Hacker News Reader

I disagree with the niche market thing.

#1 Military forward operating base #2 Anything that has to be in the middle of nowhere, think mining and agriculture, maybe even oil rigs #3 Shipping, it could make electric container ships viable. #4 Anyone who wants to rent their power, rather than build their own panels. You're shifting CapEx to OpEx. This is why cloud computing took off. You can have any amount of power you need at the flip of a switch. People will pay a massive premium for that.

For a given mass of solar panels, the space version is perhaps 50% more efficient, due to the fact they operate in space. Putting them there, in the most optimistic scenario, will cost perhaps 100x versus the ground level version. Currently it's something like 1000x, but let's asume prices go down due to Starship and what not.

So you really have to ask what kind of practical scenario affords you to use energy that is at least 60 times more expensive than what is available at ground level (but more realistically, at least 300-400x). Even if you include the capital cost of Li-ion batteries, their wear and tear and road transport for over 1000 miles, you can still get power much cheaper by just trucking batteries around. That, mind you, is before any transmission loses, investment in ground infrastructure etc., it makes no sense to put solar panels in space because the efficiency gains are simply too limited.

This leaves the military scenario indeed - which is basically sci-fi, but bras has been known to abandon common sense when planning technology, see the initial Space Shuttle concept, space laser weapons etc.

People drastically underestimate how Starship will change launch costs. Right now they throw away an expensive upper stage with every launch, and even the lower stage only gets used about a dozen times. Starship doesn't throw away anything and it's designed to be used about a thousand times. At scale it's basically just the cost of fuel and ground services, which drops cost by a couple orders of magnitude.

The book The Case for Space Solar Power has detailed cost estimates for NASA's SPS-ALPHA design. It was published about a decade ago so has old launch costs. I plugged in Starship costs and got a total cost of $0.04/kWh, which is pretty great for dispatchable power without needing storage.

The real money comes from using starship to develop spaced based industrial capacity. Instead of manufacturing solar panels on earth, it would be manufactured in space. Once we have real space industry, doing things in orbit would be much cheaper.

> Instead of manufacturing solar panels on earth, it would be manufactured in space. Once we have real space industry, doing things in orbit would be much cheaper.

The raw materials would still need to be shot up into space. And what makes manufacturing and assembly in zero-g cheaper than on the Earth's surface?

The idea is probably using ISRU(in-situ resource utilization) with regards to local asteroids to save on launch costs. Manufacturing/assembly would likely be more expensive at first until the techniques and technologies are learned(perhaps one could upsell this as "truly green manufacturing" since there would be no environmental byproducts to affect Earth).

> perhaps one could upsell this as "truly green manufacturing" since there would be no environmental byproducts to affect Earth

Except for the massive byproducts created on earth to get through the "more expensive" phase. It would be deeply deceptive and cynical to sell that as "green", surpassing even today's deceptive selling of carbon capture and storage technology on fossil fuel power plants.

But less and less is launched per pound of space-produced product as time goes on, until only rare complex (small) components such as microprocessors come from earth. There will be an industrial chain taking aluminum from lunar regolith that's catapulted out to a Lagrange point and then the raw aluminum or printed products will be lowered to more suitable orbits by ion engines that spit out ions from lunar soil.

Hate to be the one to mention it, but the military may well be paying most of the freight to set most of this up (covertly or not) since space based you-name-it including hypersonic projectiles are going to be more and more useful as the decades pass.

Starship uses methane fuel. Make that from water and ambient CO2 using clean energy sources, and the environmental impact of launch will be minimal.

Fuel is a pretty small cost for rockets anyway so using more-expensive clean fuel wouldn't be prohibitive.

Specialist components will be made on earth for some time obviously... but there are obvious benefits to maximising the volume of non mass limited payloads for things that get turned into structural components. The obvious first will be anything made from structural plastic, there are 3d printers pioneering printing in zero gravity, with some extra work you can scale that up to make a nice big printer that wouldn't even have to worry about sag, cause of the no gravity... It may not be able to print a pressure vessel, but if you need a million meters of wire conduit and plastic will do the job. Why ship up 10 launches of lightweight payload, or do all the logistics to lift, store, handle, and disperse the payload onto multiple launches with more dense payloads to get the correct volume/mass ratio... or just launch a printer you can keep using, and launch two huge payloads of printer feedstock.

This might be the most nonsensical proposal I have read in a while.

Because why should we build things on the ground and deliver the finished products in orbit, when we can push an entire industrial chain into orbit and furnish it with the vast quantities of required inputs, while taking advantage of the vast nothingness that exists there, and the 100000 times more expensive labor. It's like Elon Musk's Mars city, but in vacuum, with zero in situ resource utilization.

I think you’ve misinterpreted my point, I agreed with the point your making about complex supply chains and components. I specifically said that its only going to make sense for some specific items, and I was specifically thinking the first candidate would be bulk low density support structures. For all the structural aluminium isogrid beams, and perovskite solar panels that are the end products of specialist supply chains with the exact issues your talking about

In a large scale on orbit construction project, like a space based solar power station, we could manufacture complex interlocking units on the ground like space station modules on the ISS, and hope all the power routing, and docking, and everything else, will work together when launched… or we can build it the old fashioned way, but with robots or astronauts up in space… bolting, welding, wiring, cabling and routing everything up there from easily made standard parts that meet the requirements. Just like large photovoltaic solar farms are built on earth, with X units of photovoltaic panels, attached to structural support frames suitable for the location they are being installed at (wind loads, thermal range, etc) … and everything cabled together with hundreds of little plastic clips and miles of copper wire in weatherproof plastic conduit.

Now some things will obviously be different in space, like hundreds of plastic clips are obviously a space debris risk, so they probably won’t design cabling systems that need that sort of support. But they will need to manage lots of power and data cables which will need some level of protection from the slow but steady degradation by various forms of radiation and plasma from the Sun and deep space, and it turns out a few millimeters of plastic can get you years and years of protection from oxygen plasma which would could eat away at exposed wires and slowly increase the risk of short circuits. Yes it might need to be replaced regularly, but so will other things, and then you don’t need to use crazy specialised oxygen plasma resistant wire insulation…

Space is hard, but we build things for way more harsh environmental conditions on the ground all the time in industrial applications. What they don’t need that space currently does, is earth based industry doesn’t usually care about total weight. Starship will revolutionise space not by magic, but by slashing the cost per kilo and allowing us to use more normal engineering procedures and hardware where appropriate. Like panels and wiring and support structural girders. My point was that 3d printing lightweight plastic structural parts where they are needed, will be to space based solar power, what mixing concrete on a construction site is when compared to having everything transported to the site via cement mixers. Sometimes it makes sense for the project to mix on site, other times it won’t.

I’m not sure what there is to argue with about this thesis.

> My point was that 3d printing lightweight plastic structural parts where they are needed, will be to space based solar power, what mixing concrete on a construction site is when compared to having everything transported to the site via cement mixers. Sometimes it makes sense for the project to mix on site, other times it won’t.

I argue that any kind of large scale structure fabrication is fundamentally different than the preparation of raw inputs, like your concrete example. A large orbital 3D printer is a complex industrial machinery that needs servicing, constant supervision and human involvement, it's more akin to the Hubble telescope than to an concrete mixer. The costs to operate it in space are orders of magnitude higher than on the ground even in the most optimistic orbit cost reduction scenario, short of space elevators and the like.

Until we have self replicating and selfrepairing orbital robots, any space project will need to contend with the fact that human labor in a spacesuit in orbit is inconceivably more expensive than a guy riding his bike to the factory in sweatpants, breathing the free air our planet seems to have aplenty and drinking the water that falls from the sky from time to time. And this is, I repeat, a fundamental stumbling block no matter how far launch vehicles advance, space will always be insanely hard compared to surface.

So you better have a damn good economic case for printing structures in orbit which are 1000x more expensive than the same structures built on earth. The savings from the support structure and the lack of foldable mechanisms necessary to deploy into orbit could be significant, but definitely less than a single order of magnitude.

Costs for both paths will go down steeply over time. But launch costs can only go down so far given the sheer amount of fuel necessary to launch anything into orbit from earth. Remember maintainance robots will come, not just more reliable and cheaper printers that are very lightweight (used at the end of a long flexible arm.)

3D printing aluminum isn't, you know, rocket science.

You don't seem to resonate witn the basic premise of this discussion: the mass of the structure must stil be lifted into orbit in the form of raw materials, printer fillament, etc. We were debating the finer points of wheter the high cost of space fabrication is covered by space optimized design that will never need to experince atmospheric drag, gravity, launch forces, vibrations and volume constraints etc.

The mass of the structure need not be entirely sourced Earth not should it. Space industries must eventually transit to resource extraction from the rest of the solar system rather than solely from Earth.

Labor is extraordinary expensive, yes, at first. That is because our robotic technology isn't good enough and that people don't live and work in orbit yet. That will change with time as we begin space colonization.

No, soon enough the vast majority of mass will be catapulted from the lunar surface. There's oodles of aluminum in lunar soil.

> soon enough the vast majority of mass will be catapulted from the lunar surface

Other than naturally falling asteroids, to date we have managed to acquire about 400Kg of moon rock at a price point of roughly $200bln in today's money. That's about $500k per gram.

These speculative predictions about moon & asteroid mining imply technology advances that are so far into the future that they are basically mindless daydreaming. Things like self replicating robots, nuclear space-ships and other sci-fi staples.

We are decades or centuries away from that and the impact of such tech cannot be predicted - they will change society so much that we cannot even begin to imagine their impacts. Sort of like people in the 19th century imagined a future of personal balloons, not being able to imagine the impact of airplanes and the automobile.

It's early days, but costs per pound to space are going down fast. The tech itself isn't science fiction we can sinter a cannonball from lunar dirt or extract aluminum and make a ball to put in a catapult. It's just plenty expensive as yet, but costs are coming down.

You're right, according to me, about getting minerals from the asteroid belt, for earth use. That's not happening until the farther future: the delta is extreme, and down to earth itself not trivial. But extraction from the moon or mars moons (about the same delta to a lagrange point as the moon believe it or not) can be economical in bulk precisely because the cost to get to earth orbit can only go down so far (the cost of the fuel).

Fulton used to loiter around the fences around his steamship as it was being built; you should find and read all the scorn every single passerby had for the mad, doomed project that seems pretty ordinary now.

When I was young, maybe 1963, I had dreams in which people in the future were carrying around small devices with screens they kept looking at, about the size of my fathers pocket calender planner (3" by 6" maybe). Not a wholly original idea, Dick Tracy comics showed a TV watch back then.

My father, a very competent electronics engineer who was very well read, poured absolute scorn on the idea. He didn't just say that no known tech could do that, he said that the device was strictly impossible, and that no matter how long the world lasted, no such device would or could exist. His reason: you can't flatten a cathode ray tube that much. Time easily dispenses with our reasons why not.

In space they generate ~7x as much power averaged over the day (24hr sunlight, plus full power all the time) and during the winter in the upper latitudes it has got to be at least 100x the power you get from one on the ground.

> 24hr sunlight

That means you either need to sync the satellite with the Earth's rotation (Sun synchronous orbit), so you will deliver power at the exact time the corresponding Earth region has full sunlight (and thus, a surplus), or that you will have to go at insane altitudes. Or that you would need to launch some sort of orbital power ring that can transport the energy around the world.

Good-luck beaming power from SSO or across spans of thousands of miles. First let's see a phone that charges from across the room with no matter how bad efficiency, or an electric car charging wirelessly on the go.

The US military has already made a space-based test of power transmission from orbit via microwave (hundreds of miles down). Dunno if they published the results, possibly not.

> Good-luck beaming power from SSO or across spans of thousands of miles. First let's see a phone that charges from across the room with no matter how bad efficiency, or an electric car charging wirelessly on the go.

This is the actual impetus behind the push. Much like nuclear, the militaries of the world want a new toy (focused beam weapon), but want to pass the buck to the civilian population and claim that it's abundant cheap power with no risks.

I think the costs are going to be astronomical, and that's the difference with cloud computing - if it cost 10X as much to use AWS as it did to run your own servers onsite with IT people to manage them, AWS would never have taken off. Likewise, for mining and agriculture, onsite solar/storage is almost certainly going to be much cheaper. Electric container ships are an interesting idea, but even there it might actually be cheaper to just synthesize fuel from water and air.

In particular, transmission losses and low efficiency seem to be the fundamental problem. See previous discussion:

https://news.ycombinator.com/item?id=31521665

AWS is 10-100x the cost of on prem. I've got a client who is paying about $300k per year for a postgres instance that is basically my laptop. I think S3 is a loss leader and EC2 is close to cost. Everything else is a massive markup.

> Electric container ships are an interesting idea, but even there it might actually be cheaper to just synthesize fuel from water and air.

How does one do this without power? And if possible, why isn't it done already?!

You’d do it at the port with nuclear/wind/solar. No one does it now because marine fuel is super cheap.

> How does one do this without power?

You synthesize the fuel when and where you have power.

> And if possible, why isn't it done already?!

Hasn't hit cost parity yet. Hydrogen is the cheapest and it's still about triple the price. More portable fuels are more expensive to synthesize and cheaper to mine. Add a carbon tax (bunker fuel is filthy) and watch it happen overnight, or wait until around 2030 when the prices will cross over (but there will be an equilibrium price while capacity is built).

You say "I disagree with the niche market thing" but then as a counterexample give four niche markets.

Container ships are a huge market.

Anyone who rents power is a huge market too. AWS are buying diesel generators to backup/supplement the grid.

#2 Radio communications, mountain-top repeaters. Not dropping $4k on chopper rides to battery swap would be a boon for the environment and bank accounts.

Powering cargo planes seems like a niche market that would be suitable. The planes could have just enough power to get up above the cloud layer, then receive a charge boost each time they go below each solar powered satellite. This makes the planes lighter as they no longer need to carry as much fuel or batteries.

A second consideration is that these satellites can be factories in space. Raw materials are much easier to launch into space using rail guns as the materials can survive higher g forces then humans or pre-manufactured goods. Once in space, the raw materials can be made into higher quality goods due to the lack of atmosphere and gravity. Once they are in space they can be put to use making further satellites, orbital rings or habitats.

> Powering cargo planes seems like a niche market that would be suitable

FTA: Most designs aim to produce a beam kilometers wide so that any spacecraft, plane, person, or bird that strays into it only receives a tiny—hopefully harmless—portion of the 2-gigawatt transmission

2 GW on a _single_ km² is 2kW per m², about twice what solar power provides. “Kilometers wide” could easily mean 10km², or 200W/m².

⇒ it will be difficult to get the power density to power an airplane, even if it’s only for flying at height.

There are already planes that can fly (and do nothing else useful) on 300W/m^2

Solar panels already have sufficient mass power density. There's a pretty big gap between the 10s of MW a jumbo needs and the 700m^2 of area it has. If you were willing to trade some flight speed you could increase wing area 5x or so, and maybe make it work in the 2-4kW/m^2 range if all of the photons were at your bandgap.

It sounded like the reason for making it km wide was to avoid creating a gigawatt death laser - I assume if they were doing a flight assist thing they would send less total power and could concentrate it more

There are probably more than a few people at the Department of Defence that get very exited about the prospect of having access to a death ray that could fry anyone we happen to have the coordinates for.

Getting funding out of the military budget could speed this up immensely. So you went into green-tec to improve the world? Sigh.

I think your maths is off - there are 1m square metres in 1 sq kilometre, but 2000 kW in a GW, meaning approx 0.002 kW per square meter.

Which kind of makes this whole thing a weird economic proposition. You’re beaming energy from space at a density that approximates solar insolation, and still have to build the collector and convert it back - wouldn’t it be easier to just collect the solar insolation?

> I think your maths is off - there are 1m square metres in 1 sq kilometre, but 2000 kW in a GW, meaning approx 0.002 kW per square meter.

That's not right. There's 2000 kW in a MW, not GW. 2GW is 2e9W, so over 1e6,^2 that's 2e3W/m^2 or 2kW/m^2.

Ahh thankyou for sorting me out. My apologies OP

There are, in fact only 1000 kW in a MW, or 1,000,000 kW in a GW. Where is this factor of 2 coming from?

Call me old fashioned, but the thought of putting a massive, heavy tube into the air full of cargo without enough energy to get it to its destination seems like an idea that will get a lot of people killed. I can't even get reliable satellite TV, I don't want "keep a plane of Amazon packages from landing in my attic" reception to be a thing I need to think about.

It's not too hard to engineer a system that could handle missing a beamed refuel. One could achieve a lot of benefits from a hybrid turbine that uses fuel for take-off and "emergency events", and powered electric from satellites. I'd assume that there would be a whole constellation of satellites, so if one is missed the next would be within range, or the plane could land safely with a backup liquid fuel supply. Ideally the plane would have a high glide slope ratio. Satellites with beam forming technologies could also map out thermals and plot out ideal courses.

This is all a thought experiment on achievable engineering. Have fun with it.

Or, we could just say that two day shipping for all goods wasn't ever a sustainable idea and use—and hear me out on this—boats. Boats don't sink when they run out of fuel, and can be reliably run on electricity. Heaven forbid we sacrifice the convenience of Amazon's Buy Now button in the interest of the planet not frying to a crisp.

The primary use case here is trans-pacific. Not a lot of attics out there.

"Yea.. if it crashes, it just crashes into the ocean. What would the problem with that be?"

Move fast and break things as applied to air transport really makes for some exciting opportunities, doesn't it?

We can either be infinitely scared and do nothing or we can actually make progress in a responsible manner and get a chance to fix climate.

This is not a responsible solution, and believing so doesn't belie some "infinite fright," nor does it discount the thousands of more reasonable solutions that could be applied here first.

If you honestly thought "space based power beams" were a realistic solution, then why you would even propose a plane as opposed to a boat is entirely beyond me. This idea that we're just going to find some "plug in" solution to our current infrastructure that's suddenly net positive for the climate only invites these kinds of irresponsible flights of fancy.

If you want an honest solution, then you're going to have to reconsider the mechanisms that cause you to fly cargo across the globe on jets in the first place. These systems were built for profit, and they continue to operate with profit as their primary motivation. If you want to manage the climate "responsibly" then you need to put it first, and when you do so, you realize that our current means of distributing products around the world is patently insane.

Only by ignoring that problem or hoping that literal "power beams from outer space" will magically become available and prevent any restructuring of profit flows leads to this kind of nonsense thinking.

The answer is not to do nothing. The answer is to reject bad ideas that would turn cruise ships into potential floating craters and instead pursue ideas that are both tractable and won't treat people as collateral damage.

> bad ideas that would turn cruise ships into potential floating craters

Is the concern here is a sudden loss of power or autonomous flight?

If a beam is lost, these electric cargo planes will have enough batteries for 20-30 minutes of flight until power is restored. In the worst case, they would have to make a safe emergency landing in the ocean without hitting ships.

If the concern is about autonomous flight, then it's not really different from many other autonomous systems: they absolutely work, but have to be implemented responsibly and pass regulatory audit.

> will have enough batteries for 20-30 minutes of flight

Where does this number come from, exactly?

Also, why not circumvent these questions by using electrolysis to make zero-carbon fuel from sunlight, and just pump that the regular way?

Or ships, which can be wind powered most of the time and backed up by PV-H2 and/or beamed power in the doldrums?

We can probably also use lightweight PV coatings (or beamed power) on a blimp for air travel that won't crash when power fails. (But probably only for cargo; the Hindenburg meme will probably prevent passenger interest).

>> will have enough batteries for 20-30 minutes of flight

>Where does this number come from, exactly?

It's a ballpark. Electric planes are capable of having up to 1 hour ([1]), but I've taken a smaller number to allow for more cargo instead of batteries - there's not much difference between 20 minutes and an hour in terms of finding a safe landing.

> Also, why not circumvent these questions by using electrolysis to make zero-carbon fuel from sunlight, and just pump that the regular way?

We should try almost everything. A massively parallel approach to reducing CO2 emissions has a higher chance of success.

> Or ships, which can be wind powered most of the time and backed up by PV-H2 and/or beamed power in the doldrums?

Same - we should try all technically-sound options. Wind, green hydrogen, biofuel, ammonia, solar, nuclear, beamed power in the doldrums. Non-practical approaches will vanish, practical will find their niches.

> We can probably also use lightweight PV coatings (or beamed power) on a blimp for air travel that won't crash when power fails. (But probably only for cargo; the Hindenburg meme will probably prevent passenger interest).

Somebody should try that too.

1. https://en.wikipedia.org/wiki/Pipistrel_Velis_Electro

Yeah, that's old fashioned. A fully loaded 747 flying at cruising altitude will be able to go between 108 and 144 miles before landing if all the engines suddenly died (the difference is because of the fairly large range of heights that constitute "cruising altitude"). More modern planes can go even further from the same height. Planes designed to optimize that number can reach 360-480 miles if the engines fail (but obviously those planes carry less cargo).

Unmanned cargo planes with fail safes over ocean routes doesn’t seem too dangerous.

There's no such thing as a fail safe for a plane running out of gas.

Sure there is, prepare to safely crash with 5% left

> It's just far more efficient to build solar PV arrays on the ground, and to keep power on 24/7, build enough to store the energy for later use at night and in the winter. Throwing all that mass into orbit, even if in low-Earth orbit where decay and burnup in the atmosphere is guaranteed over time for the debris issue, is just a massive waste of resources, plus the energy loss over 100 km or so would undoubtedly very high.

That argument makes perfect economic sense ... now. The same way that burning fuel still makes economic sense in quite a few cases. But environment anxiety is not only about burning fossil fuels, it is about deforestation, farting cows, the use of land for human activities, and well, basically every industrial economic activity. There is environmental anxiety about solar and about wind energy production, and about nuclear power. And it is often an-out-of-the-blue spurt for something that happens in another continent[^1]. We are slowly but surely channeling that anxiety by taxing "Earth use".

Here is a thought experiment: try to imagine a world where human environmental footprint is bound in a way that satisfies even the most extreme environmentalists today, who--as anxiety raises--are going to be your average bread-and-butter (self-)conscious citizen of tomorrow. How does it look? How many of us?

I wish we come upon a time when we decide to have space-based solar power and space-based living. The alternative is disquieting.

[^1]: https://www.bbc.com/news/topics/cjyykdwmw58t

Emigration and performing energy intensive work where the heat can just radiate away makes sense, but a world in which we pump 100s of Terawatts through the atmosphere as microwaves because we are dissatisfied with the 100s of terawatts falling on areas we have already paved is not one driven by environmentalism, but by avarice.

There is some truth in that.

But of course we can collect solar and wind power on crop- and pastureland while still using it for those things. I doubt we will be returning much of either to nature, even though about 1/3 of the US maize crop now goes into vehicle fuel, which demand ought to start to fall off eventually. In principle that land could be returned to nature, but I doubt it will be.

> It's comparable to notions that fossil-fueled global warming could be fixed by parking a giant sunshade out at Lagrange 1.

Why can't it work? Critics of this idea often cite cost or unintended consequences, but not ineffectiveness at mitigating the temperature rise (ocean acidification etc. are a different issue).

My criticism is a massive L1 sunshade does absolutely nothing to solve the underlying problem, merely mitigates one of its effects while suppressing photosynthesis, reducing that aspect of the planet's ability to attack the underlying problem. There is only one solution to the problem which is to stop dumping carbon into the atmosphere and accept the consequences for our lifestyles. The Earth will take care of the rest.

This has been debated at length at the past, so I'll pass on that discussion.

It was your interrogative, so I suppose if you open the discussion, you can also close it. An exceptionally wise choice.

But those solar arrays on the ground are a target for light reflected from space at night, too. Throwing that into the space industry picture helps the costs.

Also, the first accident (there is always one) redirecting the beam to a populated area is going to be so spectacular I doubt it would recover after it.

Accidents probably won't be an issue (useful power density is surprisingly safe, and safeguards can be put in); malicious redirection of multiple beams, that probably is a threat.

Multiple beams can only be guaranteed safe against malicious agents by making sure only one is in any given person's sky at any given moment, which either limits us to 6 worldwide or requires the satellites to be low enough that they're in earth's shadow.

(Last time I said this on HN, I got an angry response from someone who clearly didn't understand what I was saying; I don't know if that is because words are bad at painting pictures or not, but regardless I really ought to get around to blogging this with pictures to show what I mean).