary orbit assuming a naive straight-up trajectory; with any realistic trajectory it would take way longer. The space weapon would be able to rain terrible destruction in the meantime with impunity.

An alternative would be using ground- or space-based lasers, which give near instantaneous hit, but you'd need to precisely hit a proper part to guarantee fast disabling of the weapon. Merely hitting a random spot on the dish would probably not destroy it outright. You still have to account for the time to make the observations & take the decision. And the current breed of [high power] lasers is single-use only due to the destructive nature of generation of the large energy needed.

That's all true, but a 2h window for destruction of single targets is not very useful against a modern military. There are enough ICBM silos that you can't destroy all of them before someone kills you DEATH STAR and then you still have to win the war that follows.

Indeed, it is not a useful weapon in an all out war. I'd expect it to be used as surgical strike capability instead. Special operations. You can blame its use on inclement weather due to amount of electric charge this will paint on anything in path and resulting lightning storm.

And this is how we end up with https://en.m.wikipedia.org/wiki/Kessler_syndrome

Based on a book I read on the subject, there's a way to make a phase-array microwave transmitter, which achieves the required focus by tuning into a reference signal from the receiving station on Earth. Using this system, if you try to transmit to an arbitrary spot you'll have a much more diffuse beam, and even the tight beam is only concentrated enough to raise the temperature several degrees. The receiving station is about the size of a ground-based solar array, but cheaper because it's just microwave antenna.

Another approach is to use a laser, at a frequency that gets absorbed by the atmosphere. Beam it to a high-altitude receiver station, which beams microwaves to a receiver directly below. But that way is a lot more expensive.


It'll be the largest most fragile thing in the sky - kilometers wide and trivially destabilized/neutralized. Think black paint.

We seem to be arguing It'll be terribly expensive and difficult to make work, on one hand, and It'll be trivially easy to make into a weapon, on the other.

Relatively easy to make work and quite uneconomical compared to ground solar fields. (Multiple.)

But indefinitely expandable - up to millions of square kilometers. Try that next to a big city!

> pointing multi-megawatt beam of energy at any point on the ground, with little to no countermeasures available.

Can't you use a mirror to reflect a significant part of that energy back, wrecking the transmitter? If it is several mirrors focusing on the same point on earth (like a solar thermal power station), just jiggle the mirror on land to sweep the mirrors with a beam of energy.

> with little to no countermeasures available.

In the short term, maybe. But it wouldn't take much time to put in place comprehensive defense for such a system.

In a world with ICBMs, a weapon like this is not really something I'd be worried about. It'd be a pretty stupid weapon to use for an initial attack

The only thing I'm worried about is accidents or hacks.. if that's made impossible it should be fine

>In a world with ICBMs

Quite the opposite. ICBMs have somewhat symmetrical threat and defense model. You use similar technologies for both, and you have actionable amount of time - some 20 minutes - for observation, decision taking, and deployment of countermeasures.

Assuming the station is deployed on geostationary orbit (some 36'000 km above equator), the flight time of the electromagnetic waves is just over 0.1 second. Good luck deploying any countermeasures.

They intend to beam the energy to earth through a laser. I wonder if building a space-mirror and focussing a beam of sunlight would be a much easier and scalable solution.

Though I guess you would need multiple base stations to beam it somewhere where there's no clouds?

Using a wavelength transformation to send this through the non-absorption windows means more energy hits the ground and you aren't hacking terrestrial albedo in a bad way. That global warming is about absorption of sunlight, so putting a lot more sunlight on the same absorption percent means more heat retained. It doesn't do any good to cut gwp materials in the air, then increase sunlight hitting the atmosphere in a way that makes heat retention higher.

Actually, maser - microwave laser.

The mirror is a fine grid which makes it much lighter and more resilient, and it is much easier to convert into electricity than light.

I did a quick search on the Energy Returned on Energy Invested calculations. The following site https://dothemath.ucsd.edu/2012/03/space-based-solar-power/ gave an ERoEI of 10:1 for solar panels on earth vs ERoEI of 4.2:1 for solar panels in space. Which would indicate that we would still be better laying solar panel on the ground as long as we got the appropriate empty space.

What did I missed? Have we realized enough progress since 2012 energy-wise, that the threshold has been crossed, or does it only make sense economically due to lower launch prices.

Solar panels in space also work during the night and while it's cloudy. Did your calculation include the energy costs of storage?

You do know that one can build the panels on the other hemisphere, right? It is all a small matter of building a long enough high voltage DC line. Mostly a political problem. Or you can use microwaves on ground if it is easier to build a bunch of antennas.

Ultimately also reflect off space mirrors (meshes) for much cheaper.

The main problem is theoretically weather. Space has the other kind of weather called space junk.

HVDC lines are really neat tech and I fully support solar farms in the Sahara to supply Europe, but I wouldn't be surprised if it's cheaper to launch satellites than to build a global power grid that can provide solar power over night.

I think debris is not a terrible problem right now. We manage to have a large number of satellites in high orbits.

With regards to climate change and global warming, what about putting this in between the sun and the earth to block some of the energy reaching the planet to allow it to cool?

aka geoengineering. The latest IPCC report is quite critical of it.

I wonder whether this meant to imply military capability, in the same way that the space race was basically a way to advertise ICBM capabilities.

In the sense that fancy hats imply virility.

I was pushing this a decade ago.

If you have "fusion in a bottle" and don't want to share, this is a decent way to make it harder to steal/damage/use-for-terrorism.

Also, giant beam transmission facilities and giant beam weapons aren't so different from each other. There are going to be detectable critical differences, but with some modifications one can be made into the other.

Radiation damage is going to be a challenge but satellites do with solar for a long time. Onboard robots?

Finally, somebody decided to make solar collectors in space. Been waiting for it since it was discussed in Isaac Arthur.

I bet spaceX might do something similar in the future. When they are not as encumbered with star link and the BFR.

1) It is not much more efficient due to transmission losses. (Compared to high voltage DC.)

2) The costs, on top of super expensive solar panels, make this totally pointless economically. Solar panels are hard enough to maintain on Earth.

3) It's a multi gigawatt level microwave gun in space. Want to blast some lightning onto someone? Sure, now you can. Or fry it.

Economics changes all the time. The launch costs are supposed to come down. The cost of getting land for the cells and transmission lines goes up all the time, which makes orbit actually more attractive. The weight/cost/efficiency of panels gets better all the time.

And the efficiency! 10X the solar flux outside the atmosphere! And in the right orbit, 2X again because of course night quits being an issue. So a potential 20X better utility per square.

The costs just to put the materials up there makes me wonder the long term viability of just trying to source it from the moon. By the time you are willing to spend the billions required it cannot be much more of step and provides other uses.

still I have to agree, how do you even maintain something like this. it would probably lead to a full time presence in space unless robotics become just that good. I just don't see the ROI except may be for national pride.

I'd guess you maintain it by replacing the satellites. We're pretty good at maintaining satellites by replacement.

3.5) The giant MASER in space is also a sitting duck target, so it’s shit as a weapon, and disrupting a nation’s power supply is as easy as “shooting” back.

Well... I mean china is pretty capable of breaking the kneecaps right now of whomever they want in their territory. I doubt they will use it as ion cannon.

> Solar panels are hard enough to maintain on Earth.

Isn't the main issue that they get dirty, though?

I think they just degrade the same way they do on Earth. This PDF from 2003 talks about a 0.5% yearly degradation, which, if I recall correctly, is similar to ground based pv.


How do they plan to beam energy back to Earth?

Presumably lasers.

A small part of me wants to see this turn into China deploying a solar powered death laser under the guise of a green energy initiative.

Its in the article

>China's proposal, meanwhile, appears to suggest converting solar energy into electric energy in space, before beaming back to Earth using a microwave or laser and feeding into the grid via a ground receiving system.

Short version: they can fry a target in any country in about a day.

If the country is small enough they can fry that too. I can think of a likely candidate already.

Lasers? Microwaves? I'm just speculating but I don't see another way.

Do they need to? I wonder how feasible it would be to put large computational centers in space and have space solar farms powering these computational centers. So if we want to test complex or long-running programs/models/algos, we can just beam it to the computational centers orbiting earth and wait for the results when it's finished.

How do you cool them in space?

Radiators. Need pretty big ones if the computers want to run cool.

Probably would be better to manufacture most of the stuff on some asteroid or on Mercury and have the computation radiation shielded by a significant mass.

Do you need to? Without enough gas in space, how likely is it for something even as powerful as a data center to overheat?

You have gotten it exactly backwards. Vacuum is a great insulator, getting rid of waste heat is a huge problem in space.

What do you mean? The lack of 'gas' aka material to dump heat into is exactly the issue.

They'll recharge batteries that they'll then return to earth on a cheap BFR-like reusable launcher. /joke

Internet points to anyone who computes:

- how expensive that makes the marginal kWh returned to earth likes this.

- how low launch and return-to-earth costs must go for this to become competitive with current kWh costs.

First you build a space elevator to geosynchronous orbit, then you can use full batteries as counterweight for lifting empty batteries. That way you only have to overcome friction losses.

Even better! Now it's less of a joke idea. You still have a small shuttle transfer from geosync orbit to wherever the solar farm is (heliosync or a lagrange point to avoid earth shadow), but the delta-v for that is tiny.

However space elevators are still soft SF (still missing a material with enough tensile strength), whereas BFR cheap reusable launchers are hard SF, achievable with current known tech.

If there is a beam of energy being transferred from space to the surface will it be visible?

Edit: Apparently this is a stupid question, but I still don't know the answer.

How would they prevent overheating?

How do satellites prevent overheating? Radiators!

I would use this to mine cryptocoins in space, it's way easier to send information back to earth than power

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact