This is a simple controlled trial of whether I can get a few degrees F of “radiative cooling” in my raised beds, using radiant barrier. See a few posts back for details.
In a nutshell, total failure. And I have no clear idea why.
The “treatment area” is the inside of a little radiant barrier tower. To make the treatment area, I wrapped a standard tall tomato cage in a piece of 4′ wide (tall) construction radiant barrier. Open at both ends. I then set that on an empty section of a raised bed and slightly buried the bottom edge (for air-tightness). The top is open to the sky.
The “control” area is the bed next to that.
The weather tonight is about as good as it gets for this, in Virginia, in July. The wind is almost still, currently 5 mph from the Southwest (good). The relative humidity at 9 PM is slightly below 50% (as good as it gets). And the sky is mostly clear, which is better than fair for a mid-July day. (So I’m not bothering to put a cleat plastic sheet top, because I don’t think the outside air is going to mix much with the air at a bottom of a 4′ cylinder.)
Theory tells me this will not work as well in the humid Southeastern U.S. as it does in the desert. Water vapor is a powerful greenhouse gas. But I think the lack of solid cloud cover gives me at least a fighting chance of getting an effect large enough to be observed.
The measuring devices are a pair of temperature data loggers. They are old and crude, but they work. I’ve used them enough to be confident that their temperature readings match. Or did, as of last year. They track temperatures every half-hour, to the nearest 2 F.
I set the temperature loggers to begin at 9 PM, then dropped both them on the ground. From the same height. At a few minute before 9 last night.
Results
The radiant barrier tube (blue line) kept the soil inside about 4 F warmer, overnight. So, a clear failure in terms of cooling.
First, I think this is a real effect, i.e., not experimental error, though the result is so counterintuitive, I need to check that by moving everything and re-running this. In the past, these two gauges have read identically (so it’s not a calibration error), and I double-checked the internal serial numbers to make sure I didn’t swap them.
I don’t think the explanation is that the cool night air couldn’t get inside the tube. It’s not as if there were external source of a cold breeze last night. What we had was mostly still air, subject to overnight radiative cooling. (It’s easy enough to test this, though, by swapping the radiant barrier for IR-transparent polyethylene sheet.)
I don’t think it’s plausible that this is a “waveguide” issue. Electromagnetic waves won’t enter spaces that are too small for them, meaning, I think, about half a wavelength. But the wavelength of IR — even long-wave IR — is tiny compared to the scale of this. Wikipedia lists 1 millimeter as the limit of what is considered IR.
All you need to do is look in your microwave to see that this isn’t an issue. Microwaves are longer and lower-frequency that infrared, but the waveguide that pipes microwaves into your oven is just a few square inches in cross-section.
It’s possible that the rough and crinkly surface of the radiant barrier actually reflected a fair bit of IR back down toward the soil. More than the moisture in an equivalent-sized column of air would have done. But the temperature increase — 4 F — seems vastly too high to be the product of a bit of incidental backward-reflected IR. That’s nearly as much warming as I got from literally capping off (covering) a bed with radiant barrier, in earlier experiments.
So I’m stumped. The followups need to be:
- Validate the result by moving the radiant barrier to a new location and redoing.
- Rule out “blocked flow of cool air” by replacing the radiant barrier with IR-transparent material.
No sense in speculating further until those results are in.
Bottom line is that I figured this might not work for cooling. That it appears to work for warming is unexpected. And even thought this won’t solve my problem (nights above 70F), I’d like to know why it’s doing what it’s doing.
Edit: But at the end of the day, why bother? It was a longshot. It doesn’t work. I’d say chances are good that it’s never going to work. Do I really care enough to try to figure out exactly why it won’t work? No.
It’s a dead horse. No beating required.
