Posted By jonr on 07/18/2009 3:35 PM
I haven't seen any comments on the Tyvek Attic Wrap system - it's much more than a radiant barrier - it creates a ventilation duct between the rafters.
Any difference in insulation or other side temperature of the main roof surface as compared to the eves can lead to ice damns. That's why my attic is staying vented.
Ice dams are usually a heat leak problem (either from air migration or too low an R-value.) The solution isn't always to ventilate- the leaking heat away before it can melt the snow- it's better to stop the heat leak. (Ice dams themselves can be a good way to figure out where you have a heat-leak problem.) Where possible, the better solution is to air-seal and insulate.
Ridge venting is usually blocked by snowpack when there's any decent amount of snow, so soffit/ridge venting only slows ice dams from forming when there 's scant snow to melt & freeze in the first place. Ventilated roof decks can promote drying of condensation or leaked water (which is good) but aren't great insurance against ice-dams. The air-gap of the Tyvek system probably adds up to an effective R-6, with VERY low air flow to keep the deck below freezing from heat leaking up from the house. (But the same miniscule air flow adds a path for waremer outdoor air to heat up and melt the snowpack from below on freeze/thaw days.)
The roof/attic R-value required to stop ice dams from forming depends on the anticipated snow load. Snow behaves as insulation, so the R-value of the insulation relative to that of the snow needs to ensure that the temperature of the roof deck stays below freezing over the average daily temperature.
On southern aspect eaves there can be solar heated air rising up the wall to collect under the eaves melting the snow from below only to freeze later sometimes creating an ice dam. In combination with heat-leaks up in insufficiently insulated exterior studwall this too can sometimes create ice dams where otherwise the risk would be low. If the wall is well insulated southern wall solar eave-heating usually just causes a melt-gap between the roofing and the snow, with no ice dam.
Sometimes wind-washed ridges thin out the snow layer causing a solar-heating & melting of the upper portion of the roof, but that usually results in a thin sheet of re-frozen water under the snowpack, rarely and ice-dam proper. But insulating even a small above the roof deck, including over the eaves (important to avoid localized re-freezing right at the exterior wall boundary rather than an extended sheet) will mitigate that too. Typically 2" of EPS or XPS above the roof deck (even over the eaves) is
sufficient to all but guarantee ice dams won't form in my climate (New England
~7000HDD, up to 2' of rooftop snowpack typical, rarely more.). Usually just an inch of XPS or ISO evens the roof temps out enough to do the trick. Clearly YMMV.
Building Science Corp has some generic R-value recommendations for treating ice damming problems both with/without attic/roof-deck ventilation:
http://www.buildingscience.com/documents/digests/bsd-135-ice-dams/?full_view=1
Note they're only talking about an additional R-5 for the unvented vs. vented roof deck. That's 1" of XPS. (I'm a bit surprised they didn't mention the insulated roof-deck over the eaves approach. Several roofing contractors in my area have been doing that for years with excellent results.)