Insulate a cathedral ceiling efficiently
![](https://secure.gravatar.com/avatar/436c903093eae2803ba4f5f6db192d52?s=52&d=https%3A%2F%2Fimages.greenbuildingadvisor.com%2Fapp%2Favatars%2F200x200%2Fj.png&r=g)
I just bought a house in Northern NH, zone 6 that was built in 1987. The house is about 3300sq-ft on 2 floors, so about 1600 sq-ft foot-print. The main living/top floor is entirely cathedral ceiling.
There are 2 layers of asphalt shingles on the house, and the second layer is failing with shingles brittle and starting to curl. It is particularly bad around the masonry chimney, which is probably due to poor ventilation around it. Water Leaks are starting in the ceiling, as well as ghosting on the walls. (Ghosting is where you can see dirt collect on the walls where studs are, you can “see” the skeleton of the house) I figure this means the roof is inadequately insulated and was done to 1987 standards, probably 8″ of fiberglass bats with foil face and a 1″ air gap, but this is conjecture. Framing appears to be 2×10 rafters and 2×6 walls. I walked on the roof before the winter and the deck seems solid with no soft spots.
My problem is, how best to fix it? I could throw another asphalt roof at it as is and just replace it every 15 years or so, but I would like to fix the root cause.
I see two options.
1. remove the ceiling, all 2000+ sq-ft of it, and blow in closed cell insulation and refinish.
2. Insulate above the deck using a product like raycore (or similar) to fully insulate the roof from above. re-deck, and re-roof. I would then have to correct the sophist and ridge vents.
I have read most of the articles on this sight that I think pertain to my dilemma, but I am looking for suggestions that are cost effective and environmentally friendly.
I am also considering a metal roof, but the insulation still needs to be fixed.
Thank you for your time!
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
Both of those solutions are expensive and less-than green. RayCore uses HFC245fa (a very powerful greenhouse gas, 1000x CO2) in their panel systems. Most closed cell spray foam has the same issue, though there are now a few that use HFO1234ze (<5x co2), but it's still a lot of polymer per r.
How deep are the rafters? Sounds like 2x10s? (9.25")
Try this potential solution, which has two components, and does not involve gutting the ceiling:
* Dense pack cellulose over the existing batts blowing from the exterior. The combined center cavity R of a 2x10 rafter bay with compressed batts and some cellulose will come in around R34-R35. In US climate zone 6 you need to have at least 50% of the total R on the exterior of the sheathing to have adequate dew point control. So that would take part 2:
** 7" of rigid polyisocyanurate above the roof deck. The outer couple of inches could be a nail-base panel (OSB or plywood laminated onto polyiso), or it could be an OSB or plywood nailer deck, through-screwed to the structural roof deck. One vendor of nailbase polyiso is Hunter:
https://www.hunterpanels.com/product-documents/hpanels/speciality-products/108-h-shield-nb/file
If you use their fastener spacing and specs it's pretty much the same doing it without nailbase, but with a separate nailer deck. See the documentation links on this page:
https://www.hunterpanels.com/polyiso-roof-products/h-shield-nb
To take the sting out of the wallet, in New England there are several vendors of reclaimed roofing foam heavily discounted (like 2/3-3/4 off the price of virgin-stock foam). A few of them advertise here:
https://nh.craigslist.org/search/sss?query=rigid+insulation
Mind you, the 7" of roofing foam by itself would get you to code-min performance on a U-factor basis, but it's worth filling the vent channels with air-retardent fluff such as cellulose to avoid it becoming a thermal bypass for the above-deck foam.
Polyiso is blown primarily with pentane (~7x CO2) but with used foam that hardly matters, and is generally MUCH greener, since the environmental hit has already been taken, and re-use is just piling on to the benefit side of the cost /benefit balance.
With a metal roof purlins through-screwed to the structural roof can replace the nailer deck.
I have seen many discussions on this forum about spray foam and raycore, so your answer makes sense and is not unexpected, and I really appreciate the other options.
I was concerned about leaving that 1" gap in the current roof. I guess that the dense cellulose would fill that. Never occurred to me to just blow over it from the outside and leave the batts. What can be done with the sophit vents?
I read the article on dew points on GBA, which was where I originally got the idea of using raycore panels. I will look at the hunter panel. I am a little concerned about adding 7" of insulation above the roof and potentially changing the roof-line/look. I guess I need to get over that, right?
They upped the minimum R value for roofs in my area to r-48/r-49 I think. Maybe R-50, in any event, its higher than it used to be.
I do have a follow up question, I read that polyiso decreased its performance when it got cold. Isn't using polyiso on the cold side of the roof going to be less than the advertised R values?
Can you clarify how your have added 7" above the deck? I see that hunter panels are 4". Would you suggest a 3" polyiso layer with a layer of 4" hunter panel applied over it?
Two layers of 3" reclaimed foam and a 1.5" nailbase panel would give you 7" of foam (7.5" total stackup, including the OSB nailbase facer) and would be a LOT cheaper than 4" nailbase. Read the installation documentation for the nailbase- this is a legitimate stackup. IRC code min is R49 if the insulation is between rafters or joists, but it can also meet code on U-factor basis if it's U-factor is under U0.026, which is R38.5 "whole assembly". Roofing foam is typically labeled R5.7/inch so at 7" you'd already be there with ~R40 of continuous insulation, even before counting the R-value of the roof deck, or nailer, the roofing, air films, etc.
The amount of derating for temperature necessary isn't that terrible when it's more than half the (labeled) R value. Except at higher elevations even the coldest month of the year has an average winter outdoor temp higher than 10F in your area. In Berlin NH it's more like 20F (15F in January):
https://weatherspark.com/m/26442/1/Average-Weather-in-January-in-Berlin-New-Hampshire-United-States#Sections-Temperature
Assuming the warm side of the foam is 40F (which is roughly your design goal from a dew point control perspective), the cold side is averaging 20F over the 12 coldest weeks of winter, it means the mean temp through the foam is ~30F. At that temperature generic polyiso is delivering ~R4.6/inch. See the curves in Figure 2 and Figure 3 in this bit o' marketing fluff from Dow,which is really a worst-case curve for unknown recalimed foam:
http://msdssearch.dow.com/PublishedLiteratureDOWCOM/dh_098a/0901b8038098a015.pdf?filepath=styrofoam/pdfs/noreg/179-00263.pdf&fromPage=GetDoc
7" of R4.6 foam gives you R32, but that's for the coldest weeks of winter as temperatures rise in late winter it's performance rises quickly to R5/inch+ which begins the "drying season" earlier than it would if it were a constant R-value. It's enough, despite being slightly less than 50% of the total R during the coldest weeks. If it feels too close to the margin for you, go with 8" of polyiso (x R4.6= R37).
On my daily commute I pass a pair of houses under construction that are clearly insulated at the rafters, with ~10-11" facia boards at the rafter ends. These are otherwise conventional New England-y two story houses with simple high pitch gabled roofs. It doesn't look nearly as goofy as you might think. (Maybe I should take some pictures?). An 8" facia board would be even more tame.
The mystery of how they were going to heat these off-the-gas-grid houses became apparent a few weeks ago when a pair of mini-split compressors appeared at each house, mounted on stands with the bottom pans 18-24" off the pad, which is a reasonable height for the anticipated snowpacks in that location (Sudbury, MA.)
Good info, thanks for the analysis.
My fascia would be 17 1/2", with the 2x10 rafters plus the 8" of above deck insulation.
James,
The short answer to your question is: Install rigid foam above the existing roof sheathing.
Dana has given you good advice. If, for whatever reason, you don't want to install any cellulose in the existing rafter bays, you should be able to seal up the soffit vents from the exterior of your house by temporarily removing the soffits. Seal the ventilation openings with a two-component spray foam kit. Similar work can be performed at the ridge.
Here is a link to a relevant article: How to Install Rigid Foam On Top of Roof Sheathing.
Below I will post a Building Science Corporation illustration showing one way to disguise the appearance of thick rigid foam installed above roof sheathing -- so that the fascia doesn't look fat.
.
Awesome!
I did this system in 2016 in southern white mtn. NH area with spectacular results.
The house was cold and plagued with ice dams.....no longer thanks to Dana advice.
Caution: regarding densepacking over the fiberglass......if i was not there to inspect the guys doing the work the system would have failed. I doubled the quoted price gave them my dense pack hose and had them do it proper.......count the empty bags and do some coverage calculations.
I followed up with solitex mento on original roof deck, 6" reclaimed foam ($500), 2x4 sleepers, 1/2" plywood, asphalt shinges.
ZERO ice dams. Warm house.
Hope that helps someone.