Hacker News Reader

Comment by akira2501 | Hacker News Reader

Thankfully, the planet contains abundant stores of Uranium.. and we are nowhere near peak efficiency in commercial and industrial processes. These one dimensional analysis of the problem are particularly useless, especially when their only conclusion is "we must stop growth."

The average power usage may well be 2000W, but how many average consumers are there? If we're facing a multi-mode distribution of those users, then this "stop growth" strategy instantly creates new "castes" from this distribution. As a long term strategy, it seems doomed in one way or another.

View on HN

The article accounts for this in average earth temperature. The earth’s surface would be hotter than the sun in 1000 years, and would generally be uncomfortable after 200. The only way around that would be a substantial change in the thermodynamic efficiency of our energy use and/or a non radiative means of disposing of heat.

Granted, if we did have the equivalent energy of a sun - we are unlikely to be spending it all on earth. Escaping earths gravity would be a trivial expense.

That thermodynamic argument was the least convincing part. If you wanted to, you could beam your power plant's infrared radiation towards space, or put the plant itself in space and have it beam its waste heat away from the planet. The idea that heat dissipation is limited by the black body radiation of the Earth assumes that your house is part of the power plant's radiator, which itself implies that you will be living in a power plant's radiator. :-)

That's because you don't understand thermodynamics.

You can't just put the power plant in space and keep the heat from generating the power away from the earth, because the heat gets created where the power is used to perform work. You have to avoid consuming any of that power on the earth entirely.

And you can't just beam the waste heat from every process away from the surface of the earth, that violates the second law of thermodynamics. The heat exists because the energy performed work, you are keeping the energy organized by beaming it in any direction, so it is not waste heat, it is direct infrared from the energy you've produced. You can only do this by not actually using the energy for work. This makes the whole exercise of even producing the energy pointless.

I think the biggest hole in the argument is the jump to 100% efficiency. we don't consume mor energy because we feel like it, we consume it because we need it, and a factor of 5 increase in efficiency would equate to a factor of 5 decrease in production of energy resources.

>you are keeping the energy organized by beaming it in any direction

I am afraid it is you who misunderstand thermodynamics, my friend. :-) Entropy is a matter of degree, and as long as the beam leaving the Earth is less ordered than the beam coming down, it can work. For example, imagine a hot nail on the moon suspended a few inches above a mirror. The heat will radiate into space and the moon won't get hotter.

Yes, but that energy cannot be used for work on the moon...

Presumably the work is what heated the nail up in the first place…

Yes, and if that work was done on the moon, the nail wouldn't be the only thing that was hot.

> And you can't just beam the waste heat from every process away from the surface of the earth, that violates the second law of thermodynamics.

Surely this can't be right, because if it were, air conditioning (the normal everyday kind) would be impossible, right?

What we're talking about here is loosely analogous to air conditioning the entire Earth (with the heat being sent to space rather than the outside air). Sounds like a hard engineering problem, but not impossible.

Using power to generate work is not what happens in air conditioning. With air conditioning, you are moving heat from one spot to another (another word for it: "heat pump"). That is why it's over 100% efficient.

Generating economic value from work that the OP describes would not be "air conditioning" it would be converting heat into work -- which cannot be over 100% efficient, and in fact would be significantly less than 100% efficient at the temperatures we know.

Removing that heat using a heat pump would require some off-world reservoir of cold, which doesn't exist. The reservoir of cold that we are using now (and in any conceivable future) is the earth itself, which is what the parent comment means by "you can't use the work."

> some off-world reservoir of cold, which doesn't exist

Space has an average temperature of 2.7K, and there's a lot of it. Can interplanetary space be an off-world reservoir of cold?

It is. We receive light from the sun in the visible light range. Earth emits something like 20 photons in the infrared for each incoming visible light photon. The reason is that the sun’s surface temperature is ~20x higher than Earth’s.

So now we are getting into albedo and black body radiation, but the short answer is that you cannot radiate heat out from earth (or any body) into the vacuum faster than it radiates out by itself.

An air conditioner performs the work of moving heat from one place into another. There's a reason your compressor is outside, that's because if it were inside the heat pump wouldn't work. You're expending energy to move heat generating more heat, you'd just heat up your house. The compressor wouldn't work outside if the heat wasn't convected or conducted off the radiator into the air, there has to be a temperature differential between the ambient environment and the radiator.

With the earth, there's nowhere to pump the heat to, there's only vacuum. You'd have to convert the heat to something else like light, radiate it out, expend energy to do it, and in the process generate more waste what from that energy that will then radiate out at the natural pace. At best in this process, you'd break even and just be really expending energy beaming light out into space and accomplish nothing with regard to the heat you're trying to move.

> Surely this can't be right, because if it were, air conditioning (the normal everyday kind) would be impossible, right?

Keep in mind that air conditioners are net generators of heat. The work they do moving heat around generates more heat. If you build a sealed sphere 100 feet in diameter around an air conditioner and plug it in, the air inside the sphere will heat up. This is why you cannot just open your refrigerator door to cool your house. Earth's atmosphere is warmer today around cities where a lot of air conditioners are in use. (Although excess CO2 has a much bigger effect of course.)

> What we're talking about here is loosely analogous to air conditioning the entire Earth (with the heat being sent to space rather than the outside air). Sounds like a hard engineering problem, but not impossible.

It's possible to make radiative coolers that do exactly this, within limits. It's easy at night; harder in the daytime. Here's an example.

https://www.futurity.org/radiative-cooling-system-2141732/

That's not a very descriptive or explanatory article, I'd be very interested to learn more technically about this device if you happen to know or have anything.

It's basic thermodynamics; if you leave a tray of water outside at night and you protect it from wind (convection) and insulate it from the ground (conduction) and there are no clouds, it will radiate heat away into the much cooler night sky. You may have ice in the morning, even if the ambient air temperature is above freezing.

(This is also why cloudy nights are warmer than clear nights in winter--clouds prevent some of the ground's heat from radiating away.)

The technique has been used for centuries to make ice in the desert, but it only works at night and it's inefficient.

https://www.fieldstudyoftheworld.com/persian-ice-house-how-m...

With modern technology it's more efficient and it appears it can even be made to work in the daytime:

https://energyindemand.com/2020/10/17/generations-after-peop...

https://onlinelibrary.wiley.com/doi/10.1002/adpr.202000106

But the energy transported to Earth from your space power plant still creates waste heat when it is used to do work (and also when it is transported to earth). You cannot beat the second law.

Let us, per the article's title, do the math...

* Total energy consumption = 10^26 watts (hope the caret comes through HN formatting)

* Assume some magical generation and transport mechanisms which don't involve energy gradiants (avoiding pesky thermodynamic realities), and are 99.9% efficient.

* Total waste energy is 10^23 watts.

* Radiators are on Earth, and we dedicate 1/2 of the entire surface area (not just land) to radiators. Earth has about 5.1 x 10^8 km^2, so say 2.5 x 10^14 sq. meters.

* Radiators for the waste heat would be dumping 10^23 W / 2.5 x 10^14 m^2 = 4 x 10^8 watts per square meter. That's 400 megawatts. Per square meter.

* That's a black-body temperature of 9,165 K (the Sun's surface is 5,772 K). Peak wavelength is 316 nm. So half the Earth's surface is a glowing UV light pointed at the sky. I don't think we'd have to worry about the ozone layer disappearing. (A very bright, inefficient UV lamp--plenty of infrared, hard UV, and soft X-rays to go around!)

* OK, so put the power plants in geostationary orbit. Say 1 million power stations, each with radiators 1 square kilometer.

* 1e6 stations * 1e9 square meters = 1e15 sq meters.

* So, that's only 1e8 watts/meter! Hmm. Still 6,480 K, but peak is blue-violet rather than UV. Still lots of UV and infrared, but it's space, who cares. Just don't cross the beams (ha ha, meaning no deep-space traffic). And mind the Moon! (Wouldn't want to give anbody on earth a reflected sunburn).

* The radiation pressure from the waste engery for one power stations would be about 730 kN (160,000 pounds). Would need counter-balanced beams.

* Geostationary orbit is about 26,000 miles. The surface area of a sphere that big is 2e16 square meters. So about 1/20th of the sky would be covered with powerplant radiators.

And this is just for the hypothetical 0.1% waste energy. Multiply all of the above by 1,000 for the actual used energy. So you'd have half the Earth's surface as power receivers slurping up 400 gigawatts/square meter of X-rays and hard UV. Each power station would see a force of 739 mega-newtons; you'd have to have a counter-balancing beam pointing away from Earth. And then you'd need to radiate away all that energy after it was used.

Edit: Oh, and while 1/2 of the Earth is receiving 400 GW/m^2, the other half is using that amount. Which would be the real limit. I don't think anybody would carry around a 10 GW iPhone 237+.

Why not have one square meter of radiators that emits gamma rays?

As I recall, gamma rays cannot be generated by thermal processes such as electron orbit decay - so we would need a nuclear radiator which converted waste heat into atoms which would rapidly decay into gamma rays.

Assuming we’ve done all of that… you would still need to deal with the rocket engine you’ve created. Sending out a tight beam of a suns worth of energy is equivalent to a rocket engine with 1x10^18 newtons, creating challenges of its own - including an appreciable acceleration of the planet earth.

A more plausible escape hatch for the energy may be a black hole heat sink. In theory, such a device could perfectly “recycle” waste heat into usable energy.

Oops, I miscalculated the total satellite area; (meant 1e6 x 1e6 = 1e12). To keep everything else the same, replace "each with radiators 1 square kilometer" with "each with radiators of 1,000 square km."

Large parts of the Earth's surface are going to be generally uncomfortable within 10 years, never mind 200.

Large parts of the Earth's surface are going to be generally uncomfortable now, never mind 10. Of course it'll be worse by then.

Large parts of the Earth surface were generally uncomfortable in 1850, so that's not really a useful statement.

There's no way we achieve the higher powers of ten without a big infrastructure migration to space. Really that's the promise of those higher powers, you move from a terrestrial to a at least a solar civilization.

Only a small fraction of Uranium is usable for fission without breeding. There is (IIRC from a LFTR youtube) 4-10x as much thorium as uranium.

The other issue is most designs are solid fuel rod, and they waste a huge percentage of the fuel once "spent".

Nuclear isn't price competitive with existing wind/solar, and any new nuclear project is 10-20 years away, and with wind/solar in substantial/active cost improvement and economies of scale, nuclear may never be competitive.

I loved those LFTR presentations, and I would like to think a scalable LFTR solution could be made price competitive with wind/solar once it stabilizes, but right now probably not.

LFTR can use thorium (still needs a fissile uranium starter plug), and can probably breed non-fissile uranium, I've heard it can breed existing nuclear waste into usable fuel as well. It uses 99% of the fuel according to the sales brochure, which I view as a massive huge advantage: it takes all the handwringing and NIMBY about waste storage out of the equation. IMO that should be a table stakes for any "new nuclear" design.

It's basically meltdown proof with the plug, isn't pressurized, and somewaht self-regulating in temperature if I understand the neutron economy and expansion/contraction of salts. Also a "table stakes" aspect of any new nuclear project IMO.

Scalability is probably the biggest path forward to competitive fission as well, the original MSR at ORNL was allegedly closet-sized. Massive concrete dome designs lend too much to boondoggle management and cost overruns. Scalability would enable nuclear to compete with the current kings of scalable power generation: wind and especially solar.

Maybe pebble bed can get a lot of the same table stakes as LFTR. Maybe a LFTR could serve as an onsite fuel reprocessor for a more economical design if LFTR isn't broad market feasible. I'm not a nuclear engineer, I'm just a fan of the ultracool engineering and science.

But the economics don't line up for nuclear currently.

These analyses have been predicting doom for decades. They just keep moving the goal post.

Huh, I had to re-read the article.
"But let’s not overlook the key point: continued growth in energy use becomes physically impossible within conceivable timeframes." (conceivable timeframes ~ few 100 years)
As far as I understood, the argument is not "we are doomed" but that there is a natural limit to the growth of energy consumption in the not too distant future using some naive assumptions. In any case, it is implying that 2.9% growth per year is a lot! if it is extrapolated by just a few 100 years. Do I misunderstand something? Why is this controversial?

But I did not yet read all the linked later posts.

You are correct.

It's not profound to say that sustained (infinite) growth is unsustainable, but it is of interest when it's within a conceivable timeframe.

It's not worth reading however, because the author whimsically extrapolates energy usage starting from 230 years before the invention of the lightbulb to get their exponential growth argument which doesn't hold at all for modern data.

https://data.worldbank.org/indicator/EG.USE.ELEC.KH.PC?end=2...

Reply on news.ycombinator.com