Solar panels as radiant barrier
Hi all — I can’t find it, but I recall someone here saying that solar panels were good radiant barriers for a roof. Do solar panels have a net cooling effect compared to asphalt shingles, tile, etc.? What part of the assembly is reflecting the IR?
I’m looking for ways to keep our future attic as cool as possible. I read Martin Holladay’s impaling of “powered attic ventilators”, née attic fans, so I’ve zeroed in on radiant barriers and rafter insulation. We’d have R-23 of mineral wool between the rafters (a 5½ inch batt) and a stack of two R-23 mineral wool batts on the ceiling, for R-69 total. The attic would be unconditioned, but the R-23 of rafter insulation might make it tolerable while working in it. It would be neat if the PV panels helped. Using a radiant barrier underneath them, like in the attic, is suboptimal since it would heat the panels and worsen their performance.
Thanks for your help.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Dear BlueSolar,
“Radiant barriers” are largely a marketing ploy to sell aluminized bubble wrap. Fortunately anything and everything besides air is actually a radiant barrier (with the exception of some specific plastic films).
So yes, solar panels are a radiant barrier, they will prevent light from hitting the parts of your roof that they cover. This in turn may keep your roof a little cooler and less heat will conduct into the building. They also block some air flow which may in turn let things get hotter, but are air gapped for this reason. Light colored shingles will reflect more light than dark shingles, this will similarly keep the roof cooler. There are many benefits to solar but this isn’t really one of them.
Read this article for the full rundown on radiant barriers from Martin:
https://www.greenbuildingadvisor.com/article/radiant-barriers-a-solution-in-search-of-a-problem
Also look into “cold roof construction”, it’s a method of putting a ventilation channel between the roofing and structure to prevent heat transfer.
At the end of the day limiting heat flow primarily comes down to the R value of the roof assembly.
"“Radiant barriers” are largely a marketing ploy to sell aluminized bubble wrap."
Not true. Radiant barriers definitely have their place, especially in cooling climates. Where radiant barriers gets into trouble is when unscrupulous promoters of them advertise insulating qualities they don't actually have. The comical part of this discussion is that in effort to punish the bad actors of that practice often people opposing radiant barriers are making an equally awful hash of the facts. I would hazard a guess that you are from a primarily heating climate and just don't know better.
Those of us in cooling climates are well educated on the use and misuse of radiant barriers. Radiant barriers are even written into code in California. They work. The best application is in new construction by installing roof sheathing with one surface being a radiant barrier. That side of the sheathing is on the attic side. There is a small upcharge but is totally worth it IF you live in a primarily cooling climate. As I said, its even written into the building code here because it works.
Thanks Eric. I'm thoroughly stumped by radiant barriers. Why is the shiny part facing into the attic instead of facing the sky? I've read this proviso before and it seems backward to me – what am I missing?
Holladay also described aluminum foil as low emissive, which also confused me. Aluminum is a great conductor of heat, which is why it's used for cookware. Does emissivity mean something very different from thermal conductivity?
Emissivity is about radiative heat transfer (Light generally outside the visible spectrum), and thermal conductivity is heat flow through a material. The thermal conductivity of aluminum in this case is unimportant because of how thin of a layer is used - but yes, aluminum has high thermal conductivity.
Low emissivity means a material reflects thermal radiation AND means that when that material gets hot it doesn’t emit thermal radiation. If the shiny side is facing into the attic you expect it (the shiny stuff) will get hot, but won’t radiate that heat into the attic.
Even in hot climate they only have an impact with low levels of insulation see:
https://web.ornl.gov/sci/buildings/tools/radiant/rb2/
If you take the number there with R19 attic insulation, a radiant barrier reduces heat flow by around 30%. You can get similar energy reduction by putting an extra 4" of blown insulation, which also works year around not just in the summer.
Basically with any code min level of insulation in the attic and no air handler there, an RB does squat. If it is free, doesn't hurt to put it in, but I wouldn't spend extra money on it.
The sad reality is there are a lot of houses out there with no attic insulation and air handlers/ducting up there, in that case an RB does make a big difference.
BlueSolar,
Thermal emissivity and thermal conductivity are two different things.
Usually things that are shiny conduct heat well but don't radiate well.
For example, you can't use one of those IR gun temperature sensor on shiny surfaces (without changing parameters) as it will not give the correct temperature. If you take that same shiny surface and paint it black, the IR gun now works. This is the similar reason that RB needs to stay shiny and clean for it to work.
An RB with the shiny side on the inside of the attic works because it is won't emit as much radiation. It isn't about "reflecting" the heat.
Akos,
I am honestly surprised by how high the numbers are that are quoted in the article you linked to, going to take a deeper look. I’m wondering if the RB in question also acted as an air barrier in houses without air sealing and that added to the effect.
Dear Eric,
My information is from a heating climate that’s true, but Martin’s math is pretty clear from the linked article.
The radiant barrier adds at most one or two R to an assembly, if this is put into an uninsulated building/attic that is a massive improvement. If instead it’s put into a well insulated building (let’s say R-30 roof), then it’s only a small improvement in relative terms, and something like a half inch of foam would still be better.
In your example, are you suggesting an uninsulated attic? If that’s the case I 100% agree with you that it would make a big improvement and decrease attic temperatures - but BlueSolar was already talking about adding insulated between rafters.
Yes, we'd have R-23 of mineral wool between the rafters and R-46 of wool on the ceiling below. I updated my question to clarify this.
Would it make a difference if the radiant barrier was on the exterior side of the roof, underneath the panels?
Joshua,
I have personal experience with the benefits of radiant barriers in my own home's renovation. I used to experience absolutely sweltering indoor temperatures in summer. Admittedly there was very little attic floor insulation at the time. I question strongly the figures of only 2 degrees different with a radiant barrier installed. That is not accurate. With my radiant barrier installed it was all of 10F cooler. That was with no insulation added. I could exist! My 10 degree experienced difference is not negligible.
I had always aimed to try to get away without air conditioning in my climate. . It is almost unthinkable for most here to do without some form of AC. Summers highs here average mid 90s with multiple excursions to the 100s. So my goal has been to give myself every advantage to accomplish that. Once I made the indoor temperatures tolerable with the RB, but not comfortable, I then proceeded to add R-50 blown cellulose to the attic floor. The final touches were deep overhangs for southerly and westerly windows, ceiling fans in every room and a whole house fan. All those things combined have done the trick. I am doing without AC comfortably in a climate that averages mid 90s highs in summer.
I certainly do fatigue of people telling me what to think who don't walk in my shoes. It's exactly the same thing as Midwesterners, Southerners, and Easterners saying whole house fans don't work because they bring in humid air. How would you guys feel if some arrogant person from the West said you are spending way too much time and energy on preventing cold temperature infiltration into the home because its not that bad. Close your eyes and let that picture seep into your consciousness and you will know how many of us feel! You simply do not know what you do not know.
Thanks Joshua. I didn't realize that solar panels are air gapped. The solar tiles and shingles that are coming into the market are interesting in that they replace normal tiles and shingles, so they must not be air gapped – I wonder how that will work as far as thermals and air flow. (I mean products like Tesla's solar shingles, Forward's, RGS, etc.)
I read Martin's article and I'm thoroughly confused. Why would a radiant barrier face *into* the attic? How is aluminum a low emissivity material if it's a great thermal conductor?
Physics is why good thermal conductivity and low emissivity can be linked.
In the power electronics world, heatsinks are common. Heatsinks are usually made of aluminum for a variety of reasons (it’s relatively cheap, easy to work with, and a pretty good conductor of heat). The heatsink’s purpose is to draw heat from the small but hot area of a power semiconductor device and transfer it to the cooling air. Heatsinks have a relatively thick “conducting” area and a bunch of thinner fins as a “radiating” area. This is because you want a lot of surface area to be able to transfer heat better. Heatsinks are rated for thermal performance using a unit somewhat similar to electrical resistance.
Anyway, heatsinks are usually made of BLACK ANODIZED aluminum since the black surface better radiates thermal energy, so black surfaces perform better than shiny silver aluminum surfaces.
With really high power density semiconductors, often RF amplifiers and advanced microprocessors, it is common to use a copper “heat spreader” between the power device and the aluminum heatsink. This is because copper is a better conductor of heat than aluminum, but aluminum radiates heat better than copper.
Quirks of the physical world.
Bill
By the way, Joshua, would it help if solar panels used low-e glass? I mean help reduce panel temperatures and resulting heat radiating into the roof/ceiling. I get low-e confused with the argon, but I think low-e reflects IR back into the sky.
I would assume the portion of solar radiation that is converted to electricity would be that much less energy emitted. (a 20% efficient panel would emit 20% less radiation, is my guess?). This being regardless of the emissivity properties of the surrounding materials and what-not.
I am having a hard time tracking down why, precisely, a surface reflective to IR is also intrinsically and identically less emissive of IR. I understand that if the IR is reflected, it won't be emitted (total reflected + total emitted + total transmitted = 1), but that the characteristics for IR reflectivity are identical to emissivity is a bit more of a mystery to me. This seems to be the closest thing to an answer I can find on it(from: https://irinfo.org/03-01-2014-swirnow/)
"Because the free valance electrons are not strongly attached to the atomic lattice, when hit by the incident photons, they do not transfer this energy to the lattice. The skin effect is present, the incident wave is “reflected”, and the metal does not absorb the incident energy.
This is why polished metals at low temperatures cannot efficiently absorb incident infrared energy.
Due to the weak connections between the cations and free electrons, when mechanically warmed (i.e., resistive electrical connection), the cations cannot transfer enough energy to cause surface electrons to oscillate and emit photo electrons."
In the case of low-e on the underside of the roof, a significant portion of the heat energy is conducted from the topside of the roofing (therefore not reflected at all) yet it still emits at a low rate. My read on it is that the properties that determine reflectivity apply identically to the properties that determine emissivity, even when reflectivity is not part of the equation.
In theory, it would seem low-e could just as easily go on the topside of the roof, but the drawbacks I can think of would be:
1) oxidation and other degradation of the surface condition would lower the reflectivity over time (the underside is better protected)
2) its probably undesirable to have a highly reflective surface as the visible surface of your roof
3) While it will directly reflect certain wavelengths, other higher energy radiation will not be reflected and so will be absorbed, and subsequently emitted towards the underside of the roof (unless that too was low-e). So now you've had to create a roof with two low-e surfaces instead of one.
>"we'd have R-23 of mineral wool between the rafters and R-46 of wool on the ceiling below. I updated my question to clarify this.
Would it make a difference if the radiant barrier was on the exterior side of the roof, underneath the panels?"
It would make no significant difference whatsoever.
Roofing that has high reflectance in the solar spectrum is referred to as "cool roof" materials, and there are standard for measuring & labeling a solar reflective index (SRI). While that is somewhat higher performance than the shiny stuff on the underside of the rafters (or a shiny backside to the roof deck), the affect on peak & average cooling loads or energy use would be pretty tiny on a house with R23 between the rafters and at least that much additional on the attic floor, even if there were ducts and air handler between those insulating layers.
If you are in a cooling dominated climate it's worth considering higher SRI roofing even if the total impact is small, since there is often zero up-charge. The Cool Roofs Rating Council keeps a searchable product database here:
https://coolroofs.org/directory
Solar panels with a vented space between the roof deck & panels on a steeper pitched roof (4:12 or higher) will lower the peak & average cooling loads too. While the albedo of PV panels is pretty small (= low SRI), the shade of the a panel limits the incident radiation on to the roof, and the vent space promotes convection cooling once the panels heat up in the sun. The combined effect is to lower the temperature of the roof deck, despite the low albedo of the panels themselves. Without the vent space it's about the same as having dark asphalt shingles on the roof deck (not a high SRI cool-roof). A very limited popular press discussion of that lives here:
https://blogs.scientificamerican.com/solar-at-home/the-albedo-effect/
" R-23 of mineral wool between the rafters and R-46 of wool on the ceiling below"
Depending on the climate you are in, this can be a really bad idea. You are creating an attic space that is neither inside or outside your thermal envelope.
In the winter time (might not be an issue down south) you can end up with pretty high humidity in there and a lot of mold.
The best is all the insulation on the attic floor or all the insulation in the rafters. If the insulation is in the rafters the attic should be part of your conditioned space. If the insulation is on the attic floor, then vent the attic.
Since insulation on the attic floor is the cheapest, I really don't see what benefit there is to split it and potentially create a problem.