Adding insulation from the exterior
![](https://secure.gravatar.com/avatar/1b0619e5416ecfea3d0069d0faa11de2?s=52&d=https%3A%2F%2Fimages.greenbuildingadvisor.com%2Fapp%2Favatars%2F200x200%2Fj.png&r=g)
Climate Zone 5 / Western Massachusetts
I am working on a project that involves a cape style house with existing 2 x 6 rafters spaced at 24″ o.c. Currently, the 48″ high knee walls and unused portion of the floor from knee wall to eave are insulated with fiberglass batts. The roof slope to the peak has been dense packed with cellulose insulation – no venting – and 2″ of XPS rigid insulation and strapping applied prior to drywall.
The house will need a new asphalt shingle roof next year. Would there be any problems adding more rigid insulation to the exterior, strapping with 2 x 4’s parallel and fastened to the existing rafters to create a vent space from drip edge vent to ridge? I would then sheath with plywood and install a new asphalt shingle roof, soffits and fascia.
I would also like to make the kneewall portion a conditioned space by continuing the cellulose/rigid to the floor and air sealing. All fiberglass will go to the dumpster.
Thanks in advance.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
![](https://www.greenbuildingadvisor.com/app/themes/greenbuildingadvisor/dist/img/modules/detail-library/thumb1.png)
![](https://www.greenbuildingadvisor.com/app/themes/greenbuildingadvisor/dist/img/modules/detail-library/thumb2.png)
Replies
Well you could give the fiberglass away (if it's clean, which is unlikely) to some poor schmo who can't afford anything else.
Other than your landfill contribution, the primary problem with your plan is enclosing the cellulose and existing roof sheathing between two relatively impermeable layers. You'll eliminate or substantially reduce drying to the outside (depending on what kind of foam board is used) when there is already very limited drying to the inside. And, by venting the roof and adding R-value, you'll be dramatically reducing the dominant drying motivator: solar radiant heat.
Continuing the thermal/air barrier plane all the way to the eaves makes perfect sense (it never makes sense to insulate kneewall areas), but you'll be far better off adding additional foam board on the inside where the primary air/vapor barrier should be, thus avoiding the sandwich problem.
I would recommend, however, that you still go ahead with the plans to strap and vent the new roof, since venting will increase the sheathing drying potential, allow the thermal layer to dry outward, eliminate ice dams, and reduce summer radiant gains.
If you decide not to vent the roof, then you might use foil-faced polyiso as the additional inside foam layer, tape the seams, foam the edges, and strap before drywalling (if you can sacrifice the headroom) to create a good air barrier, a radiant barrier and some additional "free" R-value.
Robert,
I feel that venting the roof is important, and so is additional roof insulation. I cannot add more insulation as an inside foam layer, since it is already finished and would interfere with door heights. I can add more foam in the knee wall space, since it is unfinished.
That being said, my only option for a majority of the roof is from the exterior. What about adding additional strapping as a vent space for sheathing prior to the rigid? Using polyiso instead of XPS? Thanks.
Jon,
If you vent the roof under the additional rigid insulation, then the insulation adds nothing to the thermal envelope. Roof venting must occur above the thermal envelope to keep the roof deck cold.
I would suggest adding the vent channels above the existing deck, but look for other areas (such as behind the kneewalls) to increase the R-value, thus increasing the average R-value of the assembly without creating problems. Also concentrate on air-tightness, perhaps using canned spray foam at the eaves and sealing any penetrations (eliminate can lights if present).
Robert,
To clarify; you suggest adding more insulation to the space from top of kneewall and floor and adding vent space only to the roof.
I know that walls are different from roofs in terms of solar exposure, weather and driving rain. Walls also are typically protected by an overhang. Are these the driving reasons behind not adding rigid insulation to a roof when walls are detailed on this site with rigid over sheathing with an air space and siding?
Jon,
While you many know that I'm not an advocate of "outsulation" of any kind, I take it that your walls don't have foam board both inside and out, such as you're suggesting for the roof assembly. I know of no "outsulation" advocate who promotes rigid foam on both sides of the structural envelope.
The theory of "outsulation" is that it keeps the sheathing above the dew point (though also at the mold-growth temperature), relies on drying to the inside (even though the dominant moisture drive in cold climates is to the outside), and allows outward drying through the drainscreen/rainscreen (which also, in combination with exterior insulation, dramatically reduces solar inward drying potential).
I believe there is a sound hygro-thermal engineering basis for the conclusion that exterior insulation is problematic, and an even more sound basis for concluding that a sandwich of rigid foam surrounding the structural envelope is highly problematic.
You can add all the interior foam board you'd like in the unused kneewall section of roof, thus increasing the average R-value of the roof assembly. With an well-insulated and air-tight space, there's very little air temperature stratification that would bias the heat loss toward the peak. It's the average R-value of any assembly that matters (high-R walls have low R-value doors and windows, but still perform well).