Ways to Simplify Foam-Free Roof assembly?

We are in the planning/engineering process of a heated barn/shop with second level living space in climate zone 5. I have made a couple posts here leading up to this, but to summarize… We are using open span, Glu-Lam gambrel rafters, 12’ on center, sitting on top of Perfect Block ICCF walls. Between rafters will be 2×10 I-Joist purlins with dense packed cellulose. We would like to avoid all spray foam, and avoid any external foam due to high wildfire risk. In order to reach an R-value of 49 (or better), and minimize thermal bridging through purlins, and overhangs, we are going to need some external roof insulation. The plan is to frame a 2×6 insulated over-roof 24” OC with (preferably mineral wool) batt insulation in the cavities. The 2x6s would run vertical, from eave to ridge, and would also be cantilevered on the gambrel ends to form our over hangs. We would use 3.5” batts to leave a 2” vent space under the roof sheathing.
MAIN QUESTION…
Am I over-complicating this in an expensive and unnecessary way?
Is the thermal bridging as detrimental as the building scientists suggest?
Due to snow load (70 psf) this cathedral roof should probably be vented regardless. Would we be better off with a deeper purlin cavity and some strapping to create vent space? With that approach we would have to design some sort of externally supported overhangs, which is not impossible but may be costly.
What could we do to simplify things with minimal loss in performance?
I will attach some photos to hopefully better explain our assembly. This is an off grid property, but we certainly don’t have an unlimited budget, so a balance of efficiency and cost is what we’re working towards. Thanks in advance for any criticism, feedback, advice, etc!
GBA Detail Library
A collection of one thousand construction details organized by climate and house part


Replies
In terms of energy use, what matters is assembly R value. Since thermal bridges are part of the calculation, you can see how much it would effect your heat loss. This is a decent calculator if you want to try different options:
https://www.ekotrope.com/r-value-calculator/
Reality is that an R30 assembly gets you most of your energy savings, so aiming for much more than that generally doesn't pay back unless it is something like cheap loose fill in the attic.
In your case you can cut cost a couple of ways. You can save a fair bit by going with dimensional lumber instead of I-joists with the existing assembly. You would still need two layers of insulation to get to r49.
Another option is to bump up the purlins to 14" or 18" (either truss or I-joist) and skip the exterior insulation layer. Dens packing a slightly thicker cavity doesn't add much cost. You would still need the rafters overtop to form the vent gap, but those can now be 2x3 or 2x4.
That is a great tool, thanks for the link!
After playing around with it, my original design total R-value is around 45. If I eliminate the over-roof I would have to go all the way up to a 16” joist for a total r-value of 47, or a 14” joist puts me at 41. Neither of those cavity depths will work with our trusses. I was hoping 12” would get me close, but that’s only r35 total.
Looks like the overroof remains our best option! If I have to add strapping anyway, it might as well be 2x6s with some exterior insulation. It is a little more work but I think the batts will be pretty easy to throw up.
Thanks again for the input. Many of your comments on other posts and my previous posts led me to this conclusion. Much appreciated, Akos!
Have you had a structural engineer involved yet? I wonder about having those I-joists leaning so far over as to be almost laying flat. Will the 2x6's be supported at the walls with a birdsmouth cut?
I don't know about the cost, but I think it's worth investigating getting Glulam purlins to match the trusses. Then overlay with 16" I-joists running vertically and use the manufacturer's standard details for attaching overhangs. Create the vent space by fastening a baffle to the underside of the upper flanges.