Historic home roof insulation solution

[Dan McGonigle]

Hello! I’m in a bit of quandary here in regards to how best to approach insulating the roof of my 1842 brick home.

Simply put, we have a third story that we don’t use. We don’t use it because it is in a state of disrepair. It has been on the project list for years and my wife and I have decided to get serious about it this year. I have plans to renovate it, not as a living space but as a clean and comfortable area for storage and space for my wife’s various crafts. The current roof system is as follows: typical a frame, rafters, purlins, original cedar shakes nailed directly to purlins, standing seam metal roof applied directly over original cedar shakes. No roof insulation. I’m not sure when the metal roof was installed, but I have to assume it’s old. No ridge vent, no soffit vents, no roof box vents, no gable vents, no venting whatsoever. How can I insulate this? I’ve read plenty of things online regarding this issue, I’ve spoke with my local building inspector, a residential architect, and the cedar shake bureau technical rep. The idea of not installing vents and applying closed cell foam to the underside of the original singles (possibly with a contact barrier in between) is debated. Some say you can, others say you can’t. My options are to explore this route (waiting to hear back from a reputable insulation contractor) or to have a ridge/soffit system installed and use fiberglass batts.

That comes to my second quandary. Current code in my area (NJ climate 4a) states roofs must be insulated to R-49. This can be achieved with fiberglass batts at 14” thick. As far as I know, my 180 year old roof rafters are 4-6” thick. There’s no way that could work. And I seriously do not want to extend my rafters by 10 inches - it would have a huge spacial and visual impact on the third story (sloped walls/knee walls and chimneys coming through). Even with closed cell foam, using the max allowable thickness of 4-5 inches (I believe) I would fall quite short. I’m looking at the NJ rehab sub code which I believe states an exception in my case - stating that if recommended R value cannot be achieved due to space constraints, fill the available cavities. Which I interpret as put as much insulation as you can in there. But, unfortunately the inspector may interpret it differently and state I “could” achieve this by undertaking more construction and increasing the depth of my rafter bays.

Any thoughts at all would be welcome!

[Ron Selzer]

Years ago I subscribed to New England journal of Light Construction (https://www.jlconline.com/magazine/) they always had a lot of good information and seemed to be on the cutting edge. Would be worth a visit there to see what you can find. Then s much as i hate to do it it would be time to get an architect and the building inspection dept. involved or decide to do it with any inspections.
Good Luck
Ron

[Jim Matthews]

I wouldn't spray foam.

Rigid insulation board overlay can be serviced, or removed more readily. Some accomodations will be required to hide the increased depth at the gutter and gable ends.

https://www.buildingscience.com/documents/insights/bsi-063-over-roofing

[Jon Nuckles]

My understanding, based on my architect, contractor, and the approval of plans for our remodel a couple of years ago, is that spraying closed cell foam completely prevents vapor penetration and you don't need to vent on the underside of the roof. I sure hope they were right!

[Ron Citerone]

Sub code sounds like it is worth running by the inspector.

[roger wiegand]

If it were me I'd go the unvented roof route (as we have in our current house and shop), fill the bays with closed cell foam, than 2" of rigid foam on the inside to stop thermal bridging, with blueboard and plaster to create your interior finished surface. (someone sells a product with blueboard glued to rigid foam for a all-in-one solution at minimal thickness.

An unvented roof can be problematic in tropical climates, shortening the life of roofing materials. I'm aware of no other downsides except for the inevitable "you can't tell where the roof is leaking". We've had one roof leak in 60 years, and it was pretty easy to find due to the branch sticking through the hole. I don't think this is a real world concern, at least if you maintain your roof.

At my son's house with a very low slope roof and no way to access the space to spray foam it needed a new roof anyway, so we put 8" of recycled polyiso rigid foam from commercial roofing on top of the existing roof, built up the edges, and applied a rubber roof over it. It has worked out extremely well.

My impression is that 90% of the naysaying on moving the insulating layer up to the rafters and going unvented comes from the guys who sell fiberglass and blown in cellulose. For us it got the HVAC equipment in the attic inside the conditioned envelope which was a huge win.

FWIW, our inspector was thrilled we were doing anything to upgrade the insulation in both houses. Code does not require you to do the impossible in an old structure.

In the future, when it is time to replace the roof you can do a couple of additional improvements, add more insulation above, and/or add a new layer of sheathing set up 3/4" off the old roof with soffit and gable vents to keep your new roof cool-- this will certainly help with summer AC loads.

[Jim Becker]

IF the roof deck wasn't old cedar shakes and was plywood or solid boards laid out as a closed deck, I wouldn't hesitate to use closed cell spray foam because it solves almost all of your issues...you can get to R-49, you get a full vapor barrier, you get a much stiffer structure, you do not need venting because inside the spray foam is considered conditioned space, etc. It also wouldn't interfere with anything exterior that has to be maintained for historical preservation if that's a factor, either. The same would be true if you had the underside of the steel on the purlins as the spray foam eliminates condensation issues on the underside of the steel.

BUT...you don't have that situation because of how the roof was previously renovated. The cedar shakes should have been removed prior to installing the steel for both functional and appearance reasons. (steel doesn't like uneven surfaces under it and can project the same to the top surface) So the best solution would be to retrofit with a ridge vent and vented soffit, have baffles in the bays for air flow, fir out the roof joists to accommodate high density/high R-Value fiberglass batts and if that doesn't get you to your required R-49, use foam insulation board cross ways over the augmented joists to get the R-value and tape the joints for air infiltration and vapor before doing drywall or other coverings.

That said, Roger is also correct that you may be relieved of being fully code compliant as a retrofit rather than new construction, and with furring out the joists you should be able to get an R-38 with the best material available in, say, 12" of space. All the rest applies, however...you do need to do the venting or you're going to have moisture issues.

[Andrew Hughes]

I would reach out to a local Roofing company. I say this because I’m a roofer and always felt like inspectors and architects were sometimes full of baloney.
Since I’m a west coast roofer I have no experience with east coast residential practices. But I’ve done commercial roofs that were supposed to be good for snow and ice dams conditions.
Dens glass sheeting with a very good peel and stick membrane.
Good Luck

[Steve Demuth]

I would absolutely go with sprayed in closed cell foam. If there are concerns about the purlin-shake confection against the foam, layer in 1" closed cell board between the rafters firmly attached to the purlins, and then cover that with foam.

My roof is a crazy concoction of non-continuous roof boards originally put down for wooden shingles and plywood, all overlain with the actual roof. Ten years ago we cleaned up the inside, extended the rafters and blew in 10" of closed cell foam before drywalling the entire mess to make a cathedral ceiling. The roof is completely unvented, as are the gable ends down to the loft floor. We have had absolutely zero moisture issues and the closed cell foam made the roof (which was probably structurally borderline before) so strong you could drive a truck on it. It's like one giant torsion box with molecular-scale webbing.

Edited to Add: I also have second house build a couple of years ago. It's 100% sprayed-in closed cell foam insulated, and is, while not airtight, sufficiently close that moisture condensation can be a problem on the colder inside surfaces (windows, e.g.). We have external ventilation with heat recovery unit that typically runs a 10% duty cycle to compensate.

[Mike Soaper]

You might consider contacting Ron Herman for a consult, he's a 3rd generation master housewright that has also done period work on historic houses.

I've enjoyed several "Woodworking show" seminars by Ron, here's some links, hope it helps.

Antiquity Builders of Ohio (http://www.antiquitybuildersofohio.com/)

Ron Herman - The Woodworking Shows (https://www.thewoodworkingshows.com/instructors/ron-herman/)

[Malcolm Schweizer]

My home was built in 1836. When I pulled the tin to replace it, I laid down treated ply over the original 1” pine and then put rigid foam insulation. The difference was amazing. The attic during mid-day is barely warmer than outside.

I have done a lot of research on spray foam insulation and if I were not reskinning the roof, I would go with that. It is in itself a moisture barrier. As stated by others, it also adds some structural rigidity.

[Tom M King]

Why not insulate at the ceiling joist level. We found a 1980 standing seam roof over an 1851 Cypress Shingle roof. I wouldn't spray foam against the wooden shingles. Not that it matters, but I doubt they're Cedar. Not sure about what was used up there in 1842, but around here everyone that doesn't know better also calls them Cedar Shakes.

"Shakes" is probably a 20th Century term. Back then, they were all made by hand, including splitting, but they were called shingles.

Here's a link to the page on my website about that roof we found. You can see the old roof in the first three pictures, which include one edge of the metal roof on top of the shingles. http://historic-house-restoration.com/CypressShingles.html All those pictures are thumbnails. You can click on one, and get a full sized version.

Historic Associations don't like spray foam, but they're fine with the way we do it by wrapping a vapor barrier over, and between ceiling joists, and blowing in insulation. The blown in stuff can be removed later without damaging any of the original parts.

It was not unusual for these old wooden shingle roofs to last 150 years. These days, they've engineered the life out of them by interlacing rows of tar paper. If one is put on according to specifications by the Cedar Shingle and Shake Bureau, they typically last about 35 years. The old ones would have lasted longer than that, had the fasteners not started to fail. The reason there were enough buildings left standing to save in the restoration of Colonial Williamsburg was because they still had the original Cypress shingle roofs on them.

[Jim Becker]

I would absolutely go with sprayed in closed cell foam. If there are concerns about the purlin-shake confection against the foam, layer in 1" closed cell board between the rafters firmly attached to the purlins, and then cover that with foam.
That's an excellent point. It's not quite as good as foam directly to the roof because there's going to be a small air gap "just because" between the insulation and the roofing materials that can hold moisture, but it's still going to be a whole lot more efficient than other insulation materials.

[Dan McGonigle]

IF the roof deck wasn't old cedar shakes and was plywood or solid boards laid out as a closed deck, I wouldn't hesitate to use closed cell spray foam because it solves almost all of your issues...you can get to R-49, you get a full vapor barrier, you get a much stiffer structure, you do not need venting because inside the spray foam is considered conditioned space, etc. It also wouldn't interfere with anything exterior that has to be maintained for historical preservation if that's a factor, either. The same would be true if you had the underside of the steel on the purlins as the spray foam eliminates condensation issues on the underside of the steel.

BUT...you don't have that situation because of how the roof was previously renovated. The cedar shakes should have been removed prior to installing the steel for both functional and appearance reasons. (steel doesn't like uneven surfaces under it and can project the same to the top surface) So the best solution would be to retrofit with a ridge vent and vented soffit, have baffles in the bays for air flow, fir out the roof joists to accommodate high density/high R-Value fiberglass batts and if that doesn't get you to your required R-49, use foam insulation board cross ways over the augmented joists to get the R-value and tape the joints for air infiltration and vapor before doing drywall or other coverings.

That said, Roger is also correct that you may be relieved of being fully code compliant as a retrofit rather than new construction, and with furring out the joists you should be able to get an R-38 with the best material available in, say, 12" of space. All the rest applies, however...you do need to do the venting or you're going to have moisture issues.

I had read that closed cell foam has a maximum thickness spec. Such that it cannot be installed in greater thickness than “x” (I believe it varies, maybe based on climate?). Did I read that wrong?

Yes, historic preservation is a huge factor - not only for me but for my towns historic preservation committee.

I think that if I furred the rafters out more than a few inches. I’d lose so much space, it almost wouldn’t be worth it. Personally, I don’t care about not achieving R-49. Obviously I want it to be reasonably climate controlled but the historic preservation of my home is far more important to me than energy efficiency. Maybe I’m a bit crazy

The more I think about ridge/soffit the more I warm up to the idea. Although I’d hate to see my old roof with a big ridge cover on the top, mixing new with old. Same goes with the soffit vents. That also allows for an easier roof job if ever needed. If down the road this roof gives out, my hearts set on cedar shingles.

[Dan McGonigle]

I would reach out to a local Roofing company. I say this because I’m a roofer and always felt like inspectors and architects were sometimes full of baloney.
Since I’m a west coast roofer I have no experience with east coast residential practices. But I’ve done commercial roofs that were supposed to be good for snow and ice dams conditions.
Dens glass sheeting with a very good peel and stick membrane.
Good Luck

Thanks! Consulting with a few roofing companies is on my to-do list

[Dan McGonigle]

Why not insulate at the ceiling joist level. We found a 1980 standing seam roof over an 1851 Cypress Shingle roof. I wouldn't spray foam against the wooden shingles. Not that it matters, but I doubt they're Cedar. Not sure about what was used up there in 1842, but around here everyone that doesn't know better also calls them Cedar Shakes.

"Shakes" is probably a 20th Century term. Back then, they were all made by hand, including splitting, but they were called shingles.

Here's a link to the page on my website about that roof we found. You can see the old roof in the first three pictures, which include one edge of the metal roof on top of the shingles. http://historic-house-restoration.com/CypressShingles.html All those pictures are thumbnails. You can click on one, and get a full sized version.

Historic Associations don't like spray foam, but they're fine with the way we do it by wrapping a vapor barrier over, and between ceiling joists, and blowing in insulation. The blown in stuff can be removed later without damaging any of the original parts.

It was not unusual for these old wooden shingle roofs to last 150 years. These days, they've engineered the life out of them by interlacing rows of tar paper. If one is put on according to specifications by the Cedar Shingle and Shake Bureau, they typically last about 35 years. The old ones would have lasted longer than that, had the fasteners not started to fail. The reason there were enough buildings left standing to save in the restoration of Colonial Williamsburg was because they still had the original Cypress shingle roofs on them.

Tom, would you mind elaborating a bit on this method? Do you mean essentially insulating the ceiling of the third story/attic, and the leaving the sloped side walls without? Or enveloping the “room” space from above the ceiling joints and down along the sides with a vapor barrier, then filling the space between the vapor barrier and backside of drywall or plaster board with loose cellulose?

The issue of moisture and roof venting comes from the conditioned spaced being exposed to the unconditioned space at the roof sheathing, correct? Moisture laden air permeating the plaster and insulation and then condensing on the cold roof sheathing. So if the conditioned spaced was completely sealed from enveloping from above with a air-impermeable barrier, the roof can stay invented. Is that right?

I try my best to make nothing irreversible. I love the idea of it being easily removed in the future. Even the thought of spraying synthetic foam against 180 year old beams makes me cringe.

Great website by the way and thanks for the clarification regarding shingles

[Tom M King]

Not getting a mental picture of "sloped sidewalls". Is there space between the ceiling joists, and rafters being used for living space? I'm not sure if my Insulation page makes anything clearer, or not. It's been years since I've worked on that website. Just a few days ago, I found hundreds of old pictures on old hard drives. I may have some pictures that can clarify.

[Tom M King]

There was a similar ventilation issue with that 1828 house on the Insulation page. Here's a picture that shows the gable end vents I added. There was no ventilation to start with, but also no insulation. The fourth floor is used in that house. There are knee walls under the rafters, and a short, flat ceiling below that gable vent.

[Dan McGonigle]

Not getting a mental picture of "sloped sidewalls". Is there space between the ceiling joists, and rafters being used for living space? I'm not sure if my Insulation page makes anything clearer, or not. It's been years since I've worked on that website. Just a few days ago, I found hundreds of old pictures on old hard drives. I may have some pictures that can clarify.

sorry that was misleading. I just meant the walls, which in my case are sloped because they are against the roof rafters. Here is a photo. 449560

The third story isn’t very big, and I’d hate to lose any more cubic footage by dramatically extended the depth of the rafters, also because the chimneys are butted up against them.

Your website is great. I see you or someone on your team can perform traditional plastering - something I’ve been fascinated by. I’ve attempted to skim coat with lime and have failed miserably.

[Tom M King]

My "team" is me, and two guys that no one else would hire, that can almost read a tape measure, if they work together.

I have a method for skim coating old plaster, with nothing but finish plaster, without using Plaster Bonder. You will read everywhere else that it can't be done.

That attic level looks very similar to the house in my picture. Will be off, and on here today, without much time to spend on it.

[Jim Becker]

I had read that closed cell foam has a maximum thickness spec. Such that it cannot be installed in greater thickness than “x” (I believe it varies, maybe based on climate?). Did I read that wrong?
There's no maximum thickness relative to ability to apply it. We have 10+ inches of it in the roof of our major addition...at R 7.5 per inch, that's not too shabby and it made the attic "conditioned space". That said, there are diminishing returns with closed cell spray foam beyond 6" or so due to the cost of material and application. In some circumstances to fully fill a given area, 2-3" of closed cell is applied and the remainder of the space is filled out with fiberglass or rockwool for cost reasons. But in other situations, the spray foam is the way to go because of fit or difficulty getting other materials in, etc. Every situation has to be judges by itself. In your home, you have 6" joists and then purlins---you have nearly 8" for the foam. You have enough space to meet or get very close to the new-construction specification without adding any sticks to the rafters.

Relative to venting, someone who knows what they are doing can replace the current cap with a vented cap and it will not stick up much more or at all from what you are used to. Modern materials permit airflow while closing off the venting to critters. They will just have to trim the steel back a little more than it likely ise now and slice out the cedar parallel to the ridge line before installing the new cap. You also need to vent the soffits to complete the task and with traditional insulation methods, you'll need baffles (air channels) up each rafter bay on the underside of the roof deck to keep insulation from closing off the venting. *** If you do the spray foam, you eliminate this work and the cost may very well cover the premium cost of the spray foam while getting you much higher R-value than you can get with anything other than closed cell spray foam. ***

To your question posed to Tom about the envelop, yes, you want the entire envelope around what will be essentially living space insulated. In space like this, the "roof" is also the walls in many situations. That means you insulate the roof (ceiling/walls) as well as the end-walls of the gables, etc. If you do install knee walls, you can technically stop the roof insulation there and then come down the knee wall, but the floor (ceiling of what's below) would also need to be insulated to complete the envelop. Assuming the space is all open now, it's easier to do the envelope all the way to where the roof meets the outer limits of the space, even if you intend to install knee walls. In fact, that makes it easier to embed storage in those knee walls because there's no faddling around with insulation in the knee wall, itself. Many folks like to do cabinets and drawers in knee walls because otherwise, it's wasted, non-walkable space.

[Tom M King]

couple more pictures of up there. I also had to figure out a way to get real ductwork in the house, without being visible other than the registers, and not using the small, high velocity stuff.

[Tom M King]

The screen was so we could see what was going on when the insulation was blown in, and to let air out. Over that went the vapor barrier. The plywood had enough fasteners in it to hold it in place. The plywood was because I had to squeeze in there, and didn't want splitters in my back. We insulated the hell out of it everywhere we could. Between the rafter at the roof slope was the only place that we just had to accept what we could get. Baffle stapled to roof sheathing from soffit up to open air above the 24" of insulation on the flat part of the ceiling.

The metal shingles were already on that house when we started to work on it, and being a museum house, no place for a ridge vent. They want me to put a Cypress shingle roof on it, like it had originally, but I'm 70 years old, and have more work lined up than I can possibly ever get done. I only work within 10 miles of home. I'd like to think that I will get to it, but have other, more important things on the to-do list.

[Dan McGonigle]

Thanks. I’m getting a better idea of the system. I’m currently looking into a product called “Intello” - claiming to be a “smart” vapor barrier that allows one to use traditional batt or fill insulation without roof venting.

Again, I’m really impressed by the work you’ve done. I’d love to focus more on historic preservation. Thanks for all your input.

In hindsight, I should have never preemptively contacted my building inspector. I normally wouldn’t, but was planning on getting a small dumpster for the some of the work and I know several people in town are friendly with him. In fact, my father was the inspector for a few of the surrounding towns and knew my towns inspector very well. I hope that grants me some leeway!

[Tom Bender]

Dan, it's an attic, not an occupied space, at least for your use. Staple some fiberglass batts in the purlins and include a moisture barrier to keep moisture from the house out of the insulation. Then drywall it. Even without any heat it will stay above freezing in the coldest weather. You will want some ventilation in summer. A gable vent on the south or west end and a powered one on the north or east end that you close for winter would make it a really good space to store crafty stuff for your kids to go thru someday.

In the meantime, lath and plaster might be fun.

[Jim Becker]

With no venting, it's critical to use an insulation method that is designed for that situation to avoid what can be serious issues over time.