Using a solar oven as a radiant refrigerator at night | Hacker News Reader

Using a solar oven as a radiant refrigerator at night

solarcooking.org

182 pointsColinWright88 comments

If you insulate it in vacuum with an IR transparent window (unfortunately $$$) you can get substantially cooler than ambient: https://www.nature.com/articles/ncomms13729

Here in Brazil we use a system with copper pipes in an insulated "box" to heat water during the day using this energy from the sun. A couple of years ago, in a cold winter day with a minimum of 5 degrees (celsius), just after the sunrise the water froze, broke the pipes and the glass above it. I couldn't understand what happened because the ambient temperature was above freezing point, maybe it was something like this effect?

Fascinating! That does seem like the most likely explanation.

This reminds me of the ancient ice ponds that made ice thousands of years ago in Persia. I read somewhere that they were able to make ice through a combination of radiative and evaporative cooling at night temperatures around the same as you experienced, about 5C.

Yes, I like this explanation, when that phenomenon happened in our house I thought one of our neighbours was dumb because he just put a blanket over the collector on the roof, as he said, "to keep it warm". But now I think he was right. Now, I don't know if this exists but it would be nice if there was a kind of glass that let radiation pass only to one side and not the other way, a kind of "valve", this could solve the problem of water frozing from radiation in our solar heating.

>Now, I don't know if this exists but it would be nice if there was a kind of glass that let radiation pass only to one side and not the other way, a kind of "valve", this could solve the problem of water frozing from radiation in our solar heating.

This doesn't exist and can't even in principle because it would be a fundamental violation of thermodynamics. Basically, it would be a Maxwell's demon for radiation that would allow you to arbitrarily reduce entropy.

But, it might be possible to have a material that has different characteristics at different temperatures, as long as it's symmetric. If it's nearly opaque to IR in its cold state, hopefully sunlight at dawn would warm it rapidly enough that it would automatically "shut off" on cold nights, and still "turn on" shortly after dawn even on cold days.

Yes I think that'd be ok. You'd basically just be taking advantage of the thermal gradient between the radiating body and the object being heated. It'd be analogous to adding more insulation as it heats up and removing insulation as it cools down.

You'd be slowing the flow of heat into your reservoir just as much as you'd be slowing down the loss of that heat later though.

You might still get some benefit in preventing freezing at the expense of needed a larger area to get the same amount of heat flow in to your water system though.

Light meter controlled / tracking louvers would be the quick fix.

We call this a one-way mirror.

https://en.m.wikipedia.org/wiki/One-way_mirror

They are not perfect (intensity & wavelength limitations), but they exist.

See also mirages and total internal reflection.

I don’t think maxwells demon is impossible, it just is a device for converting información to energy. In theory, many such devices may be possible for the conversion of information to both energy and matter. This seems like it might be support the simulation hypothesis.

Relevant link: https://en.wikipedia.org/wiki/Yakhch%C4%81l

The handwavey explanation in my head is that “ambient” for this emitter is not just the immediate physical environment, but also deep space, which is very very cold. This wouldn’t work if the temperature of space and the air were the same.

An emitter is also an absorber so if space were as hot as the emitter then it would not shed heat.

Yeah, on a clear night the pipes radiate heat to space but nothing much radiates the other way as space is cold and dark. It's also why the tops of cars get frosty on clear nights.

this is a great paper, thanks

a thing to keep in mind with research papers is that in many cases they're concerned with finding how to get the largest effect, or demonstrate the effect in a way that's most clearly due to the reasons they claim and not some experimental error, or measure the effect most precisely, and only in rare cases are they concerned with how to get the effect most cheaply, a consideration which conflicts with the others

there are a lot of cheap materials that are transparent in the thermal infrared, like ldpe, potassium chloride, sodium chloride, silicon, and rock crystal. they're mostly a pain in one way or another (not that i have any experience with this)

Checking my understanding here: in a very simplified sense, is this saying that we’re taking energy, be it from ambient temperature during the evening or that plus solar radiation during the day, and essentially bottlenecking/‘forcing’ it to radiate in the spectrum that’s atmospherically transparent and can thus escape to space? Seems the use of insulating it in a vacuum would be to minimise heat loss until it’s cooled to the respective temperature that radiates in that specific spectrum.

Incredibly cool (hah) paper to read, thank you.

You might like NightHawkInLight's videos on the topic[0] and Tech Ingredients[1] - they've been working on making paint using this technology with an at-home DIY process and experimenting with it.

[0] https://www.youtube.com/watch?v=KDRnEm-B3AI&t=285s

[1] https://www.youtube.com/watch?v=dNs_kNilSjk

Tech Ingredients is amazing

You should take an infrared thermometer and point it to the clear sky. If the sky is very clear, you will get temperatures much lower than ambient. You just let heat radiate away, and this is enough to cool.

Why isn't this replacing aircon over the whole world? Unless that IR-transparent window is made of gold dust, it has to work out cheaper than the energy cost of running an air conditioner.

If my napkin math is right, solar panels give around 2x higher electricity production per square meter than these devices give cooling power. Then that goes into a COP of ~3 for your aircon, and the solar panels have have a factor of 6x better cooling performance per area covered.

Then factor in that electricity can be used for lots of other useful things than cooling, and that solar power variation during the day is perfectly matched with cooling demand, it's a no brainer.

Adding to everyone else, but local climate makes a big difference. This isn't very useful unless you have dry still air, and it works best at night, but could maybe work during the day if you have a good view of the sky but have blocked the sun.

Most places with dry still air don't need specific cooling at night, because "everything" already cools due to this effect, and it doesn't stay hot for long after the sun goes down. Areas that need a lot of cooling overnight tend to be humid, which disrupts radiative cooling. Also, it doesn't take much airflow for convection to transfer more heat than radiation, and most places have variable wind... so you can't really count on it.

> Most places with dry still air don't need specific cooling at night, because "everything" already cools due to this effect, and it doesn't stay hot for long after the sun goes down.

Exactly this. This is reflected as well in the article, even if not promintently: they have 30C Celsius at midday (which is hot) and 4-5C at night (which is cold). 25C between day and night is close to desert behavior as far as diurnal air temperature variation [1] is concerned.

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

my air conditioner is rejecting 3000 watts to the outdoors right now, even though my bedroom is the same temperature as the nighttime outdoors. doing that with radiative cooling would require 8 square meters of emitter area at 400 watts per square meter and a clear, low-humidity sky (not necessarily night, but in the daytime, you additionally have to keep the sun off the panels). my air conditioner's condenser is significantly less than 1 square meter as seen from the sky and also works when it's cloudy or humid outside

basically forced-air convection heat transfer is a motherfucking miracle, and vapor-compression refrigeration even more so

I am disappointed you could not come up with a more concrete term for an improvement over a 'motherfucking miracle'. :)

But, yes, heat pumps are magic shite.

magic shite!

Theres videos on youtube by a dude that does Science experiments at home making radiant paint from common stuff, it's really good and easy to use and shows solid reductions below ambient.

Simple plastic film stretched over the opening would be a good start, perhaps two layers with some air between. You lose a little in radiation but gain heaps from stopping convection.

Could this be used for power generation at large scale, while cooling the Earth at the same time?

Consider this: quite powerful solar thermal generators [1] and furnaces [2] have been built, which use mirrors to concentrate sunlight. Do they create intense cold at night? No, because you can't focus coldness. The device in the article, and the one described above, depend on shielding a volume from terrestrial heat sources, and are effective only because the heat being transferred is minimized.

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

[2] https://en.wikipedia.org/wiki/Odeillo_solar_furnace

Will it work equally well on nights with cloud cover?

No. It doesn't work the same as clouds are at ~0C+ whereas space is like -270C (~3K).

This can also be used to heat and cool a building. Harold Hay experimented with a "Skytherm roof" back in the 1970s and built a house at Atascadero, CA.

https://www.sciencedirect.com/science/article/abs/pii/S09601...

Harold Hay LA Times article: https://www.latimes.com/archives/la-xpm-2007-nov-10-fi-harol...

This is how buildings heat and cool, mostly - every part of a structure that can see the sky allows radiative heating AND cooling. People forget about the second part.

And this is why we built our current house on a north facing slope, in a forest - against all common wisdom about siting.

Place stays cool in the summer, warm in the winter - the other side of the valley literally caught fire this summer, and gets intense cycles of cooling in the winter, as it gets only ~4h of sunlight and 20h of radiation - our more sheltered side gets 0h of sunlight, but the slope and tree cover drastically reduce radiative cooling at night - it’s insane how much the temperature varies as you walk around at night, and you can feel the katabatic winds flowing down the valley and side valleys.

"~4h of sunlight and 20h of radiation - our more sheltered side gets 0h of sunlight, but the slope and tree cover drastically reduce radiative cooling at night"

I still would always take that 4h of sunlight over 0h and just invest in better insulation. Don't you miss the sun in winter?

But if you don't, well, then it makes sense. North facing sites are usually quite cheaper as well.

We own both sides so the siting was by choice - the place is well insulated but having a more moderate ambient temperature outside, and having the windows shaded from the cold of space make a big difference.

As to the sunlight - while we don’t get any directly into the house for 2.5 months in the winter, the far side of the valley, which is practically barren, ends up brightly lit, bringing a lot of indirect light in.

Finally, for us, the summer is a killer - it can be over 40 for weeks on end, and the nights on the shadier side are much cooler, as we don’t have superheated rocks re-emitting IR, instead the canopy reflects and absorbs most of the solar irradiance. The daytime air will still be sweltering, but it doesn’t transfer heat anything like direct insolation. So the effect is more about the environment around the structure than the structure itself - although regarding the house the roof is exposed to the sky, as it’s heavily insulated and doesn’t matter, but the walls, which have a decent thermal mass being solid lumber, are always shaded, as are the windows, mostly through self-shading due to the orientation. Again, punishingly bright in there in the summer even without direct sunlight - most folks near us just eschew windows, or cover them with heavy shutters and so have no light and no view in the summer.

> And this is why we built our current house on a north facing slope, in a forest - against all common wisdom about siting.

Can you comment or give more detail about the level of tree coverage required to have a noticeable effect of radiative cooling?

Enough that the ground is mostly shaded on a summer’s day. Dappled shade, I suppose. You’re basically thinking about what percentage of the sky a given point on the ground can see, and a surprisingly small amount of cover goes a long way.

> katabatic winds

New word! So that’s what I experienced on a mountain side in Wyoming

Which hemisphere are you on? If you're in the southern hemisphere do you want to have a south-facing slope(aka away from the sun)?

Northern, in southern Europe.

It really depends - typically a slope facing away from the sun in the temperate zone will be more heavily forested - the sunward side can be too extreme in its variation for much beyond scrub to take hold. We chose our siting based on what was best for our environment, and on noticing how the forest didn’t frost in the winter, and kept green and lush on the floor in the summer, when all else was parched. Considerations around daylight etc., followed after choosing a site. The walls facing downhill and up/down the valley are practically all window - the bedrooms, which face the uphill slope, have small windows, which are still large by the standard for housing around us.

Many years ago, I had an indoor swimming pool with solar heating. Some days in the middle of summer that water got too warm. So I'd run the water through the black plastic solar collectors at night-time to throw away that excess heat.

I installed a new pool solar (water) heating system a few years ago. It's got electronic controls for when to send the water up based on temperature etc. It has an integrated cooling mode, which does exactly what you described. Not useful for my location but I've learned in more warm climates it can be quiet a problem.

Better than this https://www.youtube.com/watch?v=lcnr6YiT-g8

Huh, we had an outdoor pool with no heating other than just being outside in the sun, and it would regularly get too warm. We would only swim after about 11pm.

We also thought it was too cold to swim if the air temp was below 35C (95F).

Australia.

I’m missing the basic premise here, how does pointing an insulated box at the sky at night cool whatever’s in the box?

A black body will emit radiation (and receive radiation) until it's in equilibrium with its surroundings. In theory, the black bit of box radiates heat into space, and receives some of the CMB until both are at ~3K. It's how we cool satellites and spacecraft.

But the effect is quite small, and I'm suspicious of it (as the nearest surroundings are the hot planet, which transfer heat more effectively!)

If you grow crops, your crops can freeze when it's a clear sky at night with stagnant air even though the air temperature only goes close to freezing.

In some agri areas, you'll see "wind machines" to blow the warmer air around the crops to keep them from radiating their heat and frosting.

https://www.ontario.ca/page/wind-machines-minimizing-cold-in...

Wow unradiant heat, or more accurately radiant cooling I suppose. Fascinating

There are some IR bands that are transparent between the ground and space so if you optimize your emissions in those bands (because you are not an idealized black body) and insulate yourself from conduction and convection heating, you can beat your local surroundings and cool yourself lower than ambient to a useful degree.

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

There are materials and even paints that have emission spectrums designed for this.

For example: https://www.parc.com/technologies/self-cooling-paint/

Normally you would just be in the same thermal equilibrium with everything else having similar emission spectrums you'd give and receive similar amounts of radiative heating and cooling. Tweak your spectrum and you tip the balance in your favor.

it's easy to calculate how big the effect is with units(1)

    You have: stefanboltzmann tempC(20)**4
    You want: W/m**2
     * 418.76592
     / 0.0023879689
    You have: stefanboltzmann tempC(0)**4
    You want: W/m**2
     * 315.65782
     / 0.0031679874

so in the temperature range of interest you're radiating 300–400 watts/m², derated for your surface emissivity at the relevant infrared wavelengths. as i understand it, the temperature will drop until the heat emitted through radiation is counterbalanced through heat that seeps in through the insulation and through the stagnant, stratified air above your pan of water, and, probably more importantly, through greenhouse-effect radiation from the warm atmosphere and any opaque objects you have unfortunately included in your field of view

tens or hundreds of watts of cooling is quite significant indeed when we're talking about making some ice or cooling some food

> as the nearest surroundings are the hot planet, which transfer heat more effectively!

The box is insulated. In one case it's an unused refrigerator that opens on the top with more insulation apparently surrounding it. The planet may be hot but on a cold night the ground does nothing to warm me. Soil is not a good conductor of that heat at the surface nor is the air between it and the container. The planet stays "hot" because it is "insulated" as well.

The ground does a lot to warm you, radiating terrestrial temperature BBR to your entire lower hemisphere, basically it's the same ~100W/sqm.

Sleeping under some tree cover on a clear night vs sleeping under the clear sky makes heaps of difference.

That's why he insulates his box from the hot planet and enjoys the fact that the air is largely IR-transparent.

The black body temperature of the night sky on earth is much warmer than the temperature of the CMB in the universe. You can test this by aiming an IR temperature gun on a clear night.

Relevant stack exchange: https://physics.stackexchange.com/questions/153839/what-is-t...

It's possible to make ice in the desert with this effect. Not too hard to believe.

It’s also commonly freezing in the desert at night

Snow is a near perfect black body radiator

This was a key plot point in the 1990s video game: Midwinter

https://en.wikipedia.org/wiki/Midwinter_(video_game)

Therefore imagine the impact of reduced snow and ice due to global warming. It’s free cooling and we’re throwing it away.

The albedo is the part that's bad to lose. I don't think snow emits especially strongly in the frequencies relevant for this. A highly imperfect blackbody emitter could cool more effectively.

It's an issue for amateur astronomers too as you have a mirror in a long tube pointed directly at the sky: https://britastro.org/2021/dealing-with-dew

Almost all surfaces radiates heat. You can heat and cool using radiation. It’s the same way satellites is cooled, when you can’t cool with convection (interact with other atoms).

Black absorb more radiation, but it also radiates more

Great question!

“PDRC surfaces are designed to be high in solar reflectance (to minimize heat gain) and strong in longwave infrared (LWIR) thermal radiation heat transfer through the atmosphere's infrared window (8–13 µm) to cool temperatures even during the daytime.”

https://en.m.wikipedia.org/wiki/Passive_daytime_radiative_co...

I also posted this as a separate post for visibility. It’s an amazing technique that more should be aware of.

I guess that not having other warm objects pointing at the water, will not warm the water (example: a wall radiates infrared so it heats the water if the wall has "sight" over the water).

Insulating the walls of the container with the water servers the same purpose - not have the water get heat from the surroundings.

I'm guessing he just froze water when the temperature dropped below zero at night, and everything else is nonsense.

Humans have been using this technique (radiating heat into the cold of space) to freeze water in hot places, for a very long time.

https://www.realclearscience.com/blog/2018/07/09/how_people_...

That's an interesting article. I'm aware of this sort of radiative cooling, but usually convection and conduction, along with evaporative cooling will be much, much more important than radiative cooling. I'd like to see some kind of report or article showing the cooling potential in W/m^2. I remember someone made a metamaterial optimized for emitting heat in the wavelength of the atmospheric window, but the value was quite low.

You might be thinking of the work led by Professor Shanhui Fan @ Stanford some years back;

https://news.stanford.edu/2017/09/04/sending-excess-heat-sky...

This is being commercialized, but I don't know how successful it has been to date.

https://www.skycoolsystems.com/

I wonder how you prove/disprove this. I imagine if you had a control box with water exposed to the air that wasn't "radiating heat," then that should show if the solar freezer was indeed working or not.

He did try a basic control with his first version: “i got ice consistently over several nights, even though there was no frost on the ground, and pots of water sitting next to the box cooker did not freeze.”

Ah you're right! I totally missed that line. Cheers.

You could build a few test configurations.

One as stated, insulated fridge, open on top.

One, insulared fridge, open on top, under an open air shelter, no walls, just a roof, where the roof is a significant distance away from the top of the fridge, so airflow is not impacted, but there's no view or the sky. Possibly variations on this with the roof being ir transparent or not.

Another test configuration could have the fridge closed, but with forced air ventilation.

> on rainy nights my food gets wet

Indeed. But overall this guy sounds like a good person to have in your apocalypse survival community.

One phenomenon I've found peculiar on frosty mornings is that the grass directly under my (apple) trees doesn't freeze, leaving a noticeable silhouette on the lawn. I've assumed it was due to the cool, heavy air falling and the tree acting as a shield but this article makes me think there are more complex factors at play.

Cool. Might be able to do even better with self-cooling paint [0,1] during the day to reject heat and Vantablack [2] at night to radiate it? (although I wonder how Vantablack would do at radiating with its microstructured surface)

[0] https://www.parc.com/technologies/self-cooling-paint/

[1] https://www.purdue.edu/newsroom/releases/2021/Q2/the-whitest...

[2] https://www.abc.net.au/news/science/2017-01-13/vantablack-wh...

That self-cooling paint works at night as well. It's not just white paint that reflects light, it also radiates heat pretty well. The YT channel NigthHawkInLight developed a variation that can be produced with (more or less) household ingredients and it cools to 3-4 degrees below ambient temperature if I remember correctly.

Sort of related: the YouTube channel "Tech Ingredients" do some cool things with radiant cooling: https://www.youtube.com/watch?v=dNs_kNilSjk

With the exception of two vacations to California, I have lived my entire life in high humidity locations. This sounds like magic, I understand the science and all, but that it can actually work is wild. It might work for me in the winter, when we regularly get drier weather and spend a most nights flirting with freezing.

Hey, this reminds me of Bill Beaty's Human IR Sense: http://amasci.com/weird/humanIR.html

It’s good for a couple degrees I think you can get water to freeze at 36 on a clear night for example

As in when it’s 36 outside?

Units, dammit!

If it's either 36C or 36K at night, you've made a bad decision on where to live.

Don't forget Rankine, although you are making equally bad decisions as Kelvin there

TIL about the Rømer and Réaumur temperature scales which are why Rankine is not just °R but °Ra. I guess everyone was doing it back then!

https://en.wikipedia.org/wiki/R%C3%B8mer_scale

Clear F. I don’t see you passively cooling anything to below freezing when it’s 36C outside.

Innsulate it well enough and why not

I agree, but it's simple to guess from context here.

Reply on news.ycombinator.com