Are there any advantages to using a radiant barrier in an attic?
rvoisard
| Posted in Energy Efficiency and Durability on
I’m thinking of installing a radiant barrier in the attic, above the insulation and attached to the bottom of the truss top chord. For my own house.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Search and download construction details
Replies
Ryck,
The usual justification for a radiant barrier attached to the bottom of the top truss chord in an attic is to lower summertime attic temperatures, a measure that is only important if your attic includes HVAC equipment or ducts.
Of course, a much better approach is to design your attic so that it does not include any HVAC equipment or ducts.
If your attic already has HVAC equipment or ducts, I wouldn't recommend installing a radiant barrier as you propose. A better solution would be to install insulation along the roof slope to create an unvented conditioned attic.
For more information on these issues, see these three articles:
Radiant Barriers: A Solution in Search of a Problem
Creating a Conditioned Attic
Keeping Ducts Indoors
There is no HVAC in the attic. I originally wanted to install insulation on the underside of thee plywood between the truss top cords and was told that I need this space for ventilation or the underside of the plywood could rot or grow mold. I was told that the best place for the insulation was along the attic truss walls against the conditioned space?
Ryck,
I don't know what you mean by "along the truss walls."
There are two types of attics. The first kind is a vented unconditioned attic. This kind of attic has insulation on the attic floor.
The second kind of attic is an unvented conditioned attic. This type of attic has insulation that follows the roof slope.
It sounds like you have a vented attic, so your insulation should be installed on the attic floor.
For more information on this issue, see Creating a Conditioned Attic.
Where do you live, Ryck? Martin's advice that the temperature of the attic doesn't matter is substantially less relevant in a place where the summer temperature gets above 100 for weeks or months at a time--places where a vented attic can easily become a furnace and the high heat flux will overwhelm the insulation and dump a substantial amount of heat into the house. If that describes where you live, then I would advocate consulting the Australians on the matter: http://www.yourhome.gov.au/passive-design/passive-cooling
If you live in a place where it gets very snowy in the winter, then focus your efforts on air-sealing the attic floor and piling more insulation onto the floor, as deep as you can. The radiant barrier is not worth it in this climate, and may create more problems than it solves if it becomes a condensation plane and gathers moisture.
Nate,
To slow heat flow through the attic floor, you need insulation. If your attic is "dumping a substantial amount of heat into your house," then the R-value of the insulation on your attic floor is too low. To cut the rate of heat flow from your attic to your house in half, double the R-value of the insulation on your attic floor.
What locations are above 100F for weeks or months at a time? Even locations that hit north of 120F during the day reliably drop below 100F at night. Weeks of 100F+ outdoor temps would be lethal to humans without air conditioning- if such places even exist on this planet (and I'm not sure that they do), nobody lives there.
Roof decks become solar heated to well above the outdoor ambient temperature, and a radiant barrier at the bottom side of the top chords will interrupt radiated heat transfer between the roof deck and the top of the insulation layer when the roof deck is hotter than the attic air temperature. But at IRC 2009 or later code minimum insulation levels, that has an extremely small effect on the peak & average cooling loads of the house. It's measurable, but down "in the noise" of normal variations in terms of ceiling temperatures or cooling power use.
The Radiant Barrier Fact Sheet published by the Oak Ridge National Labs has moved or is no longer available on the ORNL web site. I would attach it here but it's 9.08 MB, above the 2 MB attachment limit. Current code minimums are even higher than in 2010, which means the benefits are even lower than when that document was first published, but there is no location in the US where at code-minimum attic R values radiant barriers are still financially rational, even with code-min insulated ducts in the attic.
I'm talking about during the day, when the sun is shining. It's curious that we talk up a big game about insulation needing to be surrounded by air barriers on all six sides to be effective, but we conveniently drop this for attic floor installations where it might be a challenge to implement. Heat flow is also affected by delta-T, not just insulation, but when it comes time to keep out heat we forget all about our ability to manipulate delta-T and focus entirely on insulation. It's a cold-climate mindset because you can't really change the delta-T when it's cold, but you certainly can when it's hot; heat gain is driven not only by ambient temperatures but also by solar heating; blocking or reflecting a measure of the solar heat reduces the delta-T, which reduces the heat flow across any given amount of insulation.
There's also the temporal factor. An insufficiently ventilated attic will not cool off very quickly once the sun goes down, maintaining a high delta-T through the insulation for hours after it gets dark. A well-ventilated attic can drop this delta-T quickly to reduce heat gain.
Once you're at R-50, doubling the insulation to halve the heat flow requires going to R-100, which is impossible with most low slope attic truss/rafter designs, even with raised heel trusses, not to mention it's costly. A far more sensible (and feasible) way to halve the heat flow in such a situation would be to manipulate delta-T through radiant barriers and ventilation, not increase insulation. Again, increasing insulation beyond a certain point often isn't really possible in the sorts of low-slope attics we mostly have in places where it gets really hot. My attic is 6:12 pitch and uses 2x4 trusses; towards the eaves, there's very little space for insulation, and the higher you pile it the more attic floor space isn't actually getting the full thickness and rated value. And as we know (because there are GBA articles written about it), an insulation layer of a non-uniform thickness is substantially less effective.
Nate,
It isn't possible for your attic to be "dumping a substantial amount of heat into your house" if your attic is separated from your house by R-50 of properly installed insulation. Your R-50 insulation layer must have major defects to explain the situation you are describing.
"It's curious that we talk up a big game about insulation needing to be surrounded by air barriers on all six sides to be effective, but we conveniently drop this for attic floor installations where it might be a challenge to implement."
"...we..." absolutely do NOT "...conveniently drop this for attic floor installations..."!
Air sealing the top floor ceiling is a critical preliminary before insulating an attic floor, making it part of the primary air barrier. In fact the top floor ceiling plane is one of the most critical of the 6 sides of the cube to air seal, due to stack-effect infiltration drives. Failure to air seal the ceiling in heating dominated climates often results in wet attic insulation in winter.
At sufficiently high density, fiber insulation is air retardent enough to be OK with an air barrier on just the bottom side of the insulation layer. (Cellulose does OK at any density on the attic floor with only the ceiling side sealed.)
"Heat flow is also affected by delta-T, not just insulation, but when it comes time to keep out heat we forget all about our ability to manipulate delta-T and focus entirely on insulation."
The focus isn't at all "...entirely on insulation...". Moderate & high solar reflectance index (SRI) cool roof materials are a standard part of the cooling climate energy use arsenal (even required by code in CA.) It doesn't trump code minimum-R values though. Cool roof shingles are both cheaper than and more effective than radiant barrier, though CA code specifically allows both RB and higher attic R as a substitute for cool roof shingles when using a low SRI shingle for aesthetic or other reasons.
"An insufficiently ventilated attic will not cool off very quickly once the sun goes down, maintaining a high delta-T through the insulation for hours after it gets dark. A well-ventilated attic can drop this delta-T quickly to reduce heat gain."
Not so. The rate of cooling is primarily a function of the infra-red emissivity of the roofing, the pitch of the roof (which affects convection cooling rates), the R-value of the roof deck + roofing, and the thermal mass the materials in the attic. A cool roof shingle is more effective than radiant barrier primarily because it lowers the peak temp of the roof deck, whereas radiant barriers raise the peak temp of the roof deck, which then takes longer to cool off. When the outdoor temps & roof deck drop below the attic temp a radiant barrier actively interferes with removing heat from the attic.
The ventilation of the attic has remarkably little effect on the rate that the attic cools down (unless actively ventilating at a high rate with a fan, which has other issues.) A higher-pitch to the roof (3:12 or higher) offers more convective cooling to the exterior than any amount of passive cooling via ventilation would provide for the attic. In general active ventilation increases rather than decreases the total amount of energy used even with old-school SEER 10 air conditioning equipment. A pretty good survey of the studied effects of attic ventilation in cooling dominated climates (mostly at much lower than current code-min R) can be found here:
http://www.fsec.ucf.edu/en/publications/pdf/FSEC-CR-1496-05.pdf
"Once you're at R-50, doubling the insulation to halve the heat flow requires going to R-100, which is impossible with most low slope attic truss/rafter designs, even with raised heel trusses, not to mention it's costly."
Once your at R50 taking it to R100 would make less than a 1F difference in peak & average ceiling temp, and isn't worth doing for cooling comfort or cooling energy use. (It's somtimes worth going higher than R50 from a heating energy use point of view though, but not in cooling dominated climates.) Spending the RB money or the money for the "extra" R50 on rooftop solar would buy more net power reduction per invested dollar, and would also reduce peak & average attic temperatures due to the shade provided by the rooftop solar.
Both radiant barriers and cool roof shingles lower the average temp of the roof deck year round, which is sometimes a problem in heating dominated climates, not so much in climate zones 1 & 2.
Gentlemen, Thank you for sharing your insight and knowledge with me. Please allow me to clarify. I am in zone 5 North Central Idaho and we boast the record heat for Idaho at 114 degrees farenheit and winters can be rather cool. My home is story and a half, 12/12 pitch with attic trusses. The building is 30' wide with a 16' wide room in the 1/2 story by 8' tall. There is still plenty of attic. Initially I wanted to install the batt or blow in insulation against the underside of the truss and was told by a friend that this may cause moisture problems with condensation along the underside of the decking. I have been studying in an attempt to understand how to best and most economically insulate my attic space. Performance of heating and cooling is priority and if it is more economical to insulate the underside of the roof decking then that leaves me a lot of useable and easily accessible storage space. I am also trying to understand the best way to insulate my wall spaces but that is another subject. I appreciate all of your advice.
The study of vented vs non vented attics that you referred to me from Florida seems to be inconclusive and dependent upon location, humidity, high summer time temperatures and low winter temperatures. It seems that the impetus is toward open cell foam spray insulation but the data from what I can tell is not noticeably in favor of one over the other with the exception of roof color.
The FSEC literature survey was primarily looking at advantages/disadvantages of vented vs unvented approaches, but the hourly attic temp and roof temp profiles of the vented vs unvented attics tells you just how fast the attic temperature drops, and that the rate of cooling is not dependent on ventilation. The attic temp peak lags the daily outdoor air peak by an hour or so, and the peak ceiling heat flux (with UNinsulated attics) is higher when unvented, lagging the peak with vented attic by about 3 hours, but the slope of the cooling rates after the peak is comparable, which indicates that ventilation is not the primary mechanism of attic cooling. See figure 4, p.18 (p.20 in PDF pagination.)
Insulating at the roof deck is always more expensive than insulating at the attic floor, and yes, in an unventilated attic if you put fiber insulation against the roof deck you are at significant risk of moisture problems in the roof deck. If you have no air handlers or ducts in the attic there is no great reason to go there. The cheapest way to go would be air-sealing the ceiling plane and installing 18"-20" of cellulose on the attic floor, provided you have the vertical depth for that much fluff.
If there isn't that much space and you want to go unvented, going by the book, in zone 5B you would need 40% of the total R to be low vapor permeance foam on the exterior of any fiber layer, if it's all under the roof deck, or 40% of the total R on the exterior of the roof deck, with 60% of the R as fiber (or open cell foam) directly on the underside of the roof deck. At a code-min R49 the IRC prescriptive is R20 above the fiber layer, or 41%. That gets to be pretty expensive unless you're using reclaimed rigid foam above the roof deck, and is still far more expensive than 20" of cellulose at the attic floor.
The roof deck can still be pretty safe if you cheat the book a bit using 1-2" of closed cell polyuretahnefoam on the underside of the roof deck with the rest as fiber directly below the closed cell foam.. This would still be a potential problem for wintertime moisture accumulation at the foam/fiber boundary. See Table 3 on p.11 of this document:
http://buildingscience.com/file/5809/download?token=18Y6NJQ8
Look at the intersection of 5B Boulder, and the columns for 1" ccSPF + spray fiberglass, and the 2" ccSPF + spray fiberglass. With 2" of ccSPF the roof deck is fine even with light colored "cool roof" metal roofing, and with asphalt shingles it's fine with 1" foam. The modeled total R was R38, and the 2" foam is R12, which is 31.5% of the total R, and 1" of foam would be R6, only 16%. If you keep the conditioned space at 25% relative humidity in winter that would probably still be fine, but at 40% RH it's a potential problem, and the fiber would need an interior side vapor retarder to manage that risk, preferably a "smart" vapor retarder such as 2mil nylon, that allows moisture out more rapidly than it lets it in, though a Class-II vapor retarder such as "vapor barrier latex" on gypsum could also work (albeit with a much slower drying rate.)
The dry air in your location results in fairly high diurnal outdoor temperature swings, both in winter & summer. Putting radiant barrier on the top chords or using cool roof shingles will lower the temperature of the wintertime roof deck temps, which raises the average moisture content. That would not hit problem levels if the attic is vented, but could become a problem in an unvented attic, particularly on the north side that gets little to no sun in the springtime drying season. A cool roof approach would likely increase heating energy use more than it decreases cooling season energy use, particularly with your 12:12 roof pitch, which gives it a significant convective cooling factor on the exterior.
So, how many inches of air is there between the top plates of the exterior walls and the roof deck?
The bottom cord of the trusses (which is the floor of the 1/2 story is 2x10 ans directly above the exterior wall from the top of the double plate to the underside of the roof decking is around 24" The problem area is where the interior wall space meets the underside of the roof decking that distance is only about 8". I planned to create a 45 degree angle at the wall to ceiling line to allow for ventilation and insulation if needed. I have a truss detail if that will help and there is a way to upload it.
In the area where there's only 8" of clearance you might cut'n'cobble 7" of polyiso in there to get at least R35-ish type performance for those sloped ceilings, with 18-20" of cellulose elsewhere.
If the file for the truss detail is less than 2 MB you should be able to attach it using the "File attachments" link just above the "Post" button. You have to choose the file first using the "Choose File" button that appears, then hit the "Attach" before hitting "Post".