Vaulted timber framed ceiling options
I’m weighing the differences in efficiency, cost and buildability of achieving an R-60 timber framed (vaulted) roof system in a cold climate (Western Massachusetts). I believe my options to be; structural insulating panels, a combination of SIP’s and rigid insulation, a combination of stick framing, cellulose and rigid, or trusses overlayed on the frame and filled with cellulose.
Even though it is redundant structurally and a much thicker assembly, the truss situation proves to work in terms of cost (cellulose versus rigid or SIPs), buildability (quickness and overhang detailing), and insulation value (18″ = R-60).
Going this route, would you recommend:
1. Thinning the truss down and adding a 1 1/2″ layer of polyisocyanurate as a thermal break/air barrier under the roof sheathing?
2. Using an 18″ deep truss and a Zip roof system?
3. Adding Propervents to either system full height to vent the assembly?
Thanks in advance for your suggestions.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Jon,
If you go with the system you're planning, you don't need a ventilation channel above the cellulose (although some designers and builders sleep better when it's there). If you ventilate, be sure to choose a stiff panel that is able to resist the pressure of the dense-packed cellulose.
If I were going this route, I would install rubberized membrane (Ice & Water Shield) on top of the first layer of roof sheathing. That will give you a perfect air barrier, preventing any warm, moist interior air from getting into your insulation. Of course, the rubberized membrane is also a warm-in-winter vapor barrier.
Martin,
From the inside; wood frame, drywall or wood paneling, trusses, plywood or Zip Sheathing, RooftopGuard underlayment membrane, asphalt shingles. I'm confused as to your rubber membrane 'first layer' location.
Jon,
I think your plan to install drywall on top of your timber framing may be difficult. What if it starts to rain? Where will the installers stand? How do you prevent the drywall from sagging between the widely spaced timbers?
If you install wood paneling (for example, tongue-and-groove boards) as the ceiling, then the rubberized membrane would be the next layer (on top of the boards).
Martin,
As I have done previously in this situation, I have attached 5/8" spacers on the timbers and then installed the drywall after the roof and insulation are in place, fastening to the trusses. Similarly, I have used 3/4" wood paneling versus using the 2" x 6" T&G boards. I have been able to get better quality decking in this thickness versus using 2" x 6" framing material finished for decking.
Doing it this way would prohibit your rubber membrane detail.
Martin,
Thanks for the response to my truss/cellulose part of the question. I was surprised that you were not advocating for any rigid foam approach. In following most of your comments, you appear to be a strong advocate for the insulation outside of the box system, while Robert is the Larsen truss/cellulose advocate.
Jon,
I'm not sure I understand how you intend to install the drywall. If you are planning to install the drywall in sections between rafters or purlins, how do you plan to create an air barrier between the interior and the bottom of your cellulose? An air barrier is essential.
I never commented on the use of rigid foam. I merely provided advice to you about important details for your preferred insulation method. If I were installing insulation on top of the roof sheathing, I would go with rigid foam.
Your list of options did not include nailbase. If you are unfamiliar with nailbase — rigid foam laminated to OSB on one side — you might want to research the product.
Martin,
Sorry for the confusion. I was looking for advice as to systems to help with my decision. The drywall portions are limited to floors only, so we will be dealing with wood paneling fastened to the wood frame in all roof cases.
To make the air barrier situation work, I would either have to install the barrier as you describe - behind the 2 x 6 wood decking and then the 18" deep trusses filled with cellulose or go with a SIP/rigid combination. The SIP manufacturer can provide a 7 1/4" urethane, wood clad R-38 panel or a 11" EPS wood clad R-37 panel. Both would then require an additional two layers of staggered rigid to achieve the desired R-60.
I am familiar with the nailbase, but the OSB over 2 x 4 nailers fastened through the rigid makes more sense to me.
While the truss system is redundant, it is less costly for materials, much thicker, more conventional and easily frames the overhangs. The SIP/rigid system works off the timberframe structure, installs quicker with the wood cladding, is much thinner, yet is more complex for overhangs and is less conventional.
Jon,
You have me confused, too. You say drywall is only for floors (do you mean walls?), and you said you used solid 3/4" wood panels for the ceiling and then spoke of 2x6 decking. An air barrier is almost as important on walls as on the ceiling, and I don't see how you're creating one with 5/8" spacers behind the timbers to allow the drywall to be slotted in.
On the cathedral ceiling, a perfect air barrier is imperative, whether using foam or cellulose. Imperfectly-sealed SIPS joints and air leakage around splines have proven disastrous to the roof sheathing in cold climates.
You know where I stand on foam vs cellulose, and I've long been advocating a super-trussed roof over timber frames for cellulose. With open-web trusses, there's no need for a thermal break. You can use conventional flat trusses with diagonal webs or site-built Larsen trusses, or even TJIs which create separate insulation bays and makes dense-packing easier.
I would strongly advise against any vapor barrier in the system (most cellulose manufacturers feel the same), advise a vented roof (vented roofs always outperform and are more durable than unvented roofs except perhaps in very humid climates), advise against a non-permeable roof membrane like RoofTopGuard (0.6 perm) and recommend good old #15 felt, and advise against OSB (which is too vulnerable to moisture) and recommend CDX instead.
Robert,
To clear things up; drywall will be installed only on flat ceilings under floors between timber framed beams and floor joists. As for wood decking/paneling, thickness would be relative to the span - if fastened to trusses at 2'-0" on center or timber frame at wider spacing.
I am in agreement with movement of SIPs. I work in an office with SIP roof panels that were installed with splines and spray foam at panel intersections. Every day I have to blow off insulation material off my desk.
As previously mentioned, I am leaning toward the trusses over the timberframe system. My problem mainly has to do with thickness, which by adding rigid to the trussed system would have thinned things down and help in keeping the sheathing warm, preventing condensation.
To make the system work as discussed, from the interior; timber frame, 2 x 6 wood decking, adhered rubber membrane, trusses/cellulose, CDX plywood sheathing, 15# felt, asphalt shingles. To vent the assembly, I would use a rigid vent from soffit to ridge stapled to underside of plywood. There would be no vapor barrier.
Jon,
For the warm sheathing strategy in W. MA, you would need 40% to 50% of the R-value outside the sheathing. That's not what your hybrid approach would offer.
And you most certainly do have a vapor barrier - a near-perfect one - in the form of the self-adhering rubber membrane. That would work for an air barrier but prevent any drying to the interior. I would advise against this (Martin and I disagree on impermeable layers).
Air-sealing a T&G ceiling is always a challenge, but if you're building it from the frame up, I would suggest unperforated polymeric housewrap (Tyvek or Typar) with taped seams rather than butyl rubber or rubberized asphalt.
We are estimating this project, and my cellulose installer recommends non-vented roof assemblies for the 12/12 vaulted slopes. Again, the system consists of a timber frame and wood deck, air barrier, 16" deep open web trusses at 2'-0" o.c. filled with cellulose, vent channels and sheathing. Roof material is asphalt shingles over 15 lb. felt with soffit and ridge vents. We are in western Massachusetts - climate zone 5.
He thinks about slopes filled with dense pack cellulose like walls on an angle and "they don't vent walls filled with cellulose because there is no moisture being carried from the living area into them with air, same with slopes. As for roof temperature, he says that the biggest determining factor is shingle color and venting a cathedral slope will just bring in moisture after thunderstorms in the summer while not significantly affecting shingle wear. He prefers the space be taken up with insulation. Loose fill cellulose must be vented by code and some day that will hopefully be determined by air leakage rates but I accept that it's a lot easier for moisture to make it into a flat attic as there are usually more penetrations, access hatches etc."
I'd appreciate a continued discussion. Thanks,
Jon,
Plenty of cellulose installers have had success with unvented cathedral ceilings. Since you're going to have 16 inches of dense-packed cellulose, I think you'll be fine.
John,
A roof is not a slanted wall. A vertical wall protected by roof overhangs sheds environmental water well and is less subject to wind-driven rain infiltration than a slanted assembly protected by shingles. A roof is also more susceptible to damage from falling branches or other flying debris or simply high winds.
A roof takes the full impact of the sun, raising both roofing and sheathing temperatures, accelerating the aging of composition roofing, creating a very strong temperature-driven inward moisture drive, and conducting or radiating that solar heat downwards.
The more insulation in an assembly, the more moisture storage capacity and the less energy flux to dry it to the outside. If (or, rather, when) the roof develops a leak, if there is neither air movement nor energy flux the sheathing or framing members will not be able to dry and moisture damage is possible (A Florida Solar Energy Center 10-year simulation study documented the significantly improved durability of a vented and breatheable roof).
There is a reason that even those building scientists who believe unvented roofs are workable, almost all hedge their bets with the caveat that the assembly needs to be and remain nearly perfect and acknowledge that vented roof assemblies in cold climates almost always perform better.
Also, I would disagree that flat ceilings are more problematic, since tall cathedral ceilings significantly increase the air temperature stratification and the air pressure at the top of the thermal envelope, increasing the exfiltration drive.