Solar Generator That Works At Night

I'm worried about the efficiency of this notion ... might not be anything so useful as a SolarChimney. It's something I thought up on a bet, but I can't really see why it wouldn't work. I hope wikizens with better physics heads than mine will comment/improve this:

I had a solar oven. It was a plane-mirror approximation of a parabola - 9 foot square of flat mirrors arranged so they reflect the sunlight that strikes them onto a black metal box in the middle about as big as a large soup pot. Which is to say I also had a solar refrigerator ...

No, really - you take an efficient solar oven and point it at the night sky on a cloudless night. Put an ice tray in it and block any external vents so the night air can't get in. It turns out the thermodynamics work perfectly in reverse - you're arranging "rays of cold" onto a focus. The water freezes solid.

I did this trick in the California summer several times. Bake a turkey in the same device that later ices your drinks. Very literally cool.

So the notion is you build a long trench with a parabolic cross section. Line it with an efficient thermal reflector - aluminum foil - and run a black metal pipe down the focus. You don't need a heliostat because there's no moving source - the whole night sky radiates cold. Well, except those annoyingly hot clouds. Better do this out in a desert if it's going to be your only power source.

Fill the pipe with freon gas and let the cold condense it into freon liquid. Then let the freon flow through a valve into a network of pipes warmed by atmospheric air. It will boil into a gas again and the expansion runs a piston. Use any efficient steam engine design for that and hook the piston up to an AC generator. Let the gas flow around to the solar pipe inlet to make a cycle.

If you think about it, the whole arrangement is powered by the sun, which warms the atmosphere so you can extract some work from it. So it's a SolarGeneratorThatWorksAtNight. ~ Pete.

Before I try to disassemble my old upright freezer and make one of these fool things in my backyard, I'd be obliged if someone with a real head for thermodynamics would tell me it's too silly. Note that you can always build a longer trench ...

It's "too silly". I'm left feeling that this whole page is a joke and that any serious response to it will receive ridicule simply for having taken it seriously. However, it may be my many years as a physics educator that leave me feeling that unease. If it were coming from a student, rather than one of the heavily self-educated people on this wiki, perhaps I'd feel differently.

There are no 'rays of cold' in the night sky. An object cools by the usual suspects at night as it does in any other place: convection, conduction, and thermal black-body radiation. The radiation is infra-red 'rays of heat' flying away from the object... which, if you squint and do some cartwheels, does seem a little like 'rays of cold'. Cooling by thermal radiation works better on a cloudless day (clouds reflect infrared rays shining away from the rest of Earth back to the ground, like a greenhouse). However, one should note that the 'solar generator' is not doing any work; it is the object cooling down that is firing off all those rays of heat. The presence of the mirrors, having almost no energy to redirect, is pretty much irrelevant.

The easy solution for a solar-powered refrigerator that works at night is: solar generator + batteries + electric refrigeration unit.

I did rather think it was silly. I Am Not A Physicist. I don't even play one on TV. All I meant by "rays of cold" is that the thermodynamics seems to run the same way forward as backward. So you're saying that my tray of ice cubes left out in the solar cooker on the summer night in California froze because it was insulated from the IR radiated by its surroundings, and from heating by ambient air. The mirrors had nothing to do with it ... right?

Insulation keeps cool items cool and warm items warm; it prevents convection, resists conduction, and the really good ones use an internal and/or external reflective coating to resist heat-loss and/or heat-gain by thermal radiation. A common example is the thermos (vacuum flask), which uses a vacuum to prevent all convection and conduction through the walls... and thus loses heat only via some convection/conduction off its insulated stopper and due to thermal radiation. Inside good coffee flasks, you'll find a reflective surface (though coffee flasks aren't concerned with keeping cool things cool, so they often won't be silver-colored on the outside).

A solar cooker will have a dark-colored exterior and interior surface. Under sunlight, this surface will rapidly heat to thermal radiation and conduct that heat internally... then emit it internally along with heating the internal air. At night, a warm object in the cooker would rapidly cool by thermal radiation... as the infra-red rays would fire off towards the dark walls, be absorbed, then subsequently cast into void blue yonder. Of course, that works without the oven, too. The oven serves an important purpose mostly in preventing environmental contamination (e.g. some bird saying "hey! look! that's clean water!").

The mirrors below the whole contraption will provide some insulation from thermal radiation off the ground (deflecting it, much like the silver exterior of a thermos). Thus it is, indeed, the case that a solar cooker would insulate from IR radiated from the ground below. This could make for a difference of a couple degrees, maybe even several, depending on the environment. It would be wrong to say that the mirrors are doing nothing at all. If anything, it is those mirrors that provide protection from IR... because the 'oven' portion would be horrible at that.

However, the solar cooker provides very little insulation from heating by the night air. The material of the cooker itself would approach the temperature of the night air, and that would be carried to internal objects (like that tray of ice, and the internal air). Air does not readily absorb thermal radiation (the atomic structures are too small for the wavelengths of IR, or even visible light), but it does convect off the Earth surface and carry heat to high in the atmosphere. Even so, air itself is a decent insulator (air is mostly vacuum) and can be of a lower temperature than the Earth surface. I imagine that your solar cooker was mostly insulated from the IR and most conduction from the ground (a solar cooker would be designed to not conduct heat away from the oven). I.e. it insulates the oven from the Earth (oven and earth? I've several bad puns rattling in my head now.) Anyhow, if the tray of water freezes, it's because the temperature of the night air dropped below freezing.

I haven't explained the construction of the cooker very well. You have it all right except that the black metal box was covered by an air-tight glass cone. This enabled the contents to radiate IR into the night sky while largely insulating them from convection heating by the ambient air. I don't believe the night air temperature dropped below freezing when I tried this.

Glass generally conducts heat well enough. I'm not convinced it would provide significant insulation by itself. Is there a layer of vacuum between two layers of glass? Or is it a special insulating glass?

I regret I left the cooker in the US when I came back to Oz, so can't really say. It looked like ordinary glass to me. I had read about others with solar cookers doing the same thing, so I'd be surprised if it was a special kind of glass. I can't think why you'd need a special kind of glass to make the cooker cook.

Google found some others who have made a solar refrigerator with even simpler constructions than my fancy cooker - here's links:

Insulating glass would trap considerably more heat (preventing conduction as well as convection of heat away from the cooker). In the case of freezing things, it would also make a difference... it really would prevent exterior air from affecting the temperature of internal items. Glass, like most other things, takes the temperature of the air around it. Insulating glass generally has layers such that the glass on one side can have a large heat-differential from the layer on the other side, which prevents the surrounding air from taking heat from one side of the glass to the other. It does look like normal glass, of course... people like putting insulating glass in their homes to save on heating bills (they pay for themselves after 5-10 years, depending on climate). Glass isn't nearly so bad a heat-conductor as metal (~16-50x slower than steel depending on whether it's stainless, 400x slower than copper), but it's no insulator (~40x faster thain air or cotton, ~60x faster than argon). The common insulating glass is simply layered, with air or argon-gas in the middle. There is no real convection of heat away from a sealed place, and no radiation from air or argon, so trapped air must use slow conduction to transfer the heat.

I have doubts that a small layer of glass would provide much protection against heat from the outside air. It would take that air and conduct its heat inwards more readily than the food it is guarding (though not quite so readily as the iron box). The glass mostly provides a layer of protection against convection. However, you put insulating glass in and slow that down by a factor of 40x or so, the equation seems to me a bit different. It may very well be possible to cool the object below the temperature of the night air by several degrees with heat leaving faster than it enters. It still wouldn't do so well on nights in hot, humid locales, but if the temperature normally drops to chilly, you might be able to get it down to freezing. I'd need to run through the equations by hand to be sure... most of my statements so far come from that 'feel' you get for a subject after working with it for a while.

I'll check out the links you provided.

The links discuss a changes of a few celcius degrees to up to ten celcius degrees depending on construction, which fits my understanding of the subject. You'll also note the suggestion in the solar funnel page: "It helps to place two bags around the jar instead of just one, with air spaces between the bags and between the inner bag and the jar. HDPE and ordinary polyethylene bags work well, since polyethylene is nearly transparent to infrared radiation, allowing it to escape into the 'heat sink' of the dark sky." That space there is providing the conduction shield.

Okay ... so I'm back to thinking this maybe could work. Run the aforementioned foil-lined trench down any 45 degree hill. Can't be a cliff because it has to tilt enough so the trench contents can't "see" the warm horizon. Seal the trench across the top with insulating glass or similar. Have to line the trench with an excellent insulator too of course. If your trench has enough of a fall you can put a nice turbine inline to get work from the weight of the condensed freon ...

Heck, make the trench long enough and the bottom end will get below 77K - then you can dispense with returning freon gas to the inlet - just use liquified air and let the evaporated exhaust gas escape back into the atmosphere. And to pick up the GoreAndBransonPrize in the bargain you extract the chunks of dry ice from the liquid air slurry and use the energy generated by the whole setup to fire them into orbit.

I knew the answer to global warming was going to be simple ... or at least simple-minded ;-) ... ~ Pete.

Some observations which lead to a couple of questions

Could you explain why frost appears on the metal and glass surfaces of my automobile on chilly nights, but not on equally chilly days, while not on plastic and insulated items in the same air?
Plastics and many paints are hydrophobic; water tends to roll off them. Glass (untreated with RainX or Aquapel) and many metals are hydrophilic; they transiently bind to the polarized molecules in the water and keep it attached. Frost will most readily form on surfaces to which water can easily bind... though it can form on either if there is enough water. Further, glass and metal are more likely to cool quickly (due to their conductive properties), and thus gather condensate from the atmosphere at ideal times. These would be the main reasons behind 'metal and glass' vs. 'plastics and insulated items'.
As far as 'chilly nights vs. equally chilly days': frost has a fragile existence... it only forms at the very edge of freezing... any lower, and the air will dry up; there will be nothing to freeze. Any higher and frost will melt. The most common time to see frost form is at dusk (as the temperature crosses to freezing)... though I guess the precise time is environment dependent.

Could it be that good conductors of heat are also good radiators as well?
In a way, yes: Before heat-energy can be radiated from a surface, it must be carried to that surface. Conductors of heat very readily carry heat to the surface. If you are building a contraption that is intended to carry and radiate heat away, a good heat conductor would be an important component of the design. Thus, in practical devices, radiators are often built with conductive materials. However, heat conductivity and emissivity, as material traits, are not tightly correlated. Color and emissivity have a much tighter relation: in general, black things are good radiators and white or silver things are not. That, of course, indicates only the visible light spectrum... but how an object appears in the visible light spectrum does correlate loosely with its 'color' in other spectrums.
Glass is not a good radiator of heat; it tends to let infra-red 'heat' radiation pass right on through unless tinted ('green' houses are named such for a reason), and doesn't radiate so much itself. Glass does conduct heat easily (compared to insulating materials, at least). It is an example of good (really 'not bad') conductor of heat yet poor radiator of heat.

Pete,

I am also not a physicist, but I AM a ham radio operator. And this is easily described in radio terms, since parabolic reflectors work the same way for light as they do for radio.

What makes a solar fridge is the same thing that makes a Yagi-Uda antenna work -- reciprocity.

Place a cup of water out at night, and it doesn't freeze. Why? Because all around it, there are factors contributing to its heating as much as its cooling. At some point, you reach what's called thermal equilibrium, where the cup of water won't cool any further. Any time it starts to cool below the equilibrium, the surrounding environment starts to radiate it with IR, thus bringing its temperature back up. If it heats up too much, it starts radiating IR to its environment.

Introduce a reflector, however, and now you confine the source of irradiation to the mirror itself, which contributes a negligible amount of IR itself, and that of deep space, which ... has negligible amounts of IR from wherever you're pointing the mirror. By negating the environment's contribution of IR, you now are able to provide the cup of water a more perfect sink for its own IR radiation. Note that the cup of water hasn't changed how fast it heats or cools (this is determined largely by surface area). All you did was alter the environment to which it's sinking its IR.

The radio industry has terms defined for this phenomena. It's true that a 100W light-bulb gets pretty darn warm, but it's also true that a 20W argon-ion laser can lascerate the skin. The difference is that the lightbulb models an isotropic radiator of its power, so that, in any given direction, power density is pretty small. Add in a reflector behind that lightbulb, and intensity doubles. The lightbulb is still radiating at 100W, but now the illuminated surface is seeing twice as much power as it originally received. This concept is called Effective Radiated Power, or ERP. As far as the surface illuminated is concerned, you upgraded to a 200W (isotropic) lightbulb. The obvious disadvantage, of course, is that the antenna allows light to be focused onto a smaller area. General illumination is no longer possible. Perhaps a more visible demonstration is the average flashlight. Take the bulb out of the flashlight, and it's barely bright enough to illuminate a small bathroom. Put it in front of the flashlight's reflector, and suddenly you can see many tens of feet ahead of you. Thus, when you see claims of, "it gives +12dB of gain!!", you must consider that that is directional gain, not isotropic. Like velocities, you must consider direction vectors too.

If an object is going to radiate its energy as infrared, having a reflector will allow you to sink the object's radiation in a more focused manner, to something of arbitrarily low temperature. The more surface area is devoted to reflection, the colder the equilibrium point.

Assuming the reflector is at the bottom facing upward (like a normal Newtonian telescope), then I am willing to offer the following experiment to test my hypothesis:

Requirements
- A large mug (metal is preferred).
- Two digital thermometers capable of feeding data to a computer.
- Computer software to periodically poll the thermometers for their current readings.
Procedure
- Launch the data acquisition software.
- Fill the mug with at- or near-boiling-point water.
- Place the large mug of hot water out at night, at the focal point of your reflector.
- Point it into deep-space.
- Place thermometer #1 so as to measure the temperature of the water at the top of the mug.
- Place thermometer #2 so as to measure the temperature of the water at the bottom of the mug.
Expected Results
- Initially, both temperatures should read the same.
- As the night bears on, thermometer #2 should start to register a dropping temperature first.
- Thermometer #1's readings should follow that of #2 in time.

I predict that the graph over time of the two thermometer readings should look like this:

 |++++.......
 |     +++   ......
 |        ++       ...
 |          +         .. 
 |           ++         ....
 |             ++++         .....
 |                 ++++++++++++..........
 |------------------------------------------

where the "+" indicates the bottom thermometer readings, and "." indicates the top. This is NOT drawn to scale, but only serves as a prediction of the output data.

This is why, during it total solar eclipse, your ambient temperature can *suddenly* (and I do mean suddenly!) drop 10 to 20 degrees with a marked increase in wind, for the duration of the eclipse. When your only source of atmospheric energy disappears, your heat sink suddenly becomes a heat source, that radiates out into space. And space is damn cold. :-)

Amendment: glass tops will falsify the experiment, because most glass is opaque at infrared wavelengths. A plastic that is transparent to IR is strongly recommended instead.

--SamuelFalvo?

I fail to see how your proposed experiment will test anything. You seem to be missing control tests and the step of identifying which differences (between control and experimental conditions) would deny your hypothesis and which would provide support. Controlling such things as convection of the fluid in the cup will be a big deal in this experiment. -db

A solar reflector is an electromagnetic reflector and concentrator, just like a Yagi-Uda antenna is for radio. The same concepts should apply across the entire spectrum too, including visible and infrared. If an object can "only see" the vast depths of space, then yes, naturally, it must necessarily also see a deeper radiative energy sink. Remember, there are three ways of conveying energy: convection (controlled by placing the object in a suitable container), radiation (controlled via refectors), and physical contact (not possible due to the enclosure).

The control set doesn't need to be specified -- we already have a control set (just leave comparable subject out at night without any reflector of any kind). That's pretty self-obvious. I'm surprised you didn't think of this on your own.

Identifying differences that would deny my hypothesis:
- If the reflector is useless, then the top/bottom temperature-vs-time graphs will be exactly the same with and without the reflector.
explaining what would support it
- Prediction 1. If there is any difference in the temperature graphs, then the reflector must have an influence on the cooling of the subject.
- Prediction 2. The slopes of the temperature graphs for the subject placed in the solar cooker will be steeper; hence, they'll reach their equilibrium temperature sooner than the open-air subject.

Is this pretty clear enough now? I rather thought it was self-evident given the context of the discussion.

Controlling convection in the cup is not important (and the impossibility of it explains why the temperature chart I predict has two distinct traces, not one). Indeed, the natural convection inside the cup is a necessary process for confirming the correctness of this hypothesis. Did you conveniently forget to read where I said that we should have two temperature probes, one on top and one at the bottom of the cup of water?

I'll work up a more formal experiment document -- I realize that the above is pretty much in shambles, hence some of db's confusion. --SamuelFalvo?

AprilZeroSeven

CategoryGosp

See US Patent 4624113, University of Chicago representing Argonne National Laboratory, on behalf of USDOE, 1986. For the math, see "Radiative Cooling Via Optimized Combinations of Aperture...", JB Smith, 2009. Connect a thermosyphon between a compound parabolic radiator and cabinet, and you have a refrigerator-freezer: http://skyfridge.blogspot.com

Connect a Stirling engine between a compound parabolic radiator and ambient temperature heat exchanger, and you can exploit the temperature differential to produce free power: http://stirlingengineforum.com/viewtopic.php?f=1&t=1925&sid=7a9ea9213510c7561d8760663a8523b5