Foam Insulation Requires roof Ventilation

Hey Guys,

I read an Article in the March edition of JLC titled “An Ice Dam Analyzed” by Jeffrey Hoffman, P.E.

I am amazed to find out that he recommends ridge and soffit venting for a foam insulation application. I’ve always told my past clients that ventilation is not required with a foam insulation application. Now this professional engineer has me confused.

I scanned the article in .PDF because JLC requires a membership to view the entire article on-line and it’s too big for an iNACHI upload.

Please give me your feedback…

I wonder if they used ice and water barrier, he did not say

I bet the insulation contractor goes back for some continuing education…

The manufacturer of the foam specifically recommends soffit and ridge vents with polystyrene vent chutes under the roof sheathing before the foam application for that area.

I agree with the ventilation needs and never understood how the spray foam could stop the heat transfer from within the stucture at the thicknesses being blown.

David, could you email the article to me, please?


I used to live in the Soo man I miss the snow, and shoveling and blowing and chipping… nah never mind.
I do miss the skiing:D

There are some new techniques being implemented and tested within structures to allow for, believe it or not, 100% sealing of the thermal envelope. This process normally brings the attic in to the thermal envelope (where applicable). There would still be exhaust for combustibles, dryer vents, stove vent, etc, etc. No mold issues, no moisture issues, etc.

We are currently working with an organization on this process and will be releasing several case studies and white papers on the subject. I will post a link up here once all is finished.


Here is a different opinion from a different P.E. and from my experience with the Polyurethane Spray Foam for this application, it is a cat’s meow.
3" is usually applied between the rafters and a 3" batt insulation is added with nylon straps to hold it up, this is to provide the fire resistant value to maintain the 450/25 fire qualities required to meet code.

Used it more than once and works. Air exchangers are sometimes installed in conjuction with this design because of the air tightness.

Nearly every roof application needs ventilation. I have installed over 1000 residentual applications in 30 years and all attic space needs to breath.

IKO is starting to add a warranty to unvented roof systems. The wording is somewhat confusing but after I talked to the New England rep. she told me that they are looking at the science behind bringing the attic into the thermal barrier of the house therefore warranting the shingles for ten years. Time will tell???

[FONT=Times New Roman]2. The limited warranty terms set out in this document apply to shingles installed on roof slopes of 4 in 12 (1:3) and steeper. The limited[/FONT]
[FONT=Times New Roman]warranty period for shingles installed on roof slopes of less than 4 in 12 (1:3) and down to 2 in 12 (1:6) is 12 years (If certain application procedures
are followed, see shingle product wrapper for details), and will be pro-rated for material only, exclusive of labor, etc., at an annual reduction of 8.33%.
(If certain application procedures are followed, a full limited warranty will be in effect for slopes between 3:12 to 4:12. See shingle product wrapper
for details). Armourshake, Cambridge LT and Crowne Slate will only be warranted when applied to roof slopes 4 in 12 (1:3) or steeper. Use of
an approved underlayment is strongly recommended beneath all shingle applications, below 6:12.
3. All shingles which contain a factory applied self sealing strip must be subjected to direct sunlight for several days before full sealing will
occur. Shingles installed in the fall or winter may not seal until the following spring. Shingles which do not receive direct sunlight or which are not
exposed to adequate surface temperatures may never seal. Damage to the factory applied self sealing strip by dust, sand, or foreign matter will prevent
the sealing strip from activating. This is the nature of shingles and failure to seal down under such circumstances is not a manufacturing defect.
IKO will not be responsible for any blow-offs or wind damage that may occur prior to thermal sealing having occurred. After the shingles have sealed,
however, they will be covered under this warranty if they experience blow-offs or wind damage in winds up to the levels listed in the Limited Wind
Resistance Warranty section.
4. IKO does not provide any warranty coverage to its shingles when they are installed over insulated roof deck panels.
5. IKO will not be liable for damage caused by insufficient attic ventilation or roof drainage. Ventilation must meet building code requirements.
Also, IKO is not liable for damage caused to products installed over a roof deck that is not properly ventilated with an air space between the
roof deck and the insulation. Approximately half of the ventilation must be at the eaves, with the remaining half at the top, or ridge of the roof.
6. Exclusions 4 & 5 notwithstanding any fiberglass asphalt shingles manufactured to meet ASTM D3462 applied to roof deck assemblies
(slopes > 2:12) where foam insulation is prefabricated into the roof deck system (often called “nailboard insulation”), or where insulation is installed
immediately beneath an acceptable roof deck system, or where radiant barriers are installed, with or without ventilation, directly below the deck, shall
qualify for a reduced limited warranty duration of 10 (ten) years with no Iron Clad Protection coverage, with respect to shingle problems related to
the absence of adequate roof system ventilation (linear pro-ration over 10 (ten) years). The roof design professional is responsible to ensure; 1) the
proper quality and correct application of the insulation and/or radiant barrier, 2) the provision of adequate structural ventilation and/or vapor retarders

*as they determine to be necessary, and 3) that all local codes are met (especially considering local climate conditions). *
[/FONT][FONT=Times New Roman]These important factors and[/FONT]

[FONT=Times New Roman]*decisions are *[/FONT]

[FONT=Times New Roman]not [/FONT][FONT=Times New Roman]the responsibility of IKO but are critical to assure proper deck system performance.[/FONT]

[FONT=Times New Roman]7. In all cases, replacement shingles are warranted only for the remainder of the original shingles’ limited warranty.[/FONT]
[FONT=Times New Roman]8. IKO reserves the right to discontinue or modify any of its products, including the color blend of any shingles, without notice to the original
consumer-purchaser. IKO will not be liable as a result of any such modification or discontinuance.
9. IKO will have no liability in the event that any replacement material for any shingles covered under this limited warranty may vary in color
in comparison to the original product as a result of product changes or normal weathering of the original product.
10. This limited warranty is effective as of October 1, 2009.
11. This limited warranty only confers rights upon, and may only be enforced by the original purchaser of the shingles that are covered by this
limited warranty, or any person to whom this limited warranty is transferred in accord with the TRANSFERABILITY set out above. No other person
has rights under, or may enforce this limited warranty
12. Subject to the laws of any jurisdiction that would modify this term, no action may be brought against IKO for breach of this limited warranty
or arising out of any use of or damage to the shingles that are the subject of this warranty unless it is brought within one year after any cause of action
has accrued.
Some provinces do not allow the exclusion or limitation of incidental or consequential damages, so the above limitations may not apply to you.
14. The unenforceability of any term of this limited warranty will not affect the enforceability of any other term. If any provision of this limited
warranty is unenforceable, it may be severed. In that event the remainder of this limited warranty will remain in full force and effect.
15. The installation of asphalt shingles on dimensional lumber (including shiplap or board decks) is not recommended as it may potentially
cause buckling problems. Buckling is not covered by this limited warranty.
16. With Renaissance GXL 30 shingles, the base shingle has “patches” of asphalt coating and granules which help give the shingle its unique
appearance. During exposure on the roof it is possible for these patches to show small fissures. These fissures can develop within the first few years
depending on roof exposure, climate, roof pitch, and attic ventilation efficiency and are not the result of a manufacturing defect. They do not detract
from the long term water shedding performance of the asphalt shingle roof system. (Beneath the patches, there is the shingle’s base layer of asphalt
and granules, and the glass reinforcement. Below this is the overlapped unexposed portion of the underlying shingle.)
17. Product bearing “ECONOMY NO WARRANTY” tape, is not covered by this limited warranty, and is sold [/FONT]

[FONT=Times New Roman]"[/FONT][FONT=Times New Roman]As Is.[/FONT][FONT=Times New Roman]"[/FONT]

[FONT=Times New Roman]18. Open metal valleys are recommended for best roof performance. If roof valleys are constructed using shingles, the shingles in the valley[/FONT]
[FONT=Times New Roman]are not covered by this Limited Warranty.

[FONT=Times New Roman][size=4]IKOIndRegCDA10-09PD10-[/size][/FONT]

Note: the first article is about cathedral ceilings.

Here is an article that was in the March 2007 issue of JLC


I have been building for 45 years and used SPF back in the 70’s on insulation for Potatoe barns. Protecting potatoes in the 6 month winter months is a science in itself.
No ventilation, control moisture.

Read the article.

By applying spray foam directly to the underside of the roof deck, it now insulates the attic space from the extreme heat that once radiated through the hot shingles sheathing and roof. The severe temperatures no longer exist in the attic.

I recently met with an ELK Premium Building Products rep at a trade show last year, and they claimed their shingles have no problems whatsoever being used over an un-vented roof deck. So much for curling shingles…

By insulating the roof surface with spray foam, the surface temperature exposed to the attic (the temperature driving the heat transfer) is reduced by as much as 40F.

Further information is available from ASHRAE (8700-527-4723) in a publication titled “Vented and Sealed Attics in Hot Climates”.

In a roof application, spray foam will increase the structure’s ability to handle high winds as well as bring the attic into the conditioned space. A roof application of spray foam will reduce infiltration and reduce ceiling heat transfer and duct losses. Wall and floor applications will also create better thermal and air barriers, and make better use of engineered products. Spray foam insulation can result in less conductive, convective and radiant heat transfer, lower infiltration rates, less duct losses, a more structurally sound building and can result in significantly smaller-sized heating and cooling systems and better comfort levels for the occupants.

Unvented Attics

State of the Art
The vast majority of homes are heated and cooled by ducts installed in vented attics. In hot, humid climates, in some homes, moisture in attic ventilation air can condense on the out side of inadequately insulated ductwork and/or the attic side of the unevenly insulated ceiling drywall when it is overcooled by misdirected supply air at register boots or other leak sites. This has caused moldy, soft or failed ceilings in a few homes. Attic ventilation air creates the thermal load on the duct system itself during both heating and cooling seasons. In addition vented attics are dusty and can have pests and pesticides which can contribute to poor indoor air quality in the home.
Unvented attic construction, where the insulation is applied to the underside of the roof decking can solve the moisture and IAQ problems of vented attics. Dr. Joseph Lstiburek and his colleagues at Building Science Corporation have conducted significant work in this area (Ueno, 2003). In addition, since the ductwork is now within the thermal boundary, energy savings also result (Parker et. al., 2000).
Research Needs
The unvented attic concept has been implemented in several dozen new homes in hot, humid climates. Those with tile or metal roofs seem to have no problems. Unvented attics with shingle roofs have had some problems in some homes (high moisture content in attic peaks, buckled shingles). These are generally a result of solar driven moisture into the attic as has been suggested and described by Lstiburek (2003) who has recommended the use of a vapor retarder roofing paper to solve the problem. However, some roofers are reluctant to use this slippery material.
Another issue is how best to insulate at the underside of the roof decking. Many have used vapor permeable foam insulation to avoid the problems of attaching blanket or blown insulation against gravity. Some have raised the issue of detecting roof leakage points if a roof leak developed with foam insulation.
The research needs are to:
3. Unvented Attics: Conduct measurements in unvented attics with shingles to determine the attic moisture conditions as a function of roof slope, type of roofing paper, amount and type of coupling between house and attic, amount of duct leakage. Develop predictive models and design guidelines. Conduct research to determine how roof leaks are propagated through foam insulated roof decks and suggest design or application guidelines.:slight_smile:

I understand everyone’s feedback and I agree with most of you.

This particular PE is stating that a foam insulation applied to the underside of roof sheathing will conduct any attic heat build-up right to the outer side of the roofing material, which causes shingles to deteriorate.

That’s where I get confused. I had no idea that attic heat build-up will conduct its temperature right through a foam application.

I was always educated into thinking that foam will not conduct heat and acts as a thermal barrier. I need to get re-educated on foam applications.

I did a final draw on a 4500 sq ft last summer, no power 92 deg and muggy outside, it was 74 in the home.

It is a common misconception that insulation prevents freezing (in this case, melting). Insulation can not “stop” heat transfer. It just slows things down.

If applied correctly, by the time the heat reaches the other side the weather conditions change (sun goes down…) and the heat moves back in the other direction.

ie. insulating water pipes in the attic will not stop them from freezing if water never moves (fire suppression) and replaces the heat source of the water. It may save you by extending the freeze time and hopefully the temperature comes back up above freezing by then.

Insulation does not lower the freezing temp of the water in the pipe, it only changes the rate of conductance.

The issue in this application is that there is not enough room over a conventional cathedral ceiling for the amount of insulation needed for the geographic zone (7).

Adding air ventilation between the insulation and the roof deck will also reduce the effective R-Value and the ice dam problem may not go away. Thus more insulation is needed above the roof deck. This ice dam issue could have occurred anyway, even if they vented the roof. He stated that there was less than R-38 when the bays were full. It would have dropped even more if ventilation was added and the conductive break may not have helped in the restricted space.

Foam has better performance (R/in) and seals air leaks better and eliminates potential convective heat transfer.

If the air (ventilation) baffle is not a significant air barrier between the insulation and the air space below the roof deck, huge losses could occur from air convection through the insulation (even open cell foam).

While conducting an evaluation on the effects of radiant heat transfer I came across some interesting observations. I was comparing the rate of temperature change between wet and dry insulation in a sealed container. When I opened the wet container to allow evaporative cooling, this is what I observed:

Air convection (and it’s direction) can be seen through the container and at the surface of the insulation.

There is a visible 50% division of air movement through the insulation (top of the container).
The right (warmer) side is air leaving the container through convection.
The blue side is cooler from evaporative cooling and air entering the container to replace the warmer air leaving.

You will also notice a “thermal by-pass” in the handle of the container where insulation is missing (A cool spot on the insulation directly above). Cool air rising? Yes, it is a stack effect!

Jim Bushard should like this one as it shows the insulation as an air filter!

Most important, it shows that air does in fact directionally move even in the insulation within even the smallest of spaces without the significant stack effect that a building produces.

Speaking of Jim, you can’t do this with a blower door Jim! :wink:

Excellent example Dave, Thank you.
I’m still confused on why JLC would publish an article about non vented roof decks and then print an article about the complete opposite??? Things haven’t changed that much in 3 years.

Are they the same type roof/attic config?

Duh… odd all the “Attic” talk… :wink:

These materials are not new but some of the applications are and time will tell. The attic if foam insulation is applied to seal out all air really needs to be an actively conditioned space. If not I would anticipate the development of temperature striated air layers and the potential for condensation problems. The choice of open or closed cell foams is debated as well in different aplications. I would rely on Dr. Joe quoted in Marcel’s post as the expert on this and other thermal dynamic topics he is very well respected and even entertaining in the lectures I have attended that he has given.

Yes, both articles refer to attics as well as cathedral ceilings.