Here we go again: how to insulate a new cathedral ceiling in an old 1890 house?
QUESTION: What are my options for insulating a cathedral ceiling in an old 1890 house where original flat attic was ventilated by gable vents, roof has no soffits and no soffit vents and no ridge vents?
WHAT: The home was built in 1890, is near Worcester, MA, climate zone 5. The attic is already finished, includes rooms, dormers, an octagon turret with open porch, and is original to the house. Sometime after construction, but when framing still used real dimensional lumber, a room was made into a kitchen, the dormer turned into a bathroom, and a entrance door added. ⇒ The unfinished space above the attic apartment is another 9’ tall and 18’ wide at the base (inside dimensions). We want to finish this space by moving the ceiling to under the rafters and (a) creating a 26’ x 30’ area loft; (b) opening up a 26’ x 16’ section of attic ceiling over the living room to create a high ceiling. The unfinished space also contains the trunk and ductwork for a hydronic-heated furnace and we can now bring those ducts into the new conditioned space.
* I realize this type of question has been asked many times on GBA, but all the different advice is hard to sort through (vented vs unvented roofs, foam vs loose cellulose vs dense-pack cellulose vs mineral wool vs fiberglass, vapor retarder vs no vapor retarder, vapor-permeable interior walls vs impermeable walls, air movement vs vapor movement) .
* I apologize in advance for the lengthy detail but it may reduce later back-and-forth questioning …
DETAILS:
Roof = redone about 12 years ago
asymmetrical
intersecting gable roofs and valleys
12″ x 12″ pitch
asphalt shingles over roofing paper over plywood over original 1″x 6” plank sheathing
metal edge along lower roof edge, copper (What is this called?)
real 2″ x 8″ rafters, spaced 24”-on-center
Attic = joists 2” x 4.5”, can be supported by walls below
cellulose insulation, loose blown-in, 12” – 18” deep
ventilated by 4” x 24” vents on opposing gables, 46’ apart
knob-and-tube electrical wires removed, replaced with non-metallic cable
mechanical systems = hydronic furnace heating trunk, ducts which branch out across joists to one register above each room below
Apartment = rafters become walls
horsehair plaster, wood lath
knee walls 4’ 8” high, insulated from behind
walls insulated with stuffed-in mineral wool.
ceiling (attic floor) penetrations = electrical boxes for ceiling-mounted lights, bathroom fan, heating duct registers, trunk from air handler, brick chimney
mechanical system = air handler in closet, has trunk with rises up and penetrates ceiling, then ducts branch out across ceiling to registers above each room
RESTRICTIONS:
The house is deemed historic by the town so we have to petition town to modify the exterior if changes are very noticeable. We are allowed to paint however we like.
Prefer DIY solutions because of low cash flow, but can save up funds and hire professionals if best route.
Weren’t old houses designed to breathe, be leaky?
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Joel,
Here is a link to the article which you need to read. (It's OK to read it twice if you need to -- keep reading it until you understand it): "How to Build an Insulated Cathedral Ceiling."
You wrote that the roof you need to insulate has "intersecting gable roofs and valleys." That simplifies things -- the valleys mean that a vented roof is impossible. You need an unvented roof assembly.
There are just two choices:
1. Install an adequate thickness of rigid foam above the roof sheathing (with or without additional fluffy insulation under the roof sheathing). This option requires you to install a second layer of roof sheathing above the rigid foam, as well as new roofing.
2. Install an adequate thickness of closed-cell spray foam on the underside of the roof sheathing (with or without additional fluffy insulation under the cured spray foam).
Details are in the article I linked to.
Thanks Martin for the quick reply. I'm a little leary of spray foam because of stories about mis-application, bad mixing, off-gassing, flammability. And I cannot rip off roof shingles and add another roof layer. So what if I try to vent the attic? ... by not raising the ceiling so high, flattening the ceiling 18" below the bottom of the ridge board, insulating bottom with foam board (which limits use of any type of flammable foam), installing smaller gable vents. There are no true soffits, but the fascia nailed to the rafter ends has cracks where wood meets wood and lets in outside air. The existing mineral wool is dark on the edges next to this rafter-end fascia.
I noticed that a lot of houses in this neighborhood have attic units with windows on gable walls, louvered vents above these windows just below the ridge, and no visible soffits. Assuming the ceiling for these units is a few feet above the top of the windows, those gable vents would open to a long and narrow channel.
There's no way to really vent a roof deck that's all cut up with valleys hips and dormers. Air has to come in at the bottom edge and has to have clear path to exit, on EACH rafter bay. With the local snowfall potential the little cap vents through the roof deck really don't cut it, nor do cracks in the facia boards. More often than not cap vents become heat-leak points triggering ice dams.
Worcester MA has hit the top 3 ranking in the Golden Snow Globe awards multiple times, but only #6 this past year: http://goldensnowglobe.com/current-top-10-snowiest-cities - it's really no joke. If you can't vent the cathedralized ceiling roof soffit-to-ridge (which lowers the ice-damming potential) you're better off going unvented in this climate.
If re-doing the whole upstairs it's possible to micro-zone the rooms room-by-room with local thin profile hydronic air coils running off a condensing boiler loop or condensing water heater instead of huge leaky ducts, which would allow you to get rid of the hydro-air ducts, seal up the supply & returns, and only run plumbing, which is much less lossy ,easier to insulate, etc. Local hydro coils (sized to the room loads) don't pressurize/depressurize the rooms, eliminating air-handler driven air infiltration. It's often cheaper to implement than low temperature radiators, but if the rooms ar insulated you might be able to get there with flat panel rads or low-profile cast iron radiators/baseboard etc too. What do you currently have as the heat source for you your hydro-air aka "...hydronic-heated furnace..."?
Thanks for informative reply Dana.
(1) So I guess our current attic was able to keep moisture damage down because of sheer air volume above the blown in cellulose in the space? The main roof, which has narrow 4" x 24" vents on either end is over 10' high (top of ridge board to bottom of ceiling joist), 26' wide and 46' long. There are two other lower roofs perpendicular to the main roof, sort-of like a cross;. And two dormers.
(2) The heat source is natural gas boiler in the basement (unfortunately not a condensing boiler; we converted a Utica oil burning boiler to natural gas with a Carlin EZ-Gas burner retrofit, 80-ish percent efficient).
(3) Yes, we could replace the heated air with hydronic baseboard, but the ducts give us option of ducting in cold air, too, right ... though I am not sure how shared duct system would work, especially since better to have heating ducts on floor and cold-air ducts on ceiling. My plumber will be happy if we change to baseboard hydronic heat!
(4) I wonder how the houses around us insulate their tiny attics above their attic living spacces? One neighbor has a 3500sf house (all the houses are large around here) and has no insulation at all, he says, because he still has knob-and-tube wiring.
With gable vents and free air motion through a large attic space and the insulation at the floor it can purge moisture just fine with air flows between the gables. The venting air volumes become far more restricted when you start building walls and moving the insulation layer up to the roof assembly.
Any indication on the burner (or boiler) what the BTU-rate is? Is the boiler heating your domestic hot water too? If yes, is the potable water heated with an embedded tankless coil, or is it an indirect fired tank?
The air handler probably has BTU specs at 180F and 140F entering water temps too.
Is it operated as one heating zone for the whole house?
In a multi-story building it's hard enough to run just the heating as single zone of ducted air and get satisfactory results, but using it for air conditioning too never works well. As a rule the ratio of cooling/heating load on the upper floors is proportionally larger than the ratio for lower floors. Many 2 story homes in Worcester could be cooled by a single 1-ton or 1.25 mini-split on the upper floor at the top of the staircase, letting convection bring the cool dry air down the first floor.
Before diving in, it's worth running a whole-house heat load calculation on the house & system in it's current condition as a baseline, which would also tell you your oversizing factor. With some 5th grade math you can use the boiler as a measuring instrument using fuel-use (wintertime only gas bills) and the approximate steady state efficiency of the boiler. Use +5F as the outside design temperature (it's the 99th percentile temperature bin for that location.) If you're at the higher elevations in Paxton or Holden it's reasonable to use +2F or even 0F, but not lower. For details on how fuel use load calculations work, see:
https://www.greenbuildingadvisor.com/blogs/dept/guest-blogs/out-old-new
I've seen a number of houses in Worcester with finished attics and NO attic insulation, but I've also seen them insulated unvented with open-cell foam between the rafters, sometimes with vapor barrier paint on the ceiling gypsum, sometimes not. Even without interior side vapor retarders plank sheathing seems to hold up better than OSB or plywood, even though it's technically a code violation.
It would be safer for the roof deck to go with an inch or two of closed cell foam, with the rest of the cavity filled with fiber insulation. It's safer if the closed cell foam is at least 40% of the total insulation layer. In many houses that age it's common to have full-dimension 2x6 rafters. With 2" of closed cell (R12) and 4" of compressed rock wool or HD fiberglass batt (~R17) comes in a bit over 40% foam (but still only R29 at center cavity, a bit more half the current code-minimum but WAY better than nothing.)
For 8" full dimension rafters at 2" the foam's vapor retardency protects the roof deck, but with a full 6" of fluff you're looking at R25-ish fiber, which means the foam is only 32% or so. In those cases you could go 3" on the foam to boost the ratio, or stay at 2" and use a "smart" vapor retarder or "vapor barrier latex" primer on the ceiling. In the latter solution the vapor retardency of the foam protects the roof deck, the vapor retarder limits the amount of moisture accumulation in the fiber insulation.
With rafters shallower than 6" it's worth adding Bonfigiloni strips to the rafter edges to deepen the cavity by a couple of inches, and still go with 2" of closed cell foam at the roof deck:
https://www.finehomebuilding.com/2014/11/26/breaking-the-thermal-bridge
Not that it matters, but a handful of years ago I was involved with a deep energy retrofit on an 1890s house in Worcester-proper (near Holy Cross University) that involved 6-7" of reclaimed polyiso on the roof plus spray foam in the rafter bays. It was a fairly simple roof line- a gabled roof with one large shed dormer on the northeast side, and one gabled dormer on the southwest side, making it pretty easy to use rectangles of sheet foam for most of it. Cutting the compound angles foam into valleys or hips adds a bit to the labor cost of going that route, as well as the scrap rate on the foam board. This house had only two valleys, no hips. With the built-out upper floor/attic it was divided into three separate apartments by floor, each heated & cooled with a single ductless mini-split head in open living/dining areas, with the insulated but not directly heated basement shared by all three for the water heaters & laundry/storage.
Joel,
It sounds from your responses -- I may be wrong -- that you are talking about insulating two different assemblies:
1. Cathedral ceilings cut up by roof valleys.
2. A small attic directly under the ridge of the roof.
If all you were concerned about was (2) -- a small attic directly under the ridge of a roof -- a variety of solutions might be possible, including vented solutions.
The tough part is (1) -- the cathedral ceilings that are cut up by roof valleys. These sections of your roof need to be insulated using an unvented approach. That means spray foam on the underside of the roof sheathing, or rigid foam above the roof sheathing.
Martin,
(1) I am referring to same roof. Now that I demo'd part of the ceiling, pushed aside the blown-in cellulose, and really looked around, I (finally) get it when you said that complex roofs with hips and valleys will need an unvented roof if the ceiling is "cathedralized" ... because many of the rafter bays never reach the building perimeter, so cannot have a soffit or external vent.
(2) My interest in "small attic directly under the ridge of the roof" was because I was wondering if I could achieve a raised ceiling and still get adequate ventilation (by gable vents) by just raising the ceiling flat? Of course, the square area of the flat attic floor would be reduced and more rafter bays would become sloping "sides" of the conditioned apartment. These sloping "sides" are already filled with a mix of cellulose and mineral wool.
(a) If I raise the ceiling, how should I insulate the new sloping "sides"?
Joel,
You can raise the ceiling if you want, as long your insulation approaches follow recommendations (and as long as the collar ties that you intend to raise don't have a structural function -- many do).
Will gable vents be adequate for a small attic above a horizontal ceiling? Maybe. The smaller the attic, the less likely that the gable vents will keep you out of trouble. If your ceiling is airtight, there is less chance that you'll have moisture problems in your attic.
Q. "If I raise the ceiling, how should I insulate the new sloping 'sides'?"
A. To read about all of the different ways you can insulate sloped roof assemblies, see this article: "How to Build an Insulated Cathedral Ceiling."
Martin,
(1) According to your "how to Build an Insulated Cathedral Ceiling" article, whoever (a long time ago) insulated the sloping areas of existing "walls" of my attic apartment did go against the advice of not insulating the unvented rafter areas. These rafters in the sloping areas, the knee wall space, and the floor joist space next to the knee walls were stuffed with a mix of old mineral wool and cellulose. The under-roof sheathing is 6-inch boards and they look good. What may have helped lessen any moisture buildup in the sloping rafter area is that there was (a) a large space behind the knee wall and below the stuffed sloping area; and (b) the large gable-vented flat attic of blown cellulose above the sloping area.
I wonder how these houses insulated their sloping attic roofs?
(My web browser crashed during upload of images.)
(My web browser crashed during upload of images!)
... and this ...
FWIW, I insulated the attic of my 1890's house in NJ (Climate Zone 4/5) about 7 years ago. It has a complex enough roof that we went with the unvented option. 8" rafters stuffed with fiberglass, plywood structural deck above that, Ice & Water shield on the plywood (my builder insisted, and it turned out well), then 4" of rigid foam, 1/2" OSB, felt paper and fiberglass shingles. We used spray foam at the eaves and gable floor and ceiling to do our air sealing, but that's it for SPF. Blower door testing found fist-sized holes in the spray foam that we addressed prior to drywall.
I installed a temperature/humidity datalogger under the inner layer of roof sheathing (the potential condensing plane) for a year. The RH never went above 60%. The new attic room is one of the most comfortable in the house, year 'round. Not cheap, but it really works.
It is also (as it turns out) pretty rugged. When hurricane Irene came through, we lost some of the flashing from the upper roof. It was wide open to about a foot of wind-driven rain. Between the staggered layers of foam and the Ice&Water shield, we never had any leaks inside - all the water drained out the eaves. I spent a week helping friends recover and didn't even know I had holes in the roof until I climbed up there to take a look. This is kind of a big deal. When Sandy came through a year later, we had fastened the flashings so we had no damage at all. And with power off for 2 weeks in December, an R-60 roof is not a bad thing.
Sometimes it's hard to value doing things right, but I'm pretty happy we did.
Thank you for telling your personal experience on a similiarly-aged house. I would probably go something like your solution -- installing foam on exterior roof sheathing, fluffy insulation on the interior rafter bays -- but I dont have the funds nor desire to redo our 12-year old roof at this time. On the other hand, I am leery of polyurethane spray foam on the interior (off-gassing, poor curing, poor application, flammability). I may just restuff the insulation back into the knee wall space and leave the ceiling flat and abandon raising the ceiling to use the additional 9 feet of space above as a loft. By demolishing part of the ceiling, I did see some structural details that will guide us into reinforcing the framing in the future if and when we make the loft area.
As you can see from the above photos, we installed fiberglass bats R13 in the exterior walls. My helper did this while I was out of town. The wood-lath plaster walls were originally stuffed with blown-in cellulose and/or mineral wool . I am considering removing the fiberglass bats and installing dense-pack cellulose now that I have learned about the advantages of hygroscopic cellulose and that kind of insulation was what has worked in the walls in the past. One difference is that I will drywall the walls since I need something more forgiving than plaster.
??? Thought?
??? Waste of money?
??? If we install dense-pack cellulose in the walls, we may install with Insulweb netting and then tack up the drywall. Should I install a smart vapor barrier or just leave cellulose in direct contact with the drywall?
??? Can we drywall with mold-resistant paper drywall? What about fiberglass-faced drywall? I will have to check the vapor permeability ratings of these different types of drywall.
The exterior is aluminum siding over clapboard over 1" x 6" board sheathing;. I am not sure if there is any house wrap or paper over the sheathing or under the alumimum siding. We are planning on removing the siding a few years from now and restoring the clapboard (well, it is painted in lead paint, so will remove and replace.)