Ice dams and Underlayment

Where ice dams may form along the eave because snow continually freezes and thaws or frozen slush backs up in the gutters, the underlayment application in the area of the eaves is modified to prevent ice dams from forcing water under the roofing, which could damage ceilings, walls and insulation.

Two layers of underlayment should be cemented together with asphalt cement from the lowest edge of the roof and continue up the roof to a point that is at least 24 inches inside the interior wall line of the building as show in the illustration.

The environment within the envelope of the building provides adequate warmth to prevent ice dams from forming above the heated space; therefore, the two layers of cemented underlayment are permitted to terminate 24 inches inside the interior wall line of the building. There are exceptions to the rule.


You’re not understanding ice dam formation!!! It’s exactly that heat which you claim “prevents ice dams from forming above the heated space” (wrong concept) that produces the melt water that freezes at the cold eaves.

Thank you Ben, I think that is a good illustration, sums it up nicely.

Brian, you are only partly correct. We’ve had this conversation approx 5 months ago. I suggest you go back and read the entire thread. I won’t bother re-hashing the subject again… it’s too dang hot out! Ironic, isn’t it?! :wink:

The statement has no place in building science of a truly efficient structure. It essentially says “Your heat loss from the poorly insulated/airsealed ceiling is not a problem since it prevents ice damming over the heated parts of your house”. It doesn’t say “This heat loss is costing you heating $$$$$ and is probably the main cause of the ice dams/leakage at your eaves/exterior walls.”

Heat loss at snow covered skylights on lower slope roofs has caused ice dams on the roofs below the skylight curb…not at the eave but within a foot or two or three of the lower edge of the curb, some times 10-20 feet up from the cold eave.

BTW…How many roofers have taken the time to cement together 2 layers of tarpaper? We never did in our 10 years of roofing. The time saving better solution is using “eave protection” such as WR Grace “Ice and Water Shield” and similar.

That statement is basically sharing a quote from the IRC commentary.
To have an ice dam, you need BOTH melted water and solid ice (the dam). IRC believes that once you’re past 24 inches, you will not have both. Beyond the 24 inch mark, the warm of the building will prevent both from simultaneously occurring and creating an ice dam.

It’s a bassackward quote if they are talking about ceiling heat loss saving the upper levels of the roof from ice damming!! Except for cases of winter solar attic heating, ceiling heat loss is the driving force of the ice melting to begin with…why comment on it as if it’s a positive???

BTW… If IRC refers to Canada’s group, I have had a few little “discussions” with the odd one of them about being secluded in the research field…get out and see what the home owner/buyer/builder wants…don’t do research in the absence of the real world and consumer demand!

I had one refuse to answer my question in an open forum at an energy conference about 10 years ago saying he didn’t have time!!! The question put him in a position of having to defend energy savings versus costs of an energy conservation improvement he had researched and was recommending.

What Ben is leaving out is that code is a “basic minimum” protective requirement. The IRC does not mention the intent behind the 24 inch “minimum” requirement and would not object to double underlayment going back to the peak of the roof if the roofer wished to exceed these requirements…nor is there any claim that the 24 inch “minimum” requirement will ensure against damage from ice damming.

Perhaps the recommendation was originally 60" and a political compromise was made with the NAHB.

The wording of a code does not necessarily reflect building science.

In my humble (or not so humble) opinion, Brian is correct. The reason ice dams happen at the eave is because where the wall and the roof meet is a weak point in insulation so that is the place where there is a concentrated heat loss. This heat loss is what melts the snow on the roof (which is only 8 or so inches away.) Heat loss doesn’t prevent ice dams, it causes them. The ice and water shield doesn’t prevent ice dams but it keeps the water that creeps under the shingles because of the ice dam, from reaching the sheathing. If you have trusses with a high heel, you would be able to do a better job of insulating the wall to ceiling joint + the heat loss would have a better chance to get vented away before it would melt the snow on the roof.
(I think that the cementing together of two layers of roofing felt is a recommendation of the time before ice and water shield or similar products.)

So what Bert and Brian are both agreeing on, is that if the home were to be 100% efficient and 100% heat loss free, then the home would have *zero *ice dams???

Give me a friggin break!!!

I don’t care how well and efficient an attic can be on heat loss and leakage, around here, you will have ice dams on most low emmisivity roof coverings.
Ice and water shield on the roof is your best second line of defense against damage.
It also protects the roof sheathing from the elements while pending the final roof coverings.
I have never heard of asphalting to layers of felt before nor seen it in 40 years.
Too many other products available that are more economical and quicker to use. :slight_smile:

Around here, ice dams are common and not always caused my heat loss. (although it is the #1 reason) Mostly in older homes with little to no insulation and/or venting. Its only smart to go at least 36" with Ice and Water shield, if not 72". Occasionally, the entire roof is covered with it, not just the eaves and valleys. I hope I dont ever have to take that stuff off, I hear sometimes its easier to replace the sheeting. (thats another discussion)

Good article Peter. Nice to here from you.
Hope you are keeping busy down there. :slight_smile:

I’ll give you a friggin break:)
Now that that is out of the way; have you ever seen ice dams on unheated buildings?

Can you clarify Marcel? I would have thought it to be "on most high emmisivity roof coverings.


Emissivity is a measure of the thermal emittance of a surface. Emissivity is defined as the ratio of radiant heat flux emitted by a material to that emitted by a blackbody radiator at the same temperature. Emissivity values range between 0 and 1. Emissivity may be thought of as the ability of a material to emit heat (via infrared radiation) to the surrounding atmosphere. Higher numbers indicate faster heat transfer. There are standardized methods to measure the thermal emittance of roofing materials like ASTM C1371 and ASTM E408.

While solar reflectance is the most important roof characteristic for energy savings in warmer months, emissivity can also contribute to a cool roof. In warm and sunny climates highly emissive roof products can help reduce the cooling load on the building by releasing the remaining heat absorbed from the sun.
There is also evidence that low emissivity may benefit those buildings located in colder climates by retaining heat and reducing the heating load.
Remember that emissivity and reflectivity both play a part in the surface temperature of a roof.

There are a substantial amount of black and dark surface roof coverings in this area, and during the winter months with temperatures in the mid twenties, these low reflectance materials contribute to the ice damming affects.

Hope this makes sense. :slight_smile:

Not claimed!!

From my posts #5:
“This heat loss is costing you heating $$$$$ and is** probably the main cause** of the ice dams/leakage at your eaves/exterior walls.”

and #7:
Except for cases of winter solar attic heating, ceiling heat loss is the driving force of the ice melting to begin with”

And in the ceiling heat loss, I include air warm air leakage from the house below.

There is a very rare situation that may lead to snow melt and ice damming with deep snow loading on roofs; this is not caused by house heat or direct solar heating of the attic as it can happen on dull, cloudy days also.

The conditions: the outdoor temps go above freezing for a few hours during the day and then go back to freezing at night:

-The outdoor air enters the attic and warms the roof boards (low R value) causing snow to melt and run to the eaves. As the outdoor air cools back to below freezing later , by wind or stack effect, the cold (now freezing) air enters the eave/soffit venting to first cool the snow covered eave here and freeze any moisture above the shingles at this location- Voila- beginning ice dam formation.

Of course, on any given day, any combination of the 3 ice dam formation mechanisms may be at work.

From the Canadian Roofing Contractors’ Association, here’s a quote from a 2003 Technical Paper titled “Ventilation”:

"Although the primary contributor to snow melting is heat loss from the building interior, solar radiation can also provide sufficient heat to melt snow on a roof. For example, at Ottawa, enough sunlight can be transmitted through 150 mm (6 inches) of snow cover on a clear and sunny day **to cause melting at the roof surface when the outside temperature is -10°C (14F) with an attic temperature of -5°C (23F)."

Note: When Ben talked about the IRC, I was thinking in Canadian, assuming he meant the Institute for Research in Construction…our federal gov’t building reasearch group.

**YES! **Every year. That was the (unspoken) 100% no heat loss in my prior post.

Heat that melts snow and ice does not *only *come from the structures interior.

Yes that does make sense, hence my question; you stated earlier that the low emmisivity roof coverings contributed to ice dams. I asked if you meant high emmisivity (or low reflectance.) Now I see that is what you meant.
Thanks you.