I have several related questions within this post that I feel need to be addressed at the same time. I live in northern Maine (~8300 HDD) and am attempting to build a passive solar house next summer (2-story, slab on grade ~2000 sqft). I have been reading a lot of forums regarding attic/roof insulation and air sealing and am still left with some questions. The following is what I know (or think I know) and planning presently: Standing seam metal roofing. Without crunching any numbers, I am estimating that I will require ~R-75 in the attic. Unvented cathedral trusses with 20-24" heel height (not sure if that's a realistic depth) manufactured without overhangs (allowing for wall sheathing to be run up to the roof sheathing, covering the outside face of the truss-tape on the roof/wall joint). Overhangs to be fastened to trusses, over the wall sheathing. Blown-in cellulose to fill the entire truss volume (assuming 1 inch=~R-3.5, 22"x3.5=R-77). This is where I become unsure about my materials and the order in which they are installed. Zip System roof sheathing (taped). If Zips are used, I will not require felt covering as I would with Advantech, but not sure if the savings from avoiding the felt covers the increase in Zip panel cost. Both types will provide me with a water tight barrier as all seams will be taped. One major question: To create an air tight envelope that resists condensation at the roof sheathing (or any other level), is this best accomplished with CC spray foam on the underside of the roof sheathing which then envelops the upper chord framing down to the top sill plate? I have been told from a passive home consultant that good taping is a better air sealer than spray foam and less expensive. If this is true, have I then achieved it my simply installing Zip panels that are taped? Other people have said why air seal the roof and create a conditioned space in the truss, when you should air seal the ceiling? Would seem to me that the chances of developing condensation in the attic recesses is dramatically reduced if air is not allowed to pass through the ceiling into the insulated truss. Maybe a an air barrier does the trick in the ceiling above the OSB and sheetrock. My priority is of course creating an air tight envelop to maintain a passive solar house characteristic while eliminating any chance of developing moisture in areas that are hidden. Any ideas would be much appreciated. Thanks.

Switch to some type of vented design.

90% (+/-) of moisture travels on air currents, so air sealing the ceiling will keep most of the moisture generated within the house out of the attic and therefore unable to condense on the roof sheathing. the other important consideration is that is that if you do get condensation on the roof sheathing, it may dry out and disipate with no damage, or cause mold and rot. Building Science (buildingscience.com) has seen and written about both scenarios and the conditions for each. I tend to think that will a sealed attic and a very tight building envelope you'll be fine.

Best advice I can give you is to go to the NESEA conference in Boston in March, and, if you're able, the Better Buildings conference in Burlington in February. Both have lots of seminars given by people designing and building houses like the one you're planning to build. Go, listen & talk to the other attendees. The questions you are asking are good ones, but many of the answers aren't yet cast in stone and are still being discussed and still being tried and studied.

Also, if you don't have it already, buy Builders' Guide to Cold Climate by Joe Lstiburek.

Note that Joe L is the one who wrote:

"You can’t build unvented roofs where it snows a lot even with super insulated roof."
But you may get away with it with R75.

It's always better if it can dry in both directions.
Two air barriers outperform a single one.

And it's also JoeL that retrofitted his own house with an unvented ~R60 roof that never experiences ice dams.

Metal roofs with greater than 6:12 pitch and no valleys shed snow well (but steeper is better) and almost never end up with ice damming problems even at lower-R than R75.

Standing seam metal roofing mounted on 2x purlins (rather than nailed or on any thin rainscreen or felt-type substrate) are inherently vented where it counts, reducing the ice damming risk even further. Some substrates used for metal roofing are extreme vapor barriers, and would not allow any drying toward the exterior. Purlin mounting with a moderate-permeance felt or naked-ZIP is more resiliant than ANY method that puts metal very close to the deck, since it can dry toward the exterior even in winter, even with 30" of snow (if it doesn't avalanche off the metal for a few days.)

Both the ZIP and the interior surface need to be air-sealed on the insulated-truss roof.

Points well taken. Thanks for all your help. I will try to head to one of those conferences and keep reading.

Is it true you need to replace exposed screws after 20-25 years because the rubber seal could leak around the head of the screw?

One way of getting a leak in a metal roof is to overtighten the screw thus squashing out the neoprene washer so that it is exposed to UV.  Another way is to use sheets longer than 27'.  Expansion and contraction can enlarge the screw hole until it leaks.  At least this what I have been told by metal roof companies.

One way to use long sheets without much worry about expansion and contraction is to use hidden clips on a standing seam roof that allows the metal to slide.

I agree, the best way to avoid screw hole leaks is to not have any exposed screw heads. I'm curious if glues can't be used more with steel roofs. Or possibly spot welds (say for a steel roof to steel framing).

jonr,

I think the difficulty with spot welds is that the welding can damage the protection that has been applied to the steel.  At least that is my guess since some steel has extensive protection applied at the factory.  Some standing seam types of roofs and SIP roof panels lend themselves to hidden clips that allow movement.

I wonder of they could spot weld on tabs during the manufacturing process (ie, before zinc and paint). These tabs could hang down into the attic and serve as attachment points (ie, screw tab to rafters).

Posted By Alton on 15 Apr 2013 10:52 PM

....Installed price is around $5 per square foot.

I cannot remember now if I have seen steel roofing panels with similar technology.

That is ~4x more per square foot than I just spent having 30 year architectural shingles installed.  I priced standing seam panels with the slots for hidden fasteners and was quoted $300/square for materials.  This was 29 ga metal that needed to be installed over solid decking and roofing felt similar to the shingles I installed.  My shingles cost $86/square.