Hacker News new | past | comments | ask | show | jobs | submit login
Efficient Air-Conditioning Beams Heat into Space (ieee.org)
404 points by uoaei on Sept 5, 2017 | hide | past | web | favorite | 187 comments

Huh - I'm always interested in ways to make A/C more energy efficient. Squeezing and un-squeezing gas seems so COMPLICATED compared to how easy it is to heat things up. Entropy is a cruel mistress. Anyways, it sounds like this is the important bit:


So Raman and electrical engineering professor Shanhui Fan made panels containing layers of silicon dioxide and hafnium oxide on top of a thin layer of silver. These radiate in a unique way: They send heat directly into space, bypassing the Earth’s atmosphere. The panels do this by emitting heat at infrared wavelengths between 8 and 13 micrometers. To these waves, the Earth’s atmosphere is transparent. What’s more, the panels reflect nearly all the sunlight falling on them.

For the new fluid-cooling system, the researchers made radiative panels that were each one-third of a square meter in area; they attached the panels to an aluminum heat exchanger plate with copper pipes embedded in it. The setup was enclosed in an acrylic box covered with a plastic sheet.

The team tested it on a rootop on the Stanford campus. Over three days of testing, they found that water temperatures went down by between 3- and 5 °C. The only electricity it requires is what’s needed to pump water through the copper pipes. Water that flowed more slowly was cooled more.


So it sounds like how space stations kick out heat; The ISS uses that sort of thing[1]. You flow coolant through things that heat up, like solar panels or a warm environment, and then you pump it into these bricks that don't absorb heat from light. The coolant warms up the bricks, which emit that heat as infrared radiation. Planetside, I guess that you can't let the infrared radiation be absorbed by the atmosphere, or it won't cool anything.

So, maybe this could take the compression out of our current cooling systems - compress/heat up, expose to outside temp, let cool, decompress/cool, expose to inside temp, repeat.

My main question would be, is it capable of cooling beyond the ambient outside temperature? If so, how? I might be misunderstanding or missing something about how the 'blasting energy away as IR radiation' thing works.

[1]: http://www.lockheedmartin.com/us/products/HeatRejectionRadia...

I'm one of the authors / people working on this. Yep, radiators like the ones on the ISS are pretty much the dominant/only ways to get heat out in space.

In response to your questions: yes, it can in fact cool below the ambient air temperature entirely passively (even during the day! -- which adds real value for cooling applications). We've shown in other recent work that you can use this effect to cool as much as 45-50°C below the ambient air temperature, if you insulate the radiator perfectly.

The reason this works here on Earth is that some upward thermal radiation isn't absorbed and re-emitted back to you. So if you start at the air temperature, looking upwards, you will be sending more heat out than the sky sends back to you. This allows an upward-looking surface to cool itself down until the heat going out and coming back to it balance out.

Can you apply this technology - coupled with AeroGel based insulation of radiators - to shipping containers/container walls?

This would allow for a really low cost refridgerated/temp-controlled containter. Then have a small requirement for a solar panel to pump some fluids. The result would be the ability to have shipping containers that can maintain temps with extremely low cost/energy requirements and you might revolutionize the ability to ship foods around places like Africa.

Also - can this be applied in any way as a "paint"? Such that you paint cars, roofs or whatever with this?

This is brilliant! Specifically the container idea!

> 45-50°C below the ambient air temperature

That's incredible! Isn't that enough to run a heat engine on? If so, does this allow you to extract (nearly) free energy out of the ambient heat?

Have to consider the amount of heat moved, too. It's the same difference as between pressure and volume, or volts and amps. Perhaps it can't move a large volume of heat quickly, but can achieve a decent temperature difference given enough time?

And even if it can, I am reasonably confident (from my arm chair here) that this combination of relying on a narrow band of passive IR emissions to drive a heat engine will be less effective than a photovoltaic cell generating electricity directly from incoming visible radiation across a relatively broad frequency band.

Ah, but only one of those still operates in the dark.

50C is not much for heat engines your limited to under 15% effecency in the best theoretical case. So solar + battery is going to be a net win unless your thinking of flipping panels at night for ~1% more power over pure solar panels which is silly from a cost perspective.

I thought solar was widely favored long term, but it's going to be a long time before it establishes that position.

That differential could run a Sterling engine.

That differential was quoted for perfect insulation. Plugging in a heat engine is the opposite of insulation.

My god, that could be a great asset to communities in water-starved regions of the planet.

It all comes down to units of thermal rejection per area per unit lifetime cost. If either one of those values is less than desirable expectations will fall short.

Does this work when humidity is high? Water absorbs thermal infrared and emits it back.

You can still get cooling, but you get less heat rejection per unit area in high humidity because of the effect you point out.

As someone who is very unknowledgeable about this subject area, what kind relationship between humidity and lost heat rejection would there be?

I ask because as someone who has lived most of his life in southern Florida, this kind of device could reduce electricity bills massively around here, as long as the humidity isn't too much of a problem

You heat the surrounding humid air instead of sending the heat into space.

And that surrounding air will radiate thermal infrared back heating the panel. As was pointed out in another thread, this still may work, but not as efficient as long as some infrared is not absorbed.

If there is even a slight breeze blowing, then wouldn't that move enough of the heated humid air out of the way?

No significant difference. The point is that the more opaque the air is to infrared, the more your panels are in thermal equilibrium with the surrounding air (ambient air temperature) rather than space (2.7K). They're in equilibrium with a weighted combination of those. Doesn't matter whether you heat up the surrounding air by 0.01 degrees.

> if you insulate the radiator perfectly.

I don't quite understand what that means. If your radiator is supposed to radiate away the surrounding heat, shouldn't it on the contrary be connected as much as possible to the environment?

my understanding is what you're talking about only applies for traditional conduction/convection, where you're expecting moving air/water/... to act as a heat-transfer medium, so you want to maximize your surface area of your radiator and fluid flow rates to facilitate that transfer.

what they're doing here is shifting the transmission characteristics of radiation to be within the infrared window (https://en.wikipedia.org/wiki/Infrared_window) so that it's kind of like surrounding air (and water vapor, most importantly) aren't present at all and you're radiating in to the giant cold heatsink that is space.

No, because the heat is radiated through infrared -- not conduction and convection. Any heat transfer by conduction/convection is going to be warming the thing you want to cool and will be working counter to the purpose of the device.

Conduction, convection and radiation.

Two questions:

1) Does it emit more IR at higher temperatures? In other words, would pumping hot water through/under it produce larger temperature drops?

2) Is it weather proof, or do you still need to find a coating that protects it while not blocking the IR?

Yes, it emits more radiation/ heat when hotter. A nice feature of radiation is that scales as T^4.

It's pretty durable on its own. That being said, as we've moved to the product stage we've packaged it in a way to protect it further.

I'm not sure why, but I've always thought it interesting that radiation scales as a fourth power. It's unusual to see forth power relationships in fundamental scientific processes.

Black body radiation scales with T^4, but you're relying on one particular band. I suspected it scaled, but T^4 seems rather unlikely ;-)

Most materials scale with T^4. You would need a very weird radiance spectrum to do anything different.

That said, we are talking about something that has a weird radiance spectrum as a selling point, so, expect some deviation from the standard T^4 curve.

I would assume it scales at K*T^4 where K is a constant < 1 and gets smaller as the band narrows.

Could you point the beams at something and heat it up? It seems like a waste just sending it into space.

I can't think of any scenario where you would be better off recapturing that IR beam instead of installing photovoltaics and capturing the stronger sunlight.

Also, removing heat from the Earth is probably a positive at this point in time.

The problem is that the heat is in the wrong place, right? We're not going to stop needing electricity. If you've got an IR beam and you can point it at something you can heat whatever it is and run a generator.

You need to cool the building anyway, so the added energy from displacing the beam to a (presumably turbine) generator would essentially be 'free.'

This is more efficient than the photovoltaic because you don't have to produce install and maintain a photovoltaic (all of which costs some energy), and you were going to cool this building with this method anyway (so there's little downside to harvesting the energy aside from complexity, which is a concern).

I assume you'd need line of sight from the panel to the generator, which it does seem would be tricky, but doesn't seem insurmountable. (Put the collector on a tower for example, in most areas of the country I think it wouldn't have to be higher than a few hundred feet). It just depends on how much heat you're displacing and whether you can collect enough to run the generator.

If it were "I'm just trying to produce energy", I'd agree that the photovoltaic is better, but if this system happened to be dual use and doing two things we already want to do, it seems like it would be a pretty substantial win.

My understanding of this is that the amount of heat transmission depends on the temperature of your target. You point it at outer space (cold), and you get a win. You point it at some sort of IR collector (necessarily at ambient temperature, or higher), and you've lost your cooling ability.

Right. One must remember that the target sink is going to be throwing radiation right back at the source emitter. Theoretically, they will exchange energy until they reach equilibrium between emission and absorption. Luckily for us, it will take a long time to heat space up enough to make a difference...

This also ties into why a magnifying glass cannot make something hotter than its source of light.


"Lenses and mirrors work for free; they don't take any energy to operate. If you could use lenses and mirrors to make heat flow from the Sun to a spot on the ground that's hotter than the Sun, you'd be making heat flow from a colder place to a hotter place without expending energy. The second law of thermodynamics says you can't do that. If you could, you could make a perpetual motion machine."

Thanks for that, it really helped me understand the problem with my thinking.

So if you point it at a satellite, or the ISS, and run a generator? Would that be a good backup power source for solar currently used?

That's a great question. The thought of an orbiting IR collector crossed my mind while typing that comment, but I didn't want to think about it hard enough to address that. :)

The "beam" this generates is really more of a floodlight than a spotlight, so only a tiny part of the energy would hit your satellite. If you could focus it down to a narrow beam which hit only your satellite, my gut feeling is that you would simultaneously be focusing the surface heat of the satellite onto your panel. So once again, your panel would only "see" a hot surface, and would lose the benefit of cold space.

Also, as others have calculated in other comments, the amount of energy coming from the sun dwarfs what these panels radiate, so you are better off just pointing your satellite collectors at the sun.

> Also, as others have calculated in other comments, the amount of energy coming from the sun dwarfs what these panels radiate, so you are better off just pointing your satellite collectors at the sun.

Especially in space! We lose a good amount of irradiation energy from the sun to the atmosphere.

The ISS has no shortage of warmth. Their challenge is keeping cool. Adding IR heat will only make things worse and require the use of more (solar) energy to remove that heat again through its cooling panels, which themselves work like this invention by radiating IR heat into space!

Low temperature thermal energy is really worthless unless it's in somebody's home in winter or another case where you need to keep things slightly warm in a cold environment. You can't store it or transport it very far at all. It's the waste product of just about every kind of machine. Look at the steam pouring out of this cooling tower -


Even a power station doesn't bother trying to use that for anything.

A common use is space heating in buildings in winter, but you won't be running an air conditioner in that case, nor will your neighbors want it.

You're not wrong, but it's also important to remember that they're running a heat engine to extract work. The amount of work that can be extracted is dependent on the temperature differential between the heat source and the heat sink. The water being cooled there is to keep the heat sink cold. And since the heat sink will always be colder than the heat source, they can't recycle any of that heat back to the heat source without expending energy to move it. Which would kind of ruin the point.

It's not that the heat isn't useful, or even that it's not a large amount of heat. It's that the heat is in the wrong place, and they need to efficiently get rid of it.

Maybe we could use it as a distributed way of beaming power to satellites or space probes. Have them all be capable of tilting and network them all together.

If you point it at something, and it heats up, it will radiate back heat up the thing you are trying to cool. The whole point of pointing it to space is that space is cold (in the radiation sense).

To me, it seems like a nice way to help alleviate climate change - though my guess would be that it's not enough energy to make a big difference.

If, at minimum, every air conditioner in the world becomes 21% more efficient, then that's a bit of a dent in climate change.

I guess I was mostly thinking about the effect of reusing the IR beam. It gets a little confusing to me in that AC units move more energy than they expend. So for example a 10 EER means an AC unit moves 10x more energy than it takes to run. So if a 10 kW AC unit becomes 8 kW to run (~20% more efficient), a the same time the panel would beam out ~5% of 100 kW of moved energy? Or 5kw. If you reuse 5kW somewhere, then you lose ~70% of the potential climate change heat savings? 5/(5+2)?

I'd have to recheck al the assumptions here because I put high odds that I constructed this back of the envelope set of numbers with some mistake in it...

It also means that air conditioning may become accessible to some people for whom it wasn't before, e.g. because they rely on solar power, or because electricity is expensive.

You can heat things up to the cooler temperature.

Wow, cool. Well, thanks for doing this sort of research! Better cooling could save us a whole lot of energy :)

I assume this is negatively affected by cloud cover, but I'm curious to what degree.

I was curious too, because the article was very specific about the wavelengths emitted. I found some sources saying that the 8-13 um wavelengths are in a band that isn't absorbed much by water vapor, e.g. https://wattsupwiththat.com/2008/06/18/a-window-on-water-vap....

I find hard to believe that without your special cover, the heat would travel 10-100km up to atmosphere and then reflect right back onto the radiator.

What I am missing?

You're missing the bits of atmosphere right under (and inside) your nose.

It's also not like other wavelengths are reflected right back, but when the atmosphere can absorb them, it will also emit them. (Radiation tends to be symmetric like that.) Conversely, the emitting panel will absorb this radiation, until equilibrium is reached.

That means that you need to use wavelengths where the air isn't literally glowing, if you want radiation cooling to work at all.

That's cool. But presumably any surface insulated in this way will reach equilibrium with the atmosphere/space in a similar way, correct?

Wondering if the radiation tech could work in sending the heat generated from other sources like clothes dryers etc as well?

Is this related to how ice can form in the desert at night?

this is such amazing work! how expensive are these radiators before and after scaling production? ignoring the pumps, how long should they last without repair?

The great thing is that we can use existing manufacturing processes (which we are already using) to do this extremely cheaply at large volumes. So the cost at scale will be competitive and make economic sense for customers. Our focus is on packaging this into a compelling product that delivers large electricity savings. Lifetimes of 20+ years are expected.

Thanks for jumping in! How many installed square feet of real estate do I need to be equivalent to my 3 ton air conditioner (which needs 4 square feet)?

So are those normal thin-film deposition processes? I don't know much about those, but am interested in them - are any of the materials particularly tricky to work with?

Or can you just sputter the oxides one-by-one onto a silver substrate?

any chance yall are hiring? :)

I wonder if anybody with serious resources tried to use anti-Stokes scattering for refrigeration.

Sky cooling is a super neat phenomena. It's one of the main reasons why sometimes at night, your car can frost even though the ambient temperature of the air is higher than freezing.

Sorry to be a pedant, but the singular form is phenomenon. Phenomena is the plural.

Wouldn't it be more efficient to just use water run through a coil 6 ft underground? In most areas that have the intense heat that running water over the radiators actually would help, the ground 6 ft below is pretty easily cooler than 5-10 degrees C. The electricity to pump 6 ft down couldn't be that much more either...

And when it isn't as hot out, you can actually use the ground temp to radiate heat out of the house by itself, no need for coils and a fan to run in ambient air temp, right?

This seems like a good time to raise that pumping heat into the surrounding soil doesn't work forever.

For a home aircon it might be OK, however the London Tube is having serious problems with this where many of the tube tunnels have warmed from around 14C to nearly 30C (57-86F) over the last 100 years. Leaving many of them quite hot. The surrounding ground around the tubes has warmed up and effectively heat soaked.

Heat is generated from a number of sources including diesel engines however apparently a large portion of the heat is generated by braking for each station. They've improved that by using electric regenerative braking now which in some cases can dump the power into other trains that are accelerating at the same time.

Some of the newer trains now also have air conditioning inside the trains, while that helps inside the train it doesn't help outside the train where even more heat is dumped.

Anyway for more information there is a good youtube video I can't find right now, so you could read this instead: https://www.forbes.com/sites/lauriewinkless/2017/06/22/sweat... https://www.ianvisits.co.uk/blog/2017/06/10/cooling-the-tube... https://en.wikipedia.org/wiki/London_Underground_cooling

Better solution: stop creating heat in the home. Instead of using gas or electricity to heat water, use a heat pump joined to the A/C. Store greywater and heat that up, then send it down the sewers. A liter of water has the same heat capacity as 3 cubic meters of air and a person uses 300+ liters of water per day, enough to cool 3200 square feet of house by 10 F.

I wish there were viable commercial solutions for this.

The other day, when it was 106F outside and I had to dry some clothes, I found it quite unfortunate that I was spending a bunch of energy to move hot air out of my house, while at the same time spending a bunch of energy creating hot air in my house to dry my clothes.

It would have been so nice if there were a reasonable way for me to take the heat already in my house and push it through the dryer.

(I could have hung them to dry, but I live too close to the freeway to do that -- they would have been dirtier than when I washed them)

Your dryer should be in unconditioned space, at least during the summer. That's easy if it's in its own room with a window. Summer: window open, door closed. Winter: window closed, door open, vent inside.

venting inside creates significant condensation and lint/dust issues. Not a great idea (and I'm pretty sure against relevant building codes)

There is such a thing as ventless clothes dryers, from major brands such as Whirlpool, LG, etc. [1] We had an older one in an apartment we rented; it had a water reservoir that collected the moisture which you had to empty into the sink regularly.

[1] https://duckduckgo.com/?q=ventless+clothes+dryer

A large number of clothes dryers in Europe (save for the UK, which still uses a lot of hotboxes) are condensation dryers. Some countries (Sweden I believe) actually don't allow vented hotbox dryers for new construction due to building codes explicitly barring them. Heat pump dryers have been around for about 3-4 years now, but condensation dryers have been around for decades.

In the US, but have been running Bosch condensation dryers for over a decade. Previous rental had one and the immediate difference I noticed was how much longer my clothes lasted from the cooler temps, so when I moved I bought my own W/D. No regrets.

And more recently, heat pump dryers, which can dry basically as fast as standard electric dryers, while not requiring a vent, and using much less energy. You get double benefit form the heat pump, as the hot side heats the air in the dryer more efficiently than a (resistance-based) electric heater, and the cold side chills the moist air, extracting the water to be drained.

Anyone running a heater probably needs or has a humidifier.

Of course, only works for electric!

You might need a dehumidifier too, given that the intention is to dry clothes.

Looks like heat pump clothes dryers (and hot water heaters) do exist


Also worth noting that dryers usually vent outdoors so they don't dump all the heat inside, where you'd then have to use the AC to remove it.

s you say, this seems very wasteful. Now, I don't live in a place where A/C is needed (I have lived in the tropics several times in my life but never had it), but from what I recall it causes the air to be extremely dry! That would mean you can just hang your clothes up and the water would evaporate out of them pretty quickly, no?

"Really dry" is very relative. The HVAC in my house in FL barely keeps my house in the 50% humidity range in the middle of a muggy summer. My clothes won't dry super fast at that humidity level.

People each put 300+ liters of greywater through the sewer system? Seems a tad bit high.

I took that figure from the USGS per capita household water use[1], which was 300-380 liters per day. Watering plants may make up as much as 20% of that but the biggest culprits are toilets and showers. This report[2] directly measured 1000 households and found daily direct water consumptions from 456 liters to 1815 liters(jesus).

[1]: https://water.usgs.gov/edu/qa-home-percapita.html

[2]: http://www.awwa.org/portals/0/files/resources/water%20knowle...

US puts water consumption per capita at 2,842 cubic meters per annum.

(2,842 cubic meters in liters)/365 = 7.7k liters.

I'm guessing that number is for all uses. Obviously, as being discussed, industrial and agricultural water uses would not be eligible.

The USGS [0] estimates 23,800 mega-gallons per day for domestic use by 268 million people. This is about 89 gallons or 336 liters per day per person.

[0] https://water.usgs.gov/watuse/wudo.html

That's all water use, not household or office use that could be used for this purpose.

For example:

Meat consumption accounts for 30 percent of the American figure, and sugar consumption is responsible for another 15 percent


For a while in the 70's, inspired by a book I'd read (maybe Nomadic Furniture, by Papanek and Hennessey) I took my showers with the drain closed. I would paddle around in the ankle deep warm water and leave it in the tub until cool, thus warming my chilly bathroom and small apartment for "free".

Your solution simply moves the heat into the wastewater system. What's the impact on sewers if we start filling them with much warmer water?

You can actually recover heat from sewers which can heat water or space in nearby buildings. The first example of this (in North America at least) went online in Vancouver around 2010[1].

> Similar to a geo-exchange (or geothermal) system, heat pumps transfer thermal energy from the warm sewage supply to a higher temperature range that’s effective for residential space heating and domestic hot water. While similar in concept, sewage heat recovery is more efficient and cost effective than typical geo-exchange systems. The heat source (raw sewage) has a higher temperature than ground-source heat (on average 18 degrees Celsius as opposed to 8), so it requires less energy to upgrade. Secondly, accessing the sewer line is less invasive and less capital cost intensive than drilling into the earth to access geothermal heat. It also utilizes a continuous supply of waste heat, closing the loop on a fundamental energy >> waste >> energy stream. [2]

[1] http://greenbuildingaudiotour.com/buildings/sefc_neighbourho...

[2] http://www.thechallengeseries.ca/chapter-05/neighbourhood-en...

It might reduce the buildup of sediment inside pipes. It also might increase pipe erosion. It would probably have a comparable ground-warming effect to geothermal sinking. I have no idea what effect it would have on wastewater treatment plants, but it would cause significant warming local to sewage outflows similar to how water-cooled power plants can disturb river ecosystems. It would probably make sewage maintenance much more difficult since it would be so hot and humid. It might drive out animals living in the sewers, or just make it impossible for them to survive. I have no idea what that would mean. Mainly I was thinking that it would be a great deal easier to simply deal with the heat at the wastewater plant, but there would also be an immense heat sinking effect because of all the distance traveled inside sewers.

I was only thinking of it in context of using the ground to sink heat, not as a real implementable plan. I think it goes without saying that the heat-pump-as-a-water-heater idea is far more pragmatic.

This is a common and easily available technology for residential and commercial applications. With a closed-loop system a 4-ton HVAC system requires roughly 600ft of 3/4-inch pipe as vertical wells. That typically means 4 wells maybe 15-20 feet apart. You need a decent about of property to support these systems, enough room for a well driller, appropriate drilling conditions, permits for drilling, etc.

There are other systems for people with the land which may be less expensive which don't require drilling.

I spent several summers installing these systems, they are surprisingly simple and effective.


The problem is that with ACs/power plants, the net temperature will rise if you do this because it's a closed system.

Power plants and ACs are not 100% efficient; the power plant converts chemical/nuclear/mechanical energy into electrical energy, releasing heat from work lost.

Your AC will convert electrical energy to cool the air around you, but because it is not 100% efficient, the total temperature of the system will rise.

There needs to be a way to get rid of that lost energy and remove it from the system. If you put that into the ground, it's still in the system. Ofc, even when they try to 'beam the heat to space', much of it will end up staying in the system (earth) as it reflects on clouds / water vapor, etc.

I forget the exact number, but the clear night sky is equivalent to a very cold black body radiator, around -30C or so (can't recall the actual value and it probably depends on local temperature too). Radiative equilibrium with it will push your temperature way down.

This is the main reason why it suddenly gets so cold at night. From a radiative perspective, you might as well be facing a ceiling of Arctic ice above you.

I'm an amateur astronomer. When I do observations at night, I always add an extra layer of clothing. It's amazing just how really cold it gets when all you do is sit under the clear night sky.

When it's cloudy it's different. The chill effect due to the (lack of) radiation is much less intense.

Isn’t this basically what happened with the London underground? It used to be cold down there now it’s super hot as the soil can’t absorb more heat.

Source: https://www.forbes.com/forbes/welcome/?toURL=https://www.for...

> According to this article in Rail Magazine, the average temperature inside the tunnels was around 14° in 1900.

> increased the temperature of the surrounding clay so much that, today, average tunnel temperatures are between 20-25°C.

That sounds like a good way to bank energy for later use during London winters, if you ask me.

It takes the right kind of soil/moisture for good thermal contact, and it can take a very large (length and/or depth) group loop for something like an office building.

There are limits to the technology. Skyscrapers are simply too large for any reasonable number of wells to provide adequate heating and cooling, Mr. Orio said. And it would be very difficult to install a well beneath an existing structure.


Bosch makes geothermal heating and cooling units.


These geothermal solutions are quite pricy:



a typical home of 2500 square feet, with a heating load of 60,000 BTU and a cooling load of 60,000 BTU will cost between $20,000 to $25,000 to install.


The estimate I got for a well alone was about $40,000.

Obviously, it's a renewable energy source. But this is also a premium German product. It's designed to be quiet. So you don't have a loud AC roaring in your yard all summer. That allows builders to build houses closer together in premium developments such as serenbe GA.

Maybe digging is impractical in a lot of situations?

No. The ground around the pipes will heat up very quickly.

I am super excited they built a practical demonstration system out of this technology. I would love it if more 'breakthrough' research took it to this sort of proof of concept (I know it is hard to get funding for that sort of project but it is so useful).

I would love to see self contained shade structures that would both shade and lightly cool the air underneath them.

Question - could one theoretically build a big enough array of this to counter-act globally warming? Any back of the envelope math for what it would take would be very interesting.

Radiative forcing from global warming will be somewhere between 2.6 and 6.0 Watts/m^2 by the end of the century[1]. Currently it is around 1.75 W/m^2[2], which translates to a total increase in absorption of 223 terawatts. That's about 3x the total energy produced by all photosynthesis[3][citation needed], and over 12x the energy in all forms consumed by humanity.

So, to a first estimate, no. It's absolutely beyond what we're capable of. To a second estimate: these panels reject 70 W/m^2, so to counteract current warming we would need to cover .625% of the earth's surface in them, which is 3.2 million square kilometers, which is over 6x the area you'd need to replace all of humanity's energy use with solar power.

All in all, it doesn't make much sense.

[1]: https://en.wikipedia.org/wiki/Representative_Concentration_P...

[2]: https://en.wikipedia.org/wiki/Radiative_forcing#/media/File:...

[3]: https://en.wikipedia.org/wiki/Orders_of_magnitude_(power)#Te...

It seems like solar panels work best when in sunlight, and infrared radiators work best when shaded.

If you have a square 1 m^2 solar panel, and a square 1 m^2 area to put it in, and you live at 35 degrees north latitude, you angle your panel up 35 degrees facing south. That gives you .428 m^2 of area, angled at 55 degrees, facing north, which will always be in the shadow of the panel. So stick a rectangular infrared radiator behind the solar panel.

So the area needed would be less than that needed to replace all of humanity's energy use with solar power, but the exact amount less depends on the latitude at which you place the solar panels.

Agreed on geoengineering. But given that AC is a non-trivial fraction of current energy use [1] and that there are ~2B low-energy consumers in the tropics who would enormously benefit from AC, the in-principle impact on GHG output here is substantial. Install to cool the earth: no. Install to cool human-occupied spaces: yes.

[1] https://energy.gov/energysaver/air-conditioning

Cool. I thought it wouldn't be feasible, but those numbers really help cement that idea.

Thank you.

And thank you for asking :). I had the same question and the answer is fantastic.

I really wish HN had some mechanism of extra upvotes, e.g. by being allowed to hand out one extra upvote per day or week, or by paying 5 cents (eur/usd) to buy an extra upvote (with a maximum of X per post or Y per day, or both, or whatever)... but alas.

A possible mechanism for this is the "favorite" link that you can get to by clicking on the timestamp of a comment. Currently, I don't think much is being done with it other than adding it to a private per-user list, but it could also be used to highlight or boost the chosen comments?

A quick fix might be to just add another item to the "Lists" link at the bottom that shows the top "Favorited" comments, based on how many people have it as their most recent (or 10 most recent) favorite.

these panels reject 70 W/m^2

What about plain mirrors? How efficiently do they reflect photons back into space? Presumably they're highly efficient for visible light; how easy is it to reflect infrared passively?

The problem with using mirrors as radiators is that they're terrible radiators. They might be great at reflecting the sky-photons away, but they won't launch their own photons away either.

Shiny things are great at reflecting incident radiation (which is why they're good mirrors) but bad at radiating their own heat. Dull black things are bad at reflecting but good at radiating. The technical terms are reflectance, absorptance, and emissivity.


A mirror rejects close to 0 heat compared to a matte surface, it just causes it to be absorbed in the air instead of the ground.

I think the biggest question here is what saves more dirty energy. Obviously, these units will compete with solar for roof space. The sun beats down in the summer, where these panels would be most useful, but also where solar would hit peak power generation. As long as solar can supply at least 20% of the power required to cool the building, it seems like it will be a better option, as solar is also useful for other things like charging electric cars during the winter, when A/C is not required.

Even on a flat roof, there is room to add radiator panels without compromising solar space. The solar panels will be tipped up to the south some amount based on their latitude, but you don't want to shade the low end of a panel with the high end of the next one, so there is a gap between rows. A shaded gap. Precisely the best spot to put radiator panels.

You could even imagine a combined product which would be a corrugated panel alternating "uphill" solar and "downhill" radiator segments with the dimensions tailored to the latitude. This could be installed flat on the roof.

180 degree flip panels. Solar cells during the day, IR radiators at night.

Plus it would look super cool if you got them to flip in a wave pattern across your roof.

I can see it now: malware communicating across air-gapped systems by way of flipping radiator panels on roofs.

After the CRISPR malware transmitted via actual genes, I'd barely be surprised.

That was a lot of hype and not much else. They modified the machines that were reading the specially crafted DNA.

Life always finds a way.

Yes this has to be it

Tesla's SolarCity product consists of rooftop tiles, some of which are not solar panels -- approx. 50%. Those non-collecting panels could be replaced with IR emitting panels without much change to their overall strategy.

This assumes it is most efficient to build the solar locally. These radiative panels have to be installed locally. Solar can be transmitted over wires from large scale plants.

Simply putting your outside AC Condenser Unit on the shady side of a house or have a Tree or Bush shade it can significantly reduce your energy usage. They units themselves are at least 20 Degrees cooler than being in direct sunlight.

There are also misting units that use approx 1-2 gallons of water a day (equivalent to a toilet flush) that can also reduce energy consumption by up to 30%.

Simple solutions for big savings.

Your first claim was so interesting I had to Google it. But this article suggests it doesn't hold up.


Moneyquote: "The temperature of the surrounding air has a much bigger effect on cooling efficiency than direct solar gain, and the volume of air pulled in by an air conditioner is huge."

This article you are referring to is for a heat pump and so are the % of savings.

An AC Condenser sitting in the Sunlight will naturally be a lot hotter in temperature than one sitting in the shade. For example park your car in the sunlight and then park your car in the shade? Where will the surface of your vehicle be hotter/cooler? It has nothing to do with air temperature at this point. It is the heat of the sunlight that is raising the temp of the coils in the condenser just like it would raise the temp of the surface of a vehicle parked in the sun instead of the shade.

I live in TX. I have some experience with this.

A typical small AC will have to dissipate about 5kW of heat through the radiator. If you put it into the sun, it will additionally need to dissipate the energy it gets from sunlight. Assuming that the area exposed to the sun of the radiator is about 0.25 square meters, it will receive at most 250W of energy from the sun (assuming that it is painted black and perfectly absorbs all light and miraculously doesn't emit anything).

This means that you can save at most 5% by putting the radiator in the shade.

(Metal sheets heat up quickly in the sun because they have low heat capacity. They also cool down quickly if you blow on them with a fan. Just because something feels "hot", it doesn't mean that a lot of energy is stored inside)

It's not just the unit itself, but everything the sun shining on nearby heating the air which is getting blown through the coils. The temperature increase effect is also superlinear; the efficiency goes down as the temperature increases so 250W extra heating might cause an extra 120W of loss (that number is made up, but the amount is greater than zero).

Swamp coolers are useless in high humidity climates. And similarly, shade isn't as cool relative to direct sunlight in high humidity either.

The water removes more heat from the coolant; the swamp cooler isn't for the inside of the house it's for the compressor itself.

I disagree. There isn't much difference in the _air temperature_ of air "in the shade" versus air not in the shade.

Air temperature isn't the only thing that heats a condenser. Solar radiation does that too.

It’s a difference of over 10 degrees in many places.

For a better scaling version of this technology (roll-to-roll manufacture), see: http://science.sciencemag.org/content/355/6329/1062

Atmosphere is only transparent for thermal infrared without clouds. Clouds emits and reflects ambient heat back to these panels heating them. But then one does not need typically air-conditioning on a cloudy day :)

Depends on where you live. :) Where I am, clouds do nothing.

Interesting! Do you know some examples of places where clouds influence temperature a lot and where it doesn't? And what causes this?

Maybe it's typical cloud base height?

I'd be curious if we could build a floating patch of these in the ocean and put them in front of hurricanes to rob them of the hot water they need to grow?

Anyone able to do the math? Keep in mind all the billions we spend on hurricane damage.

It would be a lot cheaper to do some cloud seeding or SRM ahead of the storm to reflect away solar radiation into space.


Professor Fan has a few incredible, highly technical lectures about the subject on YouTube if anyone's interested.[0] His work centers on nanophotonics, which goes hand in hand with research surrounding photovoltaic cell efficiency as well.

https://www.youtube.com/watch?v=Uazp_k_ckqc [0]

Howis this "20% energy savings" product different from their original "all passive" pitch[1] from 2014?


Couldn't beaming heat into space also buffer against global warming, as long as the amount of heat transmitted > the energy spent to beam it? Especially if we use renewable energy.

I think that's kind of just applying a bandaid to a wound that already has an infection. If you don't treat the underlying cause, (greenhouse gases) it's still going to get worse. Still, a massive planetary array of space radiators sounds badass.

At the risk over over-utilizing your metaphore... you still put on bandaids on infections -- even if you're taking antibiotics or applying antiseptic ointment.

Pretty neat idea.

This reminds me of sun free photovoltaics. Somehow people managed to transform heat into coherent IR waves that could be converted by the usual PV cells.

Yes this. For a reason I remember an article "sun free photovoltaics" hence my comment. But I can't find it again on this page.. odd. https://en.wikipedia.org/wiki/Sun-free_photovoltaics

Is there any material that readily absorbs heat at wavelengths between 8 and 13 micrometers? Instead of beaming the heat to space, could you instead put that special coating on a parabolic reflector and use it to heat up an absorbing material at the focal point? And then put a frying pan on top of that and cook an egg?

There's a statement of the 2nd law of thermodynamics, that forbids you from passively transferring heat from a cold body to a hot body.

Ah of course. So best case it can accelerate warming cold things to ambient temperature

It's completely unnecessary. In a system capable of doing that, you might just as well boil your eggs at the source. Thermal equilibrium, and all.

How much does this save vs. replacing these panels with solar panels?

Does that mean we can cover the deserts with these plates that run purely on solar energy and release energy as infrared radiation to combat Global Heating?

Ah, sorry, duplicate question (see below). Well, even if it does not do much, it's a start, and it shouldn't require putting in extra energy from power plants.

If this system is scalable area-wise wouldn't it be a good idea to shield the ice caps with this? A 3C to 5C degree difference could mean quite a bit in preventing further ice melting, no?

The Arctic and Antarctic are very big. If you could somehow build a structure that would withstand the harsh conditions, I imagine using a mirror or even aluminum foil would be an overwhelmingly cheaper way of reflecting energy back out.

I believe I saw a test where they put a white sheet over some ice in Greenland, and came back a year later. The ice under the sheet was multiple inches higher than the surrounding ice.

But that's just it -- it is NOT scaleable to any measure that would make a difference. The manufacturing effort alone would require quantities of raw material that would dwarf all construction efforts ever undertaken combined. There is the issue of mining, refining, manufacturing, transporation... not to mention the installation and maintenance of a hundred thousand square miles of reflective shields in an incredibly hostile (to life) environment. Even if the whole of humanity could agree this is the most pressing issue of civilization, we would wreck the earth to save the ice. It would be easier to build a shopping mall at the bottom of the ocean.

And the sun will just get warmer anyway. It is folly to think we can play God. Climate change is normal, and natural. It has been happening since the beginning of time, and will continue long after all traces of life on Earth are gone. Who are we to attempt to interfere?

We already did play God by increasing CO2 in atmosphere. There is nothing wrong in efforts to mitigate its effect.

If the efforts are effective, perhaps not. If. But just doing something/anything isn't the same as solving the problem. Building an umbrella over the ice caps is 1) not technologically possible and 2) the secondary effects of the attempt would be extreme, making the original problem much worse if one accepts the premise that human activity releasing CO2 causes global warming.

Again, imagine the amount of fossil fuel needed for mining, manufacturing, transportation, of a reflective panel that can cover 10% of the earth's surface. If indeed CO2 is at the root of this problem, ask yourself if the shade given by this sun shield would offset the manufacturing costs in terms of CO2 release.

It would be more effective to simply plant more trees, or protect the Amazon from logging or somehow convince China to stop building coal-fired power plants, or the whole world to convert to nuclear power. etc.

Or just build a ladder to the moon. This is silly.

> or somehow convince China to stop building coal-fired power plants

They are already convinced: https://www.csmonitor.com/World/Asia-Pacific/2017/0303/China...


I don't think so. There's not a lot of evidence "they" (meaning the CPC) are convinced of global consequence to stop. The desire for growth outweighs any actual concern for the future of the planet from their perspective. Without direct evidence of outcome, a tyrannical social structure operates from the top, expecting to deal with the future by sacrificing the bottom. The Aztecs, the Pharaohs, the Romans, USSR etc all fell when they external sources overwhelmed the existing mitigation methods (insidiously or violently) and China will probably travel the same route. One might hope it's not the final lasting dynasty 100 years from now.

If these become popular, maybe you could create a new solar panel that absorbs the wavelengths emitted by these air conditioners in order to create electricity

...or better yet...

These panels reject heat at a rate of up to 70 W/m^2. A good solar panel absorbs 20% of incoming light, which typically peaks at about 1000 W, so a solar panel reduces temperature by 3x as much.

So... you could just do that.

Furthermore, air conditioning isn't lowering the temperature of a system by the energy consumed. Rather, it acts like a heat pump which usually pumps more energy that is used. So 200 J/s of electric power running a typical AC unit would likely reject more than 200 J/s from a building.

Solar panel efficiency decreases with rising temperatures, so maybe increasing its emissivity in the infrared window could improve efficiency.

Could you create a solar panel that generates energy during the day and rejects heat during the night? Would it be efficient at either task?

I thought such a system already existed.

A temperature drop of 3-5 degrees C is meaningless unless other variables are supplied. For example, how much water was cooled? They tell you how long (3 days) but they don't let you how many panels they used either. Too many 'science' articles these days don't add up.

This is just a press release about the actual scientific publication. The abstract provides more information:


Cooling systems consume 15% of electricity generated globally and account for 10% of global greenhouse gas emissions. With demand for cooling expected to grow tenfold by 2050, improving the efficiency of cooling systems is a critical part of the twenty-first-century energy challenge. Building upon recent demonstrations of daytime radiative sky cooling, here we demonstrate fluid cooling panels that harness radiative sky cooling to cool fluids below the air temperature with zero evaporative losses, and use almost no electricity. Over three days of testing, we show that the panels cool water up to 5 ∘C below the ambient air temperature at water flow rates of 0.2 l min−1 m−2, corresponding to an effective heat rejection flux of up to 70 W m−2. We further show through modelling that, when integrated on the condenser side of the cooling system of a two-storey office building in a hot dry climate (Las Vegas, USA), electricity consumption for cooling during the summer could be reduced by 21% (14.3 MWh).

And you can read the full article via sci-hub to get all the experimental details.

The paper's doi:10.1038/nenergy.2017.143

You don't just get electrical savings. Unlike other AC heat transfer systems, the heat is not pumped into the local surroundings, but is ejected from the Earth.

> the panels cool water up to 5 ∘C below the ambient air temperature at water flow rates of 0.2 l min−1 m−2, corresponding to an effective heat rejection flux of up to 70 W m−2.

Can someone translate that into layman terms? The chained negative exponents are somewhat confusing. I would think 70 W m^2 would me square meters, but what's 70 W m^-2? I assume it's not really mappable to dimensions at that point...

Sorry, that's the convention of the journal. It's W/m^2 -- heat rejection per unit area.

Oh, I get it now. I totally missed that there wasn't a division there. That makes complete sense now that it was pointed out, I'm just not used to seeing stuff in that format apparently. :)

Negative exponents means its divided by. For instance s^-1 would be equal to 1/s or more formally known as Hertz. You see this every once in a while in scientific papers. This notation is convenient because it makes it easier to simplify the units.

Yeah, it's basic algebra. I just wasn't seeing it for some reason, possibly because I only think about converting exponents like that when trying to solve something (usually when helping my daughter with her Algebra homework). Yet another example of context affecting cognition, I guess.

Meh. Get rid of the big cow farms and you have removed the biggest threat to the climate. This solution here is nice, but it's just a small amount compared to the big players.

Agriculture only contributes 10% or so in the US. Hardly the "biggest threat".

  into space
Hmmm, that seems like a click-bait statement. Between the air conditioner and even low earth orbit, there's plenty of atmosphere that will soak up some of the beam's heat, no?

No offense, but did you read the article? It's directly addressed:

> So Raman and electrical engineering professor Shanhui Fan made panels containing layers of silicon dioxide and hafnium oxide on top of a thin layer of silver. These radiate in a unique way: They send heat directly into space, bypassing the Earth’s atmosphere. The panels do this by emitting heat at infrared wavelengths between 8 and 13 micrometers. To these waves, the Earth’s atmosphere is transparent. What’s more, the panels reflect nearly all the sunlight falling on them.

A little further research shows that the atmosphere is opaque at numerous wavelengths and it's probably safe to assume that they intentionally targeted the opaque ones: https://i.imgur.com/eUmTltm.png

agreed, except that in the last sentence you mean "transparent" not "opaque" (the atmosphere must be relatively transparent at a particular wavelength to allow radiant energy to escape to space at that wavelength).

While i understand you both.

A title is a title. And the title claims 'heat' not reflecting infrared wavelengths between 8 and 13 micrometers. And imho 'heat' is more then certain ranges of infrared wavelengths. So in that sense i can see why somebody uses the word clickbait in regards to this title.

edit joke While i'm just afraid the moon will start to condensate /edit

It's removing heat from a system by radiating it into space. That it's radiated using infrared is irrelevant, since it does cool the system it is attached to. Even "beam" is appropriate in this instance. There's no "bait" in the "click-bait", it's exactly as advertised.

Listen, a little tl;dr is not too much to ask for around here, especially when the salient point is buried 7 paragraphs in, when it should be present in the leading sentence. Forgive me for not locating the relevant passage, while quickly skimming this and 100 other articles, while at work.

Furthermore, regarding cloud cover, I am dubious of the premise that even the specified wavelengths penetrate otherwise opaque cloud cover. Sure, maybe that portion of the spectrum penetrates the typical atmospheric mix, including light water vapor, on an otherwise clear day, but, color me unsurprised if it fails to penetrate smog, diesel and particulate pollution, and other common dust constituents in cloud nuclei. I mean, even radar can be blinded by dense storm clouds.

But hey, maybe I'm wrong, maybe it's a 100% miracle cure. Maybe it's that awesome. I hope I really am just a big dummy, too lazy to read a blurb more than three sentences long.

> Listen, a little tl;dr is not too much to ask for around here

Then ask for a summary, or point out that it's not well written for easy dissemination of facts.. There's no need to call a title out as click bait when it isn't. There's plenty of click-bait titles to go around, we don't need to malign accurate titles without cause. A click-bait title isn't just wrong or unclear, it's purpose is to intentionally mislead people for views. An accusation of click-bait is similar to an accusation of lying, which is different than being unclear or incorrect. As such it should be reserved for cases where there's evidence to back it up, otherwise discourse breaks down, in the same way accusing someone of lying when they are unclear (or you are unclear on their meaning) doesn't lead to useful discussion.

You have questions regarding the veracity of the statement? Sure, figure that out, and then make a statement about how accurate it is. If you aren't part of the signal, you're part of the noise.

apparently you don't really know what clickbait is, though i do know what you mean, and i think thats correct, though not more correct then to say the title has a clickbait factor.

And if you think this title wasn't thought of to maximise clickthroughs (again, regardless of the content of the article as clickbait ONLY applies to the title) ... you are very naive..

Other titles could be, getting rid of excess heat via infrared into space. Which would be less interesting to majority of people, as people read it, and think 'oh airconditioning, i got that... oh a new unit that beams into space ... neato...'

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