Mechanical power generation using Earth's ambient radiation | Hacker News Reader

Mechanical power generation using Earth's ambient radiation

science.org

174 pointsdefrost66 comments

Not an engine, but some friends of mine got a mylar sheet that's black on one side and reflective on the other. We tried it out in the desert tied onto trees/vehicles. You put the shiny side down, so the hot IR radiation of the earth is reflected away, and the black side sees the extremely cold (in IR) desert sky. If you put a little hole in the middle and put a bucket under it, you get a fair bit of water, because the mylar sheet gets about 20 degrees C below ambient and a lot of water condenses on it. (even in the desert)

Now you just need a couple of droids and you could go into the moisture farming business

What I really need is a droid that understands the binary language of moisture vaporators.

But can he speak Italian lawn games?

I was going to look into this once, but instead I opted to go into Tosche station to pick up some power converters and unfortunately never quite got around to it.

Nah, those are not the droids he's looking for.

Fascinating 20 degrees C is huge. What's a fair bit of water? At what time of day and how long did you collect water?

These guys claim >40degC, and are deploying in Dubai..

https://www.i2cool.com/tideflow/uwJVdixI.html

https://baitykool.com/radiativeskycooling.html

Peak performance, I think. Considering that they got the black white sides flipped

I didn't do the measurment and it's been a while so it's possible I misremembered the temperature delta, or maybe it was degrees F. It was about a 2-3 square meter sheet and it made about a liter of water overnight.

I think I read something similar in a "Boys survival book, desert chapter" in the early 80s.

Neat. Reminds me that Applied Science made an acoustic radiometer video recently. https://youtu.be/lAeJvZfVLbE

Wow, that's awesome, and a much bigger temperature difference than I would have guessed. Did you get frost?

If anyone is interested in passive sub-ambient cooling (not for power generation, just for "free" cooling) I strongly recommend https://www.youtube.com/@Nighthawkinlight -- he has been doing a lot of experiments in this space and releasing recipes as he goes. Stuff you can do in your kitchen.

DIY radiative cooling paint from YouTuber NightHawkInLight - https://www.youtube.com/watch?v=N3bJnKmeNJY&list=PL1a2HkcVbm...

It has pretty impressive performance.

Tech Ingredients did one or two vids as well - https://www.youtube.com/watch?v=dNs_kNilSjk

Was thinking of whipping up a batch for my rv.

This concept is used all over India to cool down homes that are on the top floors.

https://www.amazon.in/EXCEL-CoolCoat%C2%AE-Reflective-Coatin...

Basically, have a highly reflective white coat on your roof, to reduce temperatures by about 3 Degrees Celsius.

Almost all homes in Urban India are made from concrete and bricks, which can hold a lot of heat.

I myself have been in houses that use this to cover only some rooms of the house (mainly the bedroom), and the temperature difference is definitely noticeable. It also makes the room livable in the extreme hot summers in India.

This is the opposite. It says, "Refelects [sic] 90% of solar infrared rays," because of its "High IR reflective Pigments [sic]," so its emissivity in the infrared is 0.1, but the IR-selective paints we're talking about here are optimized for high infrared emissivity, which means they absorb a lot of infrared.

Maybe there's some wiggle room here because solar infrared is mostly near IR and MWIR, and the place where we want high emissivity (absorptivity) is longwave IR, but to the extent that the advertisement makes any claims about infrared emissivity, it claims very low infrared emissivity, not high.

A paint with low emissivity across the spectrum will slow down the temperature rise when the sun is up, but also slow down the temperature drop when the sun is down. This can still make rooms livable, but it isn't the same as what you get with regular whitewash, where the temperature of the roof is actually lower than the temperature of the air around it.

It kind of blew my mind when I first learned about this whole phenomenon (mostly from the YouTube series I posted). Not all white paints are equal and it’s kind of interesting to think that something that looks mostly identical to our eyes has very different (passive) properties in the infrared.

I think one of the things in the paints that Ben adds is a set of microspheres that reject incident incoming infrared beyond a certain angle but allow it to pass through when radiated. Something like that.

He usually garbles the scientific theory in his videos, but I trust that he's honestly reporting his experiments, and that his theoretical errors are honest mistakes rather than intentional attempts to mislead.

You should be aware that there are rigorous constraints on how much absorptance can differ from emittance, known as Kirchhoff's Law of Thermal Radiation: https://en.wikipedia.org/wiki/Emissivity#Absorptance because without them you could get mechanical power generation from a uniform bath of thermal radiation, which would give you a perpetual-motion machine.

To be fair it does say "reflects solar infrared rays," which I would interpret as "the IR from the Sun" (aka NIR).

The product datasheet[0] claims a thermal emittance (aka LWIR) of 0.82. Having such a high value is typical for non-metallic surfaces.[1]

[0] https://5.imimg.com/data5/CA/RO/MY-653008/excel-cool-coat.pd...

[1] https://ntrs.nasa.gov/api/citations/19840015630/downloads/19...

Thank you for finding this! I was wrong, and I'm pleased that the product is in fact what the great-grandparent claimed. And the NASA publication is very helpful for putting this in context.

IIRC, the papers they're working from mention that lime works very nearly as well as the baryta they're using. Guess what people have been painting their houses white with for several thousand years?

Not with the optimal mixture for maximum-packing of limestone nanospheres.

That effect is almost not perceptible in normal milled limestone.

It would be very surprising if that were true, given the LWIR emittances in the NASA paper cited in https://news.ycombinator.com/item?id=46212591.

I totally believe that you can get significant percentage improvements with the right mix of particle sizes, but the LWIR emittance of just about all nonmetals was upwards of 75% the emittance of an ideal blackbody. That puts a ceiling of about ⅓ (4:3) on the possible improvement on that end. And plain old whitewash has a reflectivity upwards of 90% in the visible, while the best known materials, such as polished silver, are something like 96%, so you might get 3:2 on that end—but probably won't even get 2:1. https://nvlpubs.nist.gov/nistpubs/bulletin/07/nbsbulletinv7n... is one older reference on this, but https://www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=90... gives plots of silver mirrors' reflectance by wavelength, polarization, and angle of incidence.

As a side note, what people have been painting their houses with for thousands of years isn't normal milled limestone; it's slaked lime, which forms limestone by absorbing CO₂ from the air over about a month. Modern whitewash has milled limestone mixed into it, but the morphology of the final surface isn't very similar to milled limestone in a transparent binder.

Vid of the engine in action, from the team that made the paper

https://www.youtube.com/watch?v=5VSmBl8Rv_o

This one shows that it is not as unbelievable as it sounds :)

https://youtube.com/shorts/9KuTdPGqhVo

Somewhat different, but this reminds me of an approach that uses temperature gradients in the ocean to power a heat engine.

https://en.wikipedia.org/wiki/Ocean_thermal_energy_conversio...

OTECs are amazing, and step 1 of "The Millennial Project: Colonizing the galaxy in eight easy steps"[0]

[0]: https://en.wikipedia.org/wiki/The_Millennial_Project

There's a shore-based research OTEC in Hawaii, but the best is a floating, closed-loop OTEC in the ocean.

Interesting link. I would think step 7 would come before step 6 though. I thought about this for a few minutes and can't come with a reason otherwise.

The timelines are increasing powers of 2. It’ll take much longer to colonize all asteroids than to settle Mars.

wiki article states "Up to 10,000 TWh/yr of power could be generated from OTEC without affecting the ocean's thermal structure". which converts to about 500GW which... isn't that much

10 000 TWh/yr is one third of the current total electric energy generation of the whole planet, is not a small amount.

Source, page 39 of the full report:

https://www.iea.org/reports/global-energy-review-2025/electr...

This can't be correct.

10,000 TWh/y = 1e+7 GWh/y, divide it by 365.25 days/y to produce daily output of 27,379 GWh/day, then by 24 h/day to get pure power of 1,141 GW. It's still more than a terawatt, three orders of magnitude larger than the largest nuclear reactors.

oops. yes. still not that much though. i mean it's a lot but it's "one more large industrialized country" a lot not "kardashev 2" a lot

Those goalposts of yours are on a FTL ship...

Kardashev 2 has a Dyson sphere. Of course anything on a single planet can never have that much.

> 400 milliwatts per square meter

About two orders of magnitude weaker than solar panels, even over 24 hours.

E = (T2-T1) / T2

Yes, but it works at night!

Not sure if you can get the MTBF on Stirling engines higher than on LFP batteries, though.

And in the day time, if transparent and applied to solar panels, the "efficiency" gain (~10W psqm) itself will dwarf other considerations.( Remembering that bulk of radiative cooling shouldn't happen below ~1300nm) And then there's beating the efficacy of carbon capture at mitigating warming by orders of magnitudes

Nothing to sneeze at. Just be careful of midbrow high-effort dismissals from the old and wise:)

These guys are applying them to solar panels:

https://www.i2cool.com/tideflow/uwJVdixI.html

  "Case Study 2: Solar Farm in Dubai
  Problem: Solar panels lost 15–20% efficiency at 55°C+ temperatures.
  Solution: Coated panels with i2Cool’s film.
  Results:
  Panel surface temperature: ▼25.7°C (from 58°C to 32.3°C)
  Power output: ▲8% (equivalent to adding 2,400 new panels to a 30MW farm)"

Thanks, that's staggering!

RF energy harvesting in urban areas results in about 0.5-5 mW/m^2. I would guess it would be about 1-2 orders of magnitude less in rural areas.

This is like shaving nickels to make money.

Certainly, there are better energy sources like the fusion reactor in the sky and building a fusion reactor (that's perpetually 30 years away).

TIL: Active nuclear reactors of all types around the world are mappable using antineutrino detectors. It would probably also expose the location of every stationary nuclear-powered ship and submarine too.

>400 milliwatts per square meter of mechanical power with a potential for >6 watts per square meter.

What miracle gets a Stirling engine to be 15x more efficient? Stirling engines have been around for over a century.

Slightly off-topic - I was momentarily excited this was a replication of https://journals.aps.org/prapplied/abstract/10.1103/PhysRevA... (Electric Power Generation from Earth’s Rotation through its Own Magnetic Field)

I've been stalking the citations for this paper for a while now. Surely people would be scrambling to replicate these results. It could truly be transformative for the world if it works and is scale-able.

The science looks good to me

Since we're talking about stirling engines, I've always wondered how using geothermal heat for a larger stirling engine would work.

https://youtu.be/duuk_r--lqU?t=99

Even though the video uses the sun to heat the oil, I would think it would be feasible to use geothermal heat instead.

https://en.wikipedia.org/wiki/Organic_Rankine_cycle

Very fascinating article.

You boil water :)

In one of the later Foundation series books, Isaac Asimov had a whole world run on this.

Which book? Which world? I don't remember this, but it has been a few years.

foundation and earth, if i recall correctly

400 milliwatts per square meter? That's interesting that they can do it at all, but that level is completely impractical for real use.

This is plenty of power to run a microcontroller and a radio (sporadically) with an energy-harvesting setup.

> the generation of >400 milliwatts per square meter of mechanical power with a potential for >6 watts per square meter.

Keep in mind the power is fully mechanical so no electricity or control circuit is required. And based on the simplicity it seems like a good candidate to power something that you need to last 100 years with no maintenance for example.

I think the "last 100 years with no maintenance" is not likely feasible with this approach. The top plate has a coating that supports high infrared emissivity -- and I think it would need to be regularly cleaned to work well. And you can't really prevent it from getting dirty by enclosing it b/c that both substantially changes the performance and moves the maintenance burden to cleaning the enclosure.

Mechanical things don't usually work for 100 years with no maintenance. Bearings run dry, if nothing else.

Air bearings always run dry without problems.

Air bearings run dry until they get some moisture. Then they fail. Old joke about making radio enclosures: make it as watertight as possible, then drill a small hole on the bottom to let the water escape.

Until they are replaced with dust, pollution, hair, animals, leaf litter, aggressive plants, seismic events, pollen, skin particles, birdshit, fallen logs, slime mold, etc.

No moving parts in open water last without maintenance. Life, uh, finds a way.

So what? It's research, not business. Surely you didn't expect they'd found a practical source of free energy that was ready to compete with solar but somehow nobody else bothered to try before?

It is interesting to know if it has potential (pun intended) for some use. Even if that is some very niche thing.

Davis is so frickin windy, wind power would be my first choice.

Secondarily, using a deep ground source heat pump to power a Stirling cycle engine would probably be much more powerful than harvesting a few mW from ambient temperature gradient between surface and air.

That's my 2 centidollars.

Stirling engines are of course fun, but I wonder if the same approach but with the specially-coated radiator on a Peltier instead would net much - it'd avoid the moving-parts problem, at least.

I wonder if you could harness different temperatures at different water depths.

You can, it's called OTEC:

https://en.wikipedia.org/wiki/Ocean_thermal_energy_conversio...

Just like this scheme, it's not very economically efficient.

Carnot efficiency is proportional to the temperature ratio between the hot end and the cold end in degrees Kelvin. If both temperatures are in the 200's, then efficiency will be low.

OTEC does provide lots of potable water though, so that's one advantage.

Cost to build, maintain this machine? $/watt?

Great! Now I desperately need this Stirling engine for my morning coffee: https://a.co/d/6Ja2LeF

Video of how it works: https://www.youtube.com/watch?v=Q5QEBqjkNjo

Reply on news.ycombinator.com