I'm almost finished getting bids, for insulating my house, and as expected, they are all over the place from a product and price perspective. I'm likely going to use multiple contractors, as they all seem to be wedded to their products. Some also seem vastly behind the times when it comes to building science. One guy wanted to spray ccSPF over the rafter chutes, in a vented attic. I fail to see how that would be a good idea. Anyway, one guy said he does unveted attics all the time using dense pack cellulose. I can't find any information that. It sounds suspect in a cold climate, as I understand that humidity is a huge concern in the winter. Although air permanence would not be an issue, wouldn't vapor permanence be a potential problem? Seems risky. The more and more I read the "Moisture Safe Unvented Roof Systems" article, the more I think I really should have 2" of ccSPF on the underside of the roof decking (for Milwaukee's cold climate, anyway). This seems to be the least riskiest of all the other solutions, especially considering the roof was built with the idea it would be vented.

Bekkl:
National Fiber, a cellulose manufacturer in Massachusetts stands behind unvented roof assemblies insulated with their product.  http://www.nationalfiber.com/docs/S...ts0909.pdf . and http://www.nationalfiber.com/docs/U...es0909.pdf
This is the only cellulose I've used so I can't speak about other manufacturer's products - I would want to know what installation methods they would stand behind.

Where would you put the vapor barrier in this situation? I'm assuming you'd want it on between the roof decking and the cellulose? If not, wouldn't there be a drying issue in a cold climate?

While I too use (and for the most part like) their products, I disagree completely with National Fiber's assertions that a roof is just a tilted wall, and by extension, that a WUFI simulation of how dense-pack behaves in a wall is identical with how it behaves in a roof assembly.

Walls never experience saturating levels of moisture on the exterior for weeks/months on end. Snow covered roofs do. Snow cover elimates all possiblity of drying toward the exterior for those time periods, during which air transported & diffusion moisture from the warm interior will continue to accumulate in the roof deck & insulation. Dense pack may be a good enough air-barrier from a strictly thermal convection heat loss point of view, but it doesn't cut it in terms of moisture transport, and it's hygric buffering capacity is not infinite.

Whether dense-pack will work well in YOUR house an unvented cathederalized ceiling is highly climate dependent (and to some extent, a function of the roofing materials and orientation of the roof pitch.) The upshot of...

http://www.buildingscience.com/documents/reports/rr-1001-moisture-safe-unvented-wood-roof-systems

...is, if you live pretty much anywhere in the warmer mixed climate or cooling dominated south, or more temperate maritime zoned dense-pack would be fine for unvented attics. But in Milwaukee going with 2" of ccSPF against the roof deck for starters, and filling the rest of the rafter bays with wet-spray (dense packed or not), would be a better bet. (IIRC, in MA code requires at least 40% of the R to be foam and placed on the roof-deck side when going with a mixed SPF/fiber layup, but that may be a bit too conservative.)

A couple inches of SPF is only R12-ish, not nearly enough for Milwaukee's climate, but according to the BSC simulations, it's enough to protect the roof deck with an additional R38-worth of Spider (superfine spray fiberglass), which has very similar air-transport characterisics of cellulose (but none of the protective hygric buffering of cellulose.) Look at the simulations for both Chicago (with R38 total-R) & Minneapolis (R49 total-R). Going with all-densepack would indeed seem a bit risker in your climate, but in a combi-layup it with 2" of SPF (or with 2" of iso on the exterior of the roof deck) you should be fine.

Dana,
I have come to the same conclusion. The Dense-Pack guy gave me a decent price, so I was thinking of him coming in and blowing in after the foam guy does 2" on the roof deck. A few questions, if you don't mind:

1.) Would 2" of 2lb foam + the rest in dense pack cellulose give me the minimum of R38 in a 10 1/8" rafter (actual size).
2.) I'm assuming that I would not install ANY vapor barrier in this situation, since the 2" of foam would suffice. I'd think I'd want the cellulose to dry toward the inside in case moisture got up there.
3.) Would 1/2" - 1" of rigid foam be a good idea for the underside of the rafters, or just unnecessary at this point?

2" foam= R13

8" cellulose= R26 to R30

So yeah, you'd get R38+ out of it with even the crummiest of 3rd-world version of cellulose (most better-quality goods would give you R42-R44 in that stackup.)

It needs no interior vapor retarder if dense-packed. Spray applied (but not dry-blown) cellulose at that thickness also qualifies an ignition barrier for the SPF, so you wouldn't need an interior layer of gypsum, but if you put some up, a layer or two of standard latex would probably lower peak moisture levels in the cellulose (as would up to an inch of seam-sealed XPS.) You DO want an interior air-barrier though- either seam-sealed gypsum, stapled up housewrap (or both), even if dense-packed.

Interior XPS would reduce the thermal-bridging of the rafters, but the effect on overall performance wouldn't be as dramatic as it is in thinner 2x4 or 2x6 walls.

Normally in a cooling-dominated climate you'd want the strongest vapor retarder to be on the interior side to let the assembly dry toward the exterior. In a cathedralized roof there is no exterior drying happening during much of the period when winter-moisture would be condensing in there, which is why you NEED the SPF, which placed the condensation point for most of the season in a layer where interior air won't penetrate. There will be some condensation/vaporization cycling within the cellulose with temperature, but it won't cause issues unless the cellulose reaches ~30% water by weight. Cellulose will wick micro-droplets readily, and won't hold onto the moisture for long when average temps are higher as long as you don't lock it in with a strong vapor retarder. Above 1-perm on the interior should be fine (gut feel- haven't similated it against Milwaukee weather data or anything.)

Perfect. I think that will be the best "bang for my buck" at this point. Now all I have to do is convince the dense pack guy to use webbing instead of cutting holes in everything.