Image Credit: akeg, from flickr.com
Image Credit: akeg, from flickr.com Figure 11 from the FSEC literature review on attic ventilation. This graph shows the shingle temperature throughout the day of a ventilated attic vs. an unventilated attic with an insulated roofline.
Image Credit: FSEC Figure 12 from the FSEC literature review on attic ventilation. This graph shows the temperature difference between the two types of roofs (ventilated and unventilated).
Image Credit: FSEC
More Building Science
One of the most common questions I get when I describe homes with insulated rooflines is, “What does that do to the shingles?” Some roofing companies have made noise about this topic, saying that if the shingles can’t conduct heat downward into the attic, the shingle lifetime will be greatly reduced.
What’s the truth about this claim? How much does shingle temperature really rise if the insulation is right at the roofline and the attic isn’t vented? Fortunately, it’s easy to find out because the Florida Solar Energy Center (FSEC) put together a comprehensive literature review of the need for attic ventilation a few years ago.
Shingle temperature on homes with insulated rooflines
In the summer of 2000, FSEC researchers studied this issue themselves. (This experiment is described starting on page 24 of their review.) They looked at the shingle temperature for a set of houses that had insulation at the flat ceiling and a vented attic and for another set of houses that had insulation along the roofline. You can see the data in Fig. 11 below.
As you can see, the temperature difference is only a few degrees. The graph in Fig. 12 below shows how the temperature difference between the two types of homes varied throughout the day.
The largest temperature difference was about 9°F, which occurred around noon. The average temperature difference through the day was about 2°F.
Shingle temperature and durability
The paper covers a lot of ground, so here’s a short list of some of the other research results on insulated rooflines, temperature, and durability.
Here’s a bit of anecdotal evidence for you. Southface has a building constructed with structural insulated panels (SIPs) that they built in 1996. The asphalt shingles have been on the insulated roofline for 17 years now. They also have a detached garage with a vented attic. The last time I was there, I saw no difference in appearance or performance of the two roofs.
To end, let me ask you ask you a question: Do you know anyone who has ever collected on a shingle warranty? The answer is probably no. Even if it’s yes, however, shingle warranties are structured so that you’re not going to collect much if you do make a claim unless you have catastrophic failure in the first couple of years. An insulated roofline should have minimal effect, and almost certainly will not result in catastrophic failure if done properly.
Allison Bailes of Decatur, Georgia, is a speaker, writer, energy consultant, RESNET-certified trainer, and the author of the Energy Vanguard Blog. You can follow him on Twitter at @EnergyVanguard.
Weekly Newsletter
Get building science and energy efficiency advice, plus special offers, in your inbox.
23 Comments
All good points
I've been telling people the same thing for years: The temperature difference in insulated roofs is minimal, shingle life affect is limited, and it's practically impossible to collect anything worthwhile on a roof warranty because of the pro-rating and limitations. Another point to bring up is that most asphalt shingle manufacturers have taken ventilation out of their installation requirements - most of them have eliminated it entirely or now list it as a recommendation. (Warranty documents are a real exciting read).
As to Southface, when I was the interim project manager on that building, I believe it was my first insulated roofline, all done with SIPS. I always found it fun to do new things instead of just doing it the way I always did it. Rarely got me into much trouble.
More Research Needed
The warranty is to protect against early and catastrophic failure which will shorten shingle live by a relatively large increment. Conversely, the issue of excess heat is about shortening shingle life by a relatively small increment. So I do not see the failure of warranties to be enforceable against the heat problem long term as being justification to dismiss the heat problem.
This is my only concern about the heat issue: Considering the full life potential of shingles applied to a ventilated “cold roof”; how much is that life reduced if the shingles are applied to a non-ventilated “hot roof”?
If the answer is a life shortening of 1%, I would not be concerned. But if the answer is 20%, I would conclude that the cost of a hot roof in terms of shingle wear is too high.
I think there needs to be more objective research into this issue. One would think that shingle manufacturers have an interest in consumers getting the most life possible out of their shingles, and thus would speak out against hot roofs if they shorten shingle life. But then again, manufacturers might not have that interest if it causes people to buy fewer shingles.
I recently asked Certainteed about their recommendations for ventilation. They did recommend ventilation, but they would not say whether it was needed to combat heat or to remove water vapor during the heating season.
Response to Ron Keagle
Ron,
Shingle color has a much more significant impact on shingle temperatures than ventilation. So if you want your shingles to be cool (out of the logical but unproven belief that cool shingles last longer than hot shingles), then choose white shingles.
That approach is much more likely to be successful than a decision to depend on ventilation to keep your shingles cool.
Response to Martin Holladay
Martin,
Is it unproven that cool shingles last longer than hot shingles? Above, it says that shingle life can be reduced by 2 years in Miami on unvented roofs compared to vented roofs. I would like to see more information on this type of life reduction, and how the information is obtained. If heat does degrade shingles, why not use a reflective color and ventilate?
Other considerations
Temperature of shingle roofs is somewhat of a straw man. There are other reasons to have proper ventilation, and I see it more as a best practice that is trying to be swept aside in favour of easier installations.
Temperature: Higher temperatures no matter the application tends to lead to a shorter lifespan. More research is certainly required to say whether is it significant for shingles. I deal with a lot of buildings day to day, and anecdotally, I would say temperature plays a fairly significant role in a shingles lifespan (Whether that temperature comes from poor ventilation, dark colouring, or exposure).
Energy efficiency: We are insulating and tightening up our buildings, so why impose an extra 5-7 degrees on the exterior side of the insulation when proper ventilation can reduce that?
Moisture control: There is an article featured in RCI that talks about vapour permeable roofing underlay, but it shows that asphalt shingles have a fairly low permeance
http://www.rci-online.org/interface/2011-12-Lstiburek,%20Karagiozis,%20Gassman.pdf
Real life: We all strive for less air leakage, more moisture control, and greater insulation but we are still dealing with a very labour intensive process, and the human factor has to be considered. Not every installation is going to be (or stay) perfect, and having ventilation as backup to eliminate moisture issues (mold) or heat issues (ice daming) to me is the definition of best practice.
Individually they are all fairly small issues, but together I think it shows ventilation is a good practice to keep. I see a lot of push to have walls with a vented space behind the rain screen, so the push to eliminate ventilation from roofs is baffling to me.
not a settled issue
There does seem to be a greater chance of early shingle failure on unvented roofs than on vented roofs, based on my personal observations of our own projects over the years (in the Boston area: 5600 HDD; 40" of rain annually; 42° latitude). Often, the corners start to curl up after 6-8 years or so and the aggregate wears thin. I've never seen it happen that quickly on a vented roof; I've seen it happen maybe 20% of the time on unvented roofs. I don't know what the cause is; it may have nothing to do with shingle temperature -- it may instead have something to do with how the assembly deals with moisture migration. I definitely think more research is in order, maybe that doesn't look at just temperature but also at moisture content. Or maybe my data set is just too small to draw any patterns from.
re
Also Boston area:
My last house 8 sided hip, osb/iso nailbase, IKO 3 tabs installed 1990, 2005 south facet 30 percent of tabs totally nude, tabs in yard. non south facets in much better shape. Standard construction roof also south facing also not failed. Re roofed with air space. This was a white shingle roof with red accents, and I saw no evidence that the red died faster, probably the opposite.
Next house, arch shingles over 1 inch iso[!] over 3 inch t&g. Cracked shingles everywhere multiple leaks. Reinsulated with airspace and reroofed. [edit 15 year old roof]
I think the question is not how many degrees hotter the roof is but at what temp is shingle life shortened?
7 degrees does not seem like much, but if one is storing water at 39 degrees [or 205 degrees] 7 is a pretty important number
Response to Ron Keagle (Comment #4)
Ron,
You wrote, "Is it unproven that cool shingles last longer than hot shingles? Above, it says that shingle life can be reduced by 2 years in Miami on unvented roofs compared to vented roofs. I would like to see more information on this type of life reduction, and how the information is obtained."
Like you, I would love to see more research on this topic. I don't think that the Miami research is definitive. But if you are convinced that cool shingles last longer than hot shingles, that is all the more reason to follow my advice and to choose white shingles. The data on the fact that shingle color matters much more than ventilation (when it comes to shingle temperatures) is much more established than the data on whether hot shingles have a shorter life than cool shingles.
The fact is, ventilation doesn't do much. But everyone agrees that dark shingles are hot shingles.
Roof ventilation and temperature
Martin,
I believe that for hot climates, the issue of heat gain would be the biggest reason to reduce shingle temperature. But for a cold climate, superinsulated house, shingle temperature should not be able to contribute much to heat gain. For the cold climate, the main issue would be shingle life, if it is an issue. And for the cold climate, I prefer the fully ventilated “cold roof” design because it helps prevent ice dams, and removes vapor.
Actually, I am not certain to what extent shingle temperature affects shingle life. Ventilation cools the roof, and it this may be observed to increase shingle life. However, it may be that that roof ventilation is prolonging shingle life by removing water vapor. Or it may be that ventilation prolongs shingle life because it removes escaping heat before that heat can warm the roof deck and melt snow. Those two issues could be solved by better air/vapor sealing or by more insulation, as opposed to ventilation. And it may also be the case that cooler shingles do not last longer. It may also be that shingles produced by different manufacturers are affected differently by temperature and/or ultra-violet radiation
I see several different questions about this subject:
1) How much does ventilation reduce shingle temperature?
2) How much does shingle color affect temperature?
3) How much does shingle temperature affect shingle life?
4) What other objectives can be met by cooler shingles?
5) If heat shortens shingle life, would the color white and ventilation result in the longest possible shingle life?
I think all of this needs more research. Color affects radiant absorption, but so does texture. Asphalt shingles have a texture that is highly conducive to radiant absorption. Shingle color white has an aesthetic impact that some people might not want. For the question of how much ventilation affects shingle temperature, ventilation needs to be better defined. A one-inch air space under the roof of a vaulted ceiling may be fine for removing water vapor, but too restrictive for optimum cooling air flow. A more generous airflow through an attic might not place the flow in contact with the bottom of the sheathing where the heat must be removed by conduction to the airflow.
I have concluded that for any new house, I want a metal roof. For whatever reasons one might want a cooler roof, I think metal is the way to go. Aside from that, the life will not be affected by temperature. But for any other temperature concerns, metal can be made more reflective than asphalt shingles because of its smooth surface without resorting to the color white.
Response to Ron Keagle
Ron,
Bill Rose and Jeff Gordon at the University of Illinois at Urbana/Champaign have extensively studied the questions that you think are unsettled. There are far more data on these issues than you acknowledge. Bill Rose has published many papers sharing his data; they can be Googled.
For a shorthand introduction, I recommend this PowerPoint presentation by Jeff Gordon: "Attic Ventilation."
I will reproduce two slides below. The slides reference the fact that some shingle manufacturers void their warranties when their shingles are installed on unvented roofs.
.
Response to Lee Everett
Attic ventilation has less effect on shingle temperature than shingle color or orientation. (See Martin's comment below, as well as the paper I wrote the article about.) Attic ventilation does help mitigate ice dams in cold climates but doesn't help much in hot climates. Attic ventilation, especially through soffit vents, helps houses burn down more quickly when wildfires come rolling through, or when the neighbor's house is on fire.
Joe Lstiburek recently wrote an article about how to handle cold climate houses in areas with wildfires: Rocks Don't Burn.
Response to Paul Eldrenkamp and Keith Gustafson
If you're seeing significant early failures on unvented roofs, then that's something that needs to be investigated. The FSEC review by Danny Parker relates temperature increase to decomposition via Arrhenius's demonstration that chemical activity doubles for every 18° F rise in temperature. Parker calculated that in the case of unvented roofs, shingle life would be reduced by 11% as a result. In his article Understanding Attic Ventilation, Joe Lstiburek referenced a paper by Cash et al. that comes up with about the same result (10%). For reductions in shingle life greater than 10%, something else must be going on. If it is indeed related to the higher temperature, then that would suggest not using shingles on unvented roofs. But we need to know more about what's going on there.
2.7 percent increase in temperature
Martin,
I am posing questions just to frame the discussion, but posing the questions does not mean than I am asserting that those questions are unsettled, as you say. I am also not disputing the fact that there is a lot of data, nor am I refusing to acknowledge data.
Regarding the statement you quoted above:
“Yet a 2.7% increase in shingle temperature voids the warranty?”
In the context of the quote, the point of the question is to ask why manufacturers void the warranty for a 2.7% increase in temperature when just the color difference can increase the temperature by 27%.
What is the answer to that question? Are the manufacturers unaware of the fact than black shingles get 27% hotter than white shingles?
How would a manufacturer know that failed shingles were due to them suffering a 2.7% increase in temperature? When they refer to a 2.7% increase, what temperature is that increase starting from?
Response to Ron Keagle
Ron,
Q. "Why [do] manufacturers void the warranty for a 2.7% increase in temperature when just the color difference can increase the temperature by 27%?"
A. Since the shingle manufacturers' reasoning is baseless, it's hard to speculate on their reasoning. Several cynical possibilities spring to mind -- but if I stated these cynical possibilities, I would merely be speculating.
Q. "When they [University of Illinois researchers] refer to a 2.7% increase, what temperature is that increase starting from?"
A. When I reported on Bill Rose's reasearch in the July 2002 issue of Energy Design Update, here's how I (almost) answered that question: "Rose has developed his own statistical method for comparing temperatures from roof to roof. The data were represented in a form of correlational analysis. This was done using linear regression of individual cases against the base case. The slope of the regression line was used to permit the temperature of the comparison case to be expressed as “hotter than” or “colder than” the base case as a percent. The statistical method behind these comparisons is fairly complex. The method’s most important feature is its usefulness in ranking factors in order of their effect on shingle temperature."
Percent increase in temperature?
Martin, I downloaded the paper by Gordon to find out what's behind the percent increase in temperature you quoted above. He doesn't give enough info in there to know if it's done correctly, but if a roof has a temperature of 150° F, a 2.7% increase would be 16° F. You can't do percent change on Fahrenheit or Celsius scales. You have to use the absolute temperature scale, Kelvin, for that. Since Bill Rose was involved in creating this metric, I'm sure it's done right because he knows his math and science.
Edited to add the following:
Just to show the difference, if you try to use 150° F to find a 2.7% increase in temperature, you'd end up with a 4° F rise. To do it correctly, you'd have to convert 150° F to Kelvins, which would be 338 K. Then 2.7% of that would be 9 K. There are 1.8 F degrees for each Kelvin, and that's where the 16° F rise comes from.
Shingle Temperature Research
Actually there is a lot of shingle temperature research out there. I did not see the FSEC paper mentioned.
FSEC-CR-1496-05 is the pub number, available on the FSEC website.
They list a 5% increase in energy cost for R-19 average sealed ceiling, and 8 percent with lousy ductwork. Both these figures assume ducts in the attic, common for FL. No ducts in the attic, efficiency concerns near 0.
Also of importance is: http://www.professionalroofing.net/archives/past/mar02/feature2.asp
This was commissioned by the National Roofing Contractors Assn. They hold the copyright.
The conclusions of this study:
The greatest influence on roof temperature is geographic location. The mean roof temperatures for Miami and Green Bay, Wis., for example, differ by 18 degrees Celsius.
The direction a roof faces has the second greatest influence on average roof temperature (in excess of 1.44 degrees Celsius in the east through south-to-west range studied, but the real difference is greater because other directions, such as north, will be cooler).
The color of roofing materials influences the mean temperature of a roof system slightly less than direction (1.45 degrees Celsius average for these parameters).
Ventilating the area under a roof deck reduces the average temperature 0.5 degrees Celsius (about one-third the influence of the direction or color and one-thirty-sixth the influence of geographic location). Even with wind assistance, ventilation reduces average roof temperature about half as much as using white rather than black shingles.
Within the ranges studied, slope has the least influence on average shingle temperature.
Shingle Temperature Research
Actually there is a lot of shingle temperature research out there. I did not see the FSEC paper mentioned.
FSEC-CR-1496-05 is the pub number, available on the FSEC website.
They list a 5% increase in energy cost for R-19 average sealed ceiling, and 8 percent with lousy ductwork. Both these figures assume ducts in the attic, common for FL. No ducts in the attic, efficiency concerns near 0.
Also of importance is: http://www.professionalroofing.net/archives/past/mar02/feature2.asp
This was commissioned by the National Roofing Contractors Assn. They hold the copyright.
The conclusions of this study:
The greatest influence on roof temperature is geographic location. The mean roof temperatures for Miami and Green Bay, Wis., for example, differ by 18 degrees Celsius.
The direction a roof faces has the second greatest influence on average roof temperature (in excess of 1.44 degrees Celsius in the east through south-to-west range studied, but the real difference is greater because other directions, such as north, will be cooler).
The color of roofing materials influences the mean temperature of a roof system slightly less than direction (1.45 degrees Celsius average for these parameters).
Ventilating the area under a roof deck reduces the average temperature 0.5 degrees Celsius (about one-third the influence of the direction or color and one-thirty-sixth the influence of geographic location). Even with wind assistance, ventilation reduces average roof temperature about half as much as using white rather than black shingles.
Within the ranges studied, slope has the least influence on average shingle temperature.
Shingle Temperature and Sealing Attics
Now that we had read all the data regarding shingle temperatures and sealed attics, let me weigh in on the real reasons to seal attics. The first deals with resistance to hurricanes and wind-driven rain. I build only hurricane resistant homes and insist upon sealing every opening in the attics. Gable vents, soffit vents and ridge vents can all ultimately fail from hurricane force winds. Even if the house stands, the damage from water intrusion can be enormous. The second reason has to do with giving your HVAC system a chance. A hot attic continues pushing heat through the ceiling way into the evening and nighttime hours. The ductwork and airhandler a constantly fighting against 130 degree plus temperatures. My attics never get over the low 80s and are frequently within a degree or two of the temperature in the conditioned space below. In hot humid climates, I see absolutely no reason to vent an attic space and several reasons to not.
Shingle life
I dunno about you guys, but I'm from Oregon, in the valley. I see moss & algae taking shingles out just as fast as heat or UV. We don't worry about silly things like ice dams. OTOH, we do have to deal with people who don't understand how water thinks.
Question: Venting Cathedral Ceiling
Regarding the power point by Jeff Gordon that Martin linked to post #10, what does the author mean by this following statement in his presentation?
“You cannot cool the upper part of a cathedral ceiling roof
with venting. If a cathedral ceiling cavity is vented top
and bottom, and faces south, then air moves through the
cavity driven by buoyancy. Air enters at the soffit. As it
moves up the vent slot it becomes heated and exits the
ridge as heated air. If a cavity has a slot but no vents,
the air in the slot will have a certain temperature. The air
in the vented slot will reach that same temperature after,
say, 10 feet. The remaining higher part of the slot will
receive no cooling effect from the moving air.”
Response to Ron Keagle
The way I read that is that whether you have venting or not in a cathedral ceiling, the temperature at the top will be the same. The cooler air drawn in for venting won't be cool anymore after moving up the roofline toward the ridge vent. He says that it'll be the same temperature as air in there with no venting. I guess that's one of the things they've measured at the University of Illinois.
Response to Allison Bailes
Allison,
Just to be clear, I assume Jeff Gordon is referring only to air within the ventilation slot or chute inside of the insulation cavity; and not referring to air near the ceiling in the living space.
In this part of Jeff Gordon’s statement, he is perfectly describing the soffit inlet-to-ridge outlet of convention-driven venting of a cathedral ceiling:
“If a cathedral ceiling cavity is vented top
and bottom, and faces south, then air moves through the
cavity driven by buoyancy. Air enters at the soffit. As it
moves up the vent slot it becomes heated and exits the
ridge as heated air.”
So I do not understand why he precedes that perfect description of roof venting with this sentence:
“You cannot cool the upper part of a cathedral ceiling roof
with venting. “
But the most perplexing part is where he concludes by saying this:
“If a cavity has a slot but no vents,
the air in the slot will have a certain temperature. The air
in the vented slot will reach that same temperature after,
say, 10 feet. The remaining higher part of the slot will
receive no cooling effect from the moving air.”
He seems to be comparing an air chute with an inlet and outlet to an air chute with no inlet and outlet. Then he concludes that the former will behave like the latter in the portion of the chute above 10 feet from the bottom.
Why would that be the case? Why would a chute with an inlet and outlet act like it has no inlet and outlet?
Shingle tempature research
1st of all.....separate fiberglass shingles from the organic paper based. In most areas of the country this is a given. In the Upper Midwest, not so much
#2: The south side does take the biggest beating, either from more direct sunlight or the UV rays.
I vote for both!
#3: I've been telling my customers for years that there are very valid, good reasons for ventilation done right. (Along with insulation. They are NOT Exclusive). Many times, it is an economic benefit.
The least valid reason is extending shingle life. The shingle warranties are written to protect the manufacturers. Since the vast majority of homes and buildings have inadequate insulation and ventilation, it is an out for them.
(I just settled a warranty claim w/ Certainteed. 3 yr. old roof. They paid Stormchaser rates to replace)
This is a case where research will verify common sense.
Log in or create an account to post a comment.
Sign up Log in