Opinions on this Insulation Retrofit Plan
![](https://secure.gravatar.com/avatar/90f0b268ae24c5f9ed3772bfe84de678?s=52&d=https%3A%2F%2Fimages.greenbuildingadvisor.com%2Fapp%2Favatars%2F200x200%2Fd.png&r=g)
Background:
1960s split level with 2×4 construction in Zone 6. Along with the construction of a 500 sqft grade level addition (with 500 sqft basement), I’m planning to retrofit exterior insulation to the existing walls (currently thinking 4” semirigid rockwool, given the existing polyethylene vapour barrier) and significantly improving the roof/attic insulation. The roof is a hip roof with a 4:12 slope. The truss heels are only about 4” high and there is only about 4’ to 5’ of vertical clearance in the attic at the peak. It’s a tight space. The existing roof is vented and has insulation consisting of two layers of fiberglass batt (~R24) without ventilation baffles. Some of the batts are quite dirty as the air must be flowing through them. Given the minimal amount of insulation I can provide on the inside at the eaves, I’m considering converting the vented roof to unvented and installing roof top insulation.
The Plan (also see sketch, Option 2):
– Install 6” of GPS/XPS/Polyiso (~R30) on top of the roof sheathing,
– Seal between the top of the walls and roof sheathing with spray foam,
– Carefully rip out the poly vapour barrier above drywall from above,
– Install 2 layers of overlapping ~R12 batts directly above the ceiling or blowing in about 7” of cellulose (~R24) for about R50 total effective R-value.
Questions:
1. Does this plan make sense? Am I missing anything?
2. This is going to be expensive, isn’t it?
3. Is it actually okay to have the large air space between the batt insulation and the underside of the roof deck in an unvented attic? Typically I’ve see details for cathedral ceilings where the batt is installed right against the roof sheathing. I could do this here, but it would be more difficult given the limited access, also it may be hard to get more than R12 between the 2”x4” top truss cords.
4. How easy is it to join roof top insulation panels at hips and valleys? Would you fill gaps with spray foam and perhaps reinforce the joint with tape/mesh at each layer?
5. Given that I want the addition’s roof to match the existing (shape style and elevation), could these insulation details work just as well for new construction or is there a better way?
Thanks.
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
Hi Dave,
Your plan sounds reasonable, but be sure to review Martin's article on calculating foam minimums. (See https://www.greenbuildingadvisor.com/article/calculating-the-minimum-thickness-of-rigid-foam-sheathing.)
Before you do too much, I would get a blower door test to determine how leaky your house is. Then I would air seal it and measure again. The payback on a deep energy refit (which seems what you are contemplating) can be really long. It might make more sense to air seal and add additional insulation to the attic while keeping it vented. (And maybe this approach would allow you to add solar and further reduce your energy needs.)
To answer one of your questions, you would need to install the air permeable insulation against the underside of the roof sheathing when splitting the total insulation between the outside and inside. (See here for more info: https://www.greenbuildingadvisor.com/article/combining-exterior-rigid-foam-with-fluffy-insulation.)
If you decide to move ahead with exterior foam, rockwool on the walls would allow drying to the exterior. Reclaimed rigid foam on the roof sheathing would help with your materials cost, but you still might run into difficulties hiring a roofer who is comfortable with this type of project.
Thanks. Appreciate your input.
But why does the permeable insulation need to be directly against the underside of the roof sheathing? If the whole attic space is unvented and sealed, wouldn't the air space between the insulation laying in the ceiling and the roof sheathing just be like a very large air gap like when strapping a wall. As long as the roof top insulation meets the minimum thickness to prevent condensation on the underside of the roof sheathing (like in the first link you cited) I can't really see what could go wrong. Maybe there is something that could happen, I just don't know what, the second link you cited didn't really address my question, as I think it assumes the more common situations of a cathedral ceiling or accessible attic space with deep rafters, not an inaccessible attic with shallow roof trusses.
>"If the whole attic space is unvented and sealed, wouldn't the air space between the insulation laying in the ceiling and the roof sheathing just be like a very large air gap like when strapping a wall. "
Yes- those gaps are referred to as a "thermal bypass", which is something that you should never intentionally design in to your thermal layers. With foam up top it's more reliable and more effective to install the fluffy stuff tight to the underside of the roof deck so that air can't simply find a convection path around it. Well fitted batts between the rafters can work just fine.
It's usually OK to leave any older insulation on the attic floor as-is, unless it's a substantial amount. It can't count toward the code-R due to the large thermal bypass, but that's not to say it does nothing. If there's R25 or better on the floor it will affect your roof deck temperatures and dew point control if you're also installing R20+ on the underside of the roof deck.
In addition to the low-verditude problems associated with the extreme greenhouse gas blowing agents used in XPS, the flame spread issues with polystyrene (XPS or EPS) can be a real problem in roof assemblies, since it melts, drips & flows when burning. (There are some real horror stories around fighting roof fires on EPS core SIP roofs.) Polyisocyanurate and polyurethane do not have those issues- they will burn, but they char in place, and don't drip burning liquids onto whatever happens to be below.
Interesting. Thanks for giving me the term "Thermal bypass". Will give me something to read about. Also I was unaware that polyisocyanurate performed differently in a fire. Thanks for your input Dana.
We're not big fans of XPS around these parts as it is the least green of the commonly available rigid foams. I would leave that out of the assembly. If this were my project, I'd go all-polyiso. If your concern is the "cold weather performance of polyiso" problem (which is often overrated), put a layer of EPS or GPS as the outermost layer, or just derate the polyiso to R5 per inch and don't worry about it.
I would probably spray foam the inntermost layers of rigid foam in the vallys, then leave the outermost layer cut to a tight fit so that I didn't have to do as much trimming to keep the roof deck flat. Spray foam tends to expand out more than you might think and in the worst spots, so you want to minimize the need to trim it to save on time/labor.
The air gap between insulation layers is not good. Try not have any gap. If you use a thick enough layer of exterior rigid foam, you don't need spray foam on the inside, either, you can just use batts or other fluffy types of insulation. The thickness of that "thick enough layer" is defined in some of those linked articles and depends on your climate zone.
This won't be a cheap assembly, but it will perfom pretty well. I would try to detail the interior ceiling -- and the ends of all the rafter bays -- in an airtight manner though to limit moisture ingress into the roof assembly.
Bill