Thin Roofs...and foil

Even now at the winter solstice, with the noon solar radiation reduced some 20 percent, any thin darkish thin surface (such as a metal roof or a tile roof) heats up to 50 degrees C or more, with the heat being almost immediately transferred undiminished to the underside, as you can see from the temp reading of the underside of this rubberized canvas tent which is supposed to provide cooling shade but really provides heating shade which will radiate on the order of (52-32)*5=100 watts per square meter to my balding head, plus perhaps as much again convected to the air below. Not as bad as the 700 watt sun, you might say, but not comfortable either – and suffocating if the underside of the roof is not radically ventilated. So next time you are sitting under one of these tents at a Thai wedding or funeral, be warned -- it's the heat, not your emotions, that is causing you to feel like you are going to faint.

Here is an excerpt from the manual of the Australian company Bluescope Steel who make (and promote) steel roofing in Thailand. This indicates a good reduction in heat transmission through thin roofs (steel, cement tile, etc) with foil insulation, properly installed (air space above and below foil, and foil with shiny side down.) For a dark roof, the thermal transmission is said to drop from 165 to 22 watts, almost 90 percent.

Foil sales folks must be delighted with this conclusion. But be aware that it contains at least one hidden but important assumption -- very low convective transfer from the roof to the room below because the air beneath the foil is taken to be perfectly still. But this is not true in any case for many reasons. Convective cells develop even in sealed spaces, desirable fans move air, and required ventilation to replace that air which will otherwise heat at about 2 degrees per minute (ventilation rate of approx one room air change per minute--that's right, not one or two air changes per hour, as in most attic spaces even with vents) so there is in fact much more downward convective transfer from the roof than this chart suggests.

My estimates do not yield quite such promising results – my 1935 Eshbach Engineers Manual (college edition) indicates an R value bonus of only about 2 or so for a single layer of foil. And a direct measure of the efficacy of shiny vs. black spray paint surface on the foil which can be seen in the simple room simulation which I performed last month yields a reduction of about 50 percent in heat transmission. This implies that the heat flux from downward radiation and convection are about the same. Th foil brigade will try to tell you that here is not convective transfer downward at all, and this just ain't true.

Want to know the exact ruth? The fact is that the physics of downward heat flow in roofs is extremely complex, the stuff of learned papers by mathematically oriented professors.

But I don't question that properly installed foil will improve matters-- from very bad to not as bad--in the case of a thin tile, cement, or metal roof. Studies such as those done in Florida suggest aircon energy savings of perhaps 15 percent with the addition of foil. My caveat is that in most practical situations one cannot rely on foil alone to prevent uncomfortable heat buildup from a foil-insulated roof.

Some reasonably unbiased web sites describing use of foil insulation are:

Florida studies indicating reduction of energy 25 percent and temp 6 deg:

http://www.fsec.ucf.edu/en/publications/html/FSEC-CR-1231-01/index.htm

Wikipedia:

http://en.wikipedia.org/wiki/Building_insulation

Q&A on foil insulation by experienced researchers:

http://www.fsec.ucf.edu/en/publications/html/FSEC-EN-15/index.htm