Rock wool is not thermal conductive or has “cold or hot spots”, as a matter of fact it has a flame spread of zero(see MSDS). It will not freeze since it cannot absorb water (see MDSA freeze point is “N/A”).

Of course rock wool is thermally conductive, and of course moisture can form on (not in) the fibers when the temperature of the fiber is at the temperature of the entrained air.

In a framing cavity the dew point (no confusion here- but I'll abuse it anyway) of the entrained air will be no higher than the coldest surface, which in winter would be the OSB, but as ice crystals grow on the OSB, they will extend into & between the rock wool fibers.  Are rock wool fibers are more thermally conductive than the ice fibers? (Beats me.) But either way, as more moisture enters via vapor diffusion from the interior side, it'll end up in that outer layer of growing ice crystal.

Dana1: Are we having fun yet. I love it when you create your own Roxul spec sheets and say how it is thermally conducive but give no value, and how you state moisture can form on it…etc that knocks down the r-value which contradicts the Roxul claims. I think you should call them and correct their data sheets, Greenguard certification, third party testing, and Engineers. Don’t take this personally, but I have looked at their data sheets in depth and talked to their Engineers and I will design to their data, not Dana1 data r-value knock downs, freeze points, or whatever else you dream up, more so in the particular assemble. So I’m done with the debate.

I’m in the middle of a similar test on a two piece split seal design which has two materials in contact with one another, and chem treated steel. The parameter in question is 'glass transition" of the seals, and supplier...The France client is paying us large sums of money to figure out why this seal is failing in service. It is the stupidest thing, I could have told them before they hired us to fault find the corrective action is a complete re-design, it is bad but since they designed it they refuse to admit it. The flow is incorrect(as in the flow of a vented attic or wall) that cause failures. We simulate the conditions in the field in a test chamber with various pressure, temp, salt, etc, cycles. It is failing at -40 F. I think the geometry and the contraction, expansion, opens the seal for contamination, then 3000 psi deforms it and it leaks. You read the BSC wall test and how difficult it is to induce or simulate a situation such as this above with pressure and temp, and gage measure it at the right locations(thermal, pressure gages, etc) in cavities. We use the same CO2 they did in its liquid state expand it to a gas to get the freezing temps. Perhaps you read BSC results of the knock-down r-values in 8 wall designs, in a seal case it is leakage rate too. A sill seal can leak in the same manner over time, so can peel and stick tapes, caulks, adhesives, SPFs, if cycled. If they hire me to redesign this I will try and design out of the seals, lessen the loads they see, or into a single one that takes the cyclic loads better. You don't want to subject materials to cycles unless you have to. That is what this company should have done along time ago but, they don’t listen like some people out here ;) As long as they are paying my bill I don’t care. It will end up on my tube for a redesign sooner or later. When it does my CFD model will be accurate to the lab and field data this particular design has seen since it was place in service a decade ago. The new design model will be re-calibrated to new lab and field data for future use.

TLP - OP lives in Alberta so ABC applies. As posted earlier •••
Our code (ABC) also adds an explanation (See Appendix A)
•••recent research indicates that venting of attic or roof spaces is generally still required.The exception provided in Article 9.19.1.1. recognizes that some specialized ceiling-roof assemblies, such as those used in some factory-built buildings, have, over time, demonstrated that their construction is sufficiently tight to prevent excessive moisture accumulation. In these cases, ventilation would not be required.•••

Now I appreciate that our code officials and the national research team are not as smart as you and roxol, nevertheless, the code trumps. Experience in our climate shows that it take very little infiltration to ice up an assemble that is not vented. Furthermore, even with Roxal, there is a stack effect inside the sloped roof assemble which both draws from the inside and concentrates at the top.

Maybe it does make a difference where you live.

FBBP, long day at work. I been thinking about HVAC and stack effect and putting it to good use in tight houses past few days I'll address that next. As far as code, been looking at mine...I can see a potential battle brewing up with them since as stated alot of code does not account for dense materials, for example, FG was used to set the code but, there is no way it understands all the different material properties that can be in an assemble like this... there is no room for opinions or guessing here....that takes a good data base on similarity or modeled performance design and empirical data, such as BSC has started to produce and manufactures testing, developing the knock down r-values as Roxul has which is zero! You also see ALOT of building "science" that decouples materials technology and does not understands it. They seem to be physics related which is not design-build reality. I'll explain more later......There is a difference between physics theory that seems to change on a whim and practical engineering design. The terminology is even different. I've done ALOT of HVAC design too, it has to integrate with the wall and roof assemblies.

Posted By FBBP on 02 May 2014 05:58 PM
TLP - OP lives in Alberta so ABC applies. As posted earlier •••
Our code (ABC) also adds an explanation (See Appendix A)
•••recent research indicates that venting of attic or roof spaces is generally still required.The exception provided in Article 9.19.1.1. recognizes that some specialized ceiling-roof assemblies, such as those used in some factory-built buildings, have, over time, demonstrated that their construction is sufficiently tight to prevent excessive moisture accumulation. In these cases, ventilation would not be required.•••

Now I appreciate that our code officials and the national research team are not as smart as you and roxol, nevertheless, the code trumps. Experience in our climate shows that it take very little infiltration to ice up an assemble that is not vented. Furthermore, even with Roxal, there is a stack effect inside the sloped roof assemble which both draws from the inside and concentrates at the top.

Maybe it does make a difference where you live.

Looks like there is a loop hole in your code to not vent, which is what I would pursue when using a dense insulation. Otherwise, the answer is simple, if code "trumps" and "Experience in our climate shows that it take very little infiltration to ice up an assemble that is not vented"....then tap into that "experience" by prescriptive methods to find detailed design info and answers and stop wasting your time out here. Sounds like the assemble will be perfect with no r-value knock downs too. Good luck! ;)

Part of the problem is that you can build a good roof assembly that would work for most people and then someone comes along and maintains a high interior humidity or does something to (further*) pressurize the house (driving moisture outwards). Don't do these things, air seal it really well, pack it tight with moisture absorbent cellulose and keep the exterior side slightly vapor permeable (no vapor impermeable synthetic underlayment) and it should be fine without a vent.

* - stack effect alone will cause some roof pressurization in a neutral house. But blocking a return duct in a closed door room will make it much worse.

Iron slag has a thermal conductivity. There's no need to look up what that is- spinning into fiber doesn't change the thermal conductivity of the material. No spec sheet required.

The thermal conductivity of the combined air + fiber layer has a thermal conductivity too, as measured by ASTM C519 protocols for R-value labeling purposes in the US.

Dana 1: Iron slag has a thermal conductivity.

TLP: Who said anything about iron slag? “Slag Wool” has not been used much if any since 1840 since it was injurious to workers. It has use to this day in temperature cooling or smelting molten metal due to its low thermal conductivity. Roxul uses “Mineral Wool” (see spec sheet) derived from rock and/or minerals. Minerals and iron slag are two totally different materials in case you don’t know. There are over 5000 species of minerals. Roxul also adds resign as a binder and oil that makes it "char" more than thermally conduct, so, good luck guessing the thermal conductivity or other properties.

Dana 1: There's no need to look up what that is- spinning into fiber doesn't change the thermal conductivity of the material.

TLP: Since thermal conduction occurs at the molecular structure and bond, internally, lower density fiber has lower thermal conductivity than HD fiber.
Dana 1. No spec sheet required.

TLP: I’d recommend you start looking at all the material properties in the manufactures spec sheets, if that spec data was independently developed by test since it is obvious you have no clue what you are writing about.

Dana 1: The thermal conductivity of the combined air + fiber layer has a thermal conductivity too, as measured by ASTM C519 protocols for R-value labeling purposes in the US.

It’s not just a matter of “combined air + fiber” and “thermal conductivity”, "r-value" (steady state thermal resistance), that needs to be considered when designing an assemble. Heat transfer as a result of mating material properties, interaction of those properties with dynamic pressures and temperatures are more applicable. The traditional methods that work most of the time, to understand all this is to begin with, is the manufacture spec sheets, either field data by similarity, or hot box testing and design development.

Any slag/stone/whatever has a thermal conductivity.

It's simply preposterous to assert that the stone fiber has no thermal conductivity.

The thermal conductivity of the batt is higher than the conductivity of air, so...

Nothing about the materials in the fiber or the sheathing will trump the physics of water's phase changes, even if it can sometimes have minor effects at the temperature margins.

Dana1: Any slag/stone/whatever has a thermal conductivity. It's simply preposterous to assert that the stone fiber has no thermal conductivity.

TLP: You keep twisting things around to suit being right and I got a feeling you will continue until you are in your own mind. Again, Roxul is not made of ‘stone’ it is made of ‘mineral’ which is completely different than stone and iron slag. You can't the material right so how do expect to know everything about it...It has such a low steady state thermal conductivity it does not matter. See below .04, wood .15. Wood is MUCH more of bridging concern, that is why we need to insulate and isolate it. Roxul mineral is lower than .04 until it reaches 400F then it starts to climb, that is why their spec sheet shows N/A since most walls/roof will not see 400F.

Dana1:The thermal conductivity of the batt is higher than the conductivity of air, so...

TLP: Ambient Air(,02), pressurized lower depends on pressure and direction, both air and insulation very low…..

Dana1: Nothing about the materials in the fiber or the sheathing will trump the physics of water's phase changes, even if it can sometimes have minor effects at the temperature margins.

TLP: I’m not sure what the heck you are talking about here, but, assuming you are referring to the states (not “phase”) of air and water being a gas, liquid, or solid (frozen), and the relation those states have with the states and composition and strength of materials when confined in a pressurized cavity, that is either vented or non-vented, and relative combined heat transfer rates or flux of the assembly, input-to-output. Pressure and temperature gradients will definitely have an effect on each other in different states by consuming or releasing energy; more importantly air and water states combined with temperature and pressures differentials will definitely have an effect on the material properties.

Conductive heat flow occurs in direction of the decreasing temperature since higher temperatures are associated with higher molecular energy. In a vented assemble the temperature and pressure gradients will be higher than unvented in the direction of the vent. Thermal cycles will be higher too that fatigue the strength of materials, especially foams, seals, caulks, tapes, etc with adhesive additives or low densities materials. I know I see it in hox box test every day.

“Physics”: Seems to me alot of the "gurus" around “building science” confuse everyone with their theories that seem to change. I’m guessing these Physics, etc, majors could not find a job figuring out how fast an apples fall from a trees, so they came to this industry and are trying to turn “Design-Build” into “Physics”. This is applied science, more specifically thermodynamics, fluid dynamics, aerodynamics, structural statics and dynamics, not “physics 101”. Those disciplines are support functions to the design that has to satisfy them, and a lot of other things like actual applied structural loads, and manufacturing or the ‘build” . In design, we are not interested in general physics theory that cannot be put to practical use. I would suggest that any “physics” theory that these so called “building scientist” have, if it does not include a practical design-build application, how material properties and manufactures specs for example, are effected by "physics", dynamics, and if it is not validated by at least a hot box test, ignore it! If they do not have lots of proven experience designing and building homes, or something structural or mechanical, or understand the design-build processes, ignore them. Don’t buy into their post, blogs, websites, conventions, and bs. They should rename the term to “Building Engineering” or “Building Design-Build”. Put the “Scientist” in Silicon Valley, CA think tanks where they belong, see how they do around their peers, stop trying to impress builders and contractors or get them to follow them as icons they obviously are not.

http://www.engineeringtoolbox.com/thermal-conductivity-d_429.html

Again, Roxul is not made of ‘stone’ it is made of ‘mineral’ which is completely different than stone and iron slag.

From the Roxul web site:

WHAT IS STONE WOOL?
ROXUL insulation is a rock-based mineral fiber insulation comprised of Basalt rock and Recycled Slag. Basalt is a volcanic rock which is abundant in the earth, and slag is a by-product of the steel and copper industry. The minerals are melted and spun into fibers.

WHAT HAPPENS IF ROXUL INSULATION GETS WET?
ROXUL insulation is moisture resistant yet vapor permeable. In the event the insulation becomes damp or wet, the insulation, when thoroughly dried, will maintain the original performance characteristics. ...its thermal properties will be restored after drying,...

Posted By jonr on 07 May 2014 10:09 AM

Again, Roxul is not made of ‘stone’ it is made of ‘mineral’ which is completely different than stone and iron slag.

From the Roxul web site:

WHAT IS STONE WOOL?
ROXUL insulation is a rock-based mineral fiber insulation comprised of Basalt rock and Recycled Slag. Basalt is a volcanic rock which is abundant in the earth, and slag is a by-product of the steel and copper industry. The minerals are melted and spun into fibers.

WHAT HAPPENS IF ROXUL INSULATION GETS WET?
ROXUL insulation is moisture resistant yet vapor permeable. In the event the insulation becomes damp or wet, the insulation, when thoroughly dried, will maintain the original performance characteristics. ...its thermal properties will be restored after drying,...

Yet their CAS # points to a 'glass' product. I'll call hopefully talk to a senior engineer this time. Meantime, I got a seal failing in the same -40F ambient temp hot box. First time it leak, second part is good at the same -40 zero leak. 25 things it can be, no consistency, a new design we are trying to proof test before it goes out the field. I didn't do the initial design, but I can see probs with the drawing it is being built to, tolerance accumulations that can cause variations. I got to go look at the cost and capability if I tighten them up. The seals were spec'ed out at a supplier that mates to another supplier and looked good on paper for thermal capatibility, it could be due to thermal mismatches or structural loads too. Now this is typical design-build integration, and why I am trying to get away from a supplier or seal controlled building design. I'll try and get some pics up of the hot box after Roxul answers my questions.

Gee I dunno, I guess I don't really understand physics, but they awarded me the degree anyway- go figure.

Most of the building design doesn't take a physics degree, to be sure.

Posted By Dana1 on 07 May 2014 02:29 PM
Gee I dunno, I guess I don't really understand physics, but they awarded me the degree anyway- go figure.

Most of the building design doesn't take a physics degree, to be sure.

I found the perfect job for you and your Compadres, a “Rocket Scientist” vs “Building Scientist” @ Boeing in a lab coat where you belong. The position requires you ‘push the boundaries of what is technically impossible” which you’re good at. https://jobs.boeing.com/JobSeeker/JobView?reqcode=14-1006442 If that fails you might try GBA or BSC writing a bunch of BS...just kidding! ;)

I've designed stuff that is flying on a Boeing satellite (or did it crash, and I just didn't hear about it? :-) )- not sure I could handle dealing with that many layers of mis-management on a regular basis, but thanks!

Posted By Dana1 on 08 May 2014 05:51 PM
not sure I could handle dealing with that many layers of mis-management on a regular basis, but thanks!

You got that right!! don't get me started. Roxul Senior Engineer never called me back, been two days, so I called left a message for that dept's supervisor. They have contradictions all over their website, if I don't get answers soon I won't use their product. Problem here is quantifing moisture and air flow with mechanical heat transfer and publishing it for design guides. That can require some expensive R&D and testing I see they did not do. We model it using a Ansys Fluent CFD modeling to get the heat flux loads or we test just to develop loads, then export the loads to FEM then test lab to calibrate the model, pass or fail the part. It can be a real headache and complex. That is why I get a good chuckle out of seeing people trying to use r-values & conduction off the internet alone. TOTALLY mislead. The model(s) only gives an idea when comparing design options which would perform better nominally. The test lab tells us if we screwed up. Then we try and move thermocouples and pressure gages around so we don't look like complete idiots to "mis-management". If that don't work there always the unemployment line : 0 ) http://www.ansys.com/Products/Simulation+Technology/Fluid+Dynamics/Fluid+Dynamics+Products/FLUENT+for+CATIA+V5

Just an fyi for FBBP
I spoke with a larger spray foam contractor today. He told me that every roof they spray foam is unvented.
Now I know just because all the roofs they do are unvented it doesnt mean that's the best way to go but it does mean that the code doesn't seem to be preventing the practice and it is quite common in Edmonton.
Im still not sure what to do in my case. I start excavating next week

Roxul supervisor finally called me back yesterday. They are busy opening a new plant in MS. Roxul is popular in Europe and CAN not the US yet. It is made out of "Salt Rock" that is locally sourced, and iron slag from the iron industry. The MSDS is a generic test form, data like vapor pressure, boiling and freeze point is not required that is why it is "NA" . They heat the composite material up to 2850 F (where they get their high fire rating) to smelt it, then use resign to spin it, and de-dust oil for dust and to better the moisture property. The installed properties are as stated on their website, little to no r-value knock down due to low sorption rate, water insoluble, drainage, etc....Of course as with any installation, if you design it to stay in a puddle of water it won't perform as well compared to being completely dry. It can freeze on the outer layers, not internally, it would be so small that it would have no effect on r-value. Turns out BSC has done some additional testing to verify it, in addition to greenguard, etc.....Doug sounded like a sincere Engineer, noting it;s strengths and weakness (which are few). I'm convinced, and impressed with a product that welcomes third party testing and has such a long history of proven usage, it's a great product and we will be switching to it for restoration and new construction.

Dave, your best shot is a look at what is working in your area with some scrutiny. The best source is renovation and restoration contractors/AHJ vs new construction to find out what designs are having issues. I'd NEVER seal in SPF with all it's known fungi additives(IE: Bromine), but to each their own. I suggest if you are wanting to try something new to the area hire an Engineer, or, get with the manufactures Engineer that can point your to third party testing that includes conditions in your climate zone. Good luck!

•••I spoke with a larger spray foam contractor today. He told me that every roof they spray foam is unvented. ••••

Dave - it is the same here in Calgary. The thing is, while there has been a few sprayed roof in Alberta that are more then twenty years old, the bulk are less then ten. Long term damage is yet to be estbished.

Interesting side story. a few years ago, I was looking for a cargo trailer to use for the ICF business. CPA auctions had two on the block. Both good looking units. Added benefit, they had been used by a spray foam outfit so where all sprayed inside. Figured that with a small heater the boys would have a nice coffee room during cold weather. Happen to notice a 16" long piece of foam hanging loose from one of the roof cross ribs so checked it out. What use to be square tubing was now u channel. The steel insulated by foam had rusted out completely on the bottom. A little more checking showed that half the cross ribs in both trailer where compromised.
Just food for thought.

The other thing that hasn't been discuss is the fact that no matter how much insulation is placed, there is always some heat loss. With no venting, this is transmitted through the sheathing. If you have valleys, it will melt the underside of the snow pack, leading to ice dams.

So after a number of delays the roof should be framed next week. I ran Dana's suggestion by the building codes officer:
"But short of that a couple inches of closed cell foam on the interior, cellulose fill and a smart vapor retarder actually works."
Unfortunately they wouldn't go for it. If I'm going unvented then all they'll let me do is cc spray foam up to at least code minimum R value. No other insulation and no other vapour barriers "smart" or otherwise.

Im considering trying Bob's idea of the double sheathed roof. I have thought about it some more ...I have very large overhangs so if the 1st layer of sheathing just goes to the exterior walls and the 2nd layer goes all the way to the end of the overhang then I think there will be enough intake opening. The 1st time I read the idea I imagined the two layers being identical in area which didn't seem to offer much in my mind but with top layer extending past the bottom layer the surface area of the opening is much larger and its protected by the overhang itself.

If Bob is still reading this or if anyone has a detail for this kind of setup id appreciate it. Specifically I'm wondering how big the gaps should be between the lengths of 2x4. Obviously if I just mirror the rafters and run the 2x4s end to end with no gaps I wont gain anything.

Thanks