Sorry to hear that. A 2" shot of closed cell would have been good enough to protect the roof deck, a smart vapor retarder would have mitigated against mold in the rock wool.

With 2" of vent space between a structural roof deck and a nailer-deck for the shingles above that you'd be OK with the tried & true poly vapor barrier below. It's effectively the same drying rate that you'd get from a traditional vented cathedral ceiling that had a 2" gap between the insulation & roof deck. Codes typically only require 1" gap, but even on decently pitched roofs 2" is better.

building a Vented roof:
Install your rafters and sheathing as you would normally. Air seal the sheathing - the best way is to tape the joints, so you can use ZIP sheathing & tape. Most tapes will not seal well to bare OSB or plywood, so if that it what you have, use SIGA 60mm tape which sticks to anything. You can also cover the roof with an air barrier but it has to be detailed properly - just stapling it up doesn't do the job. You can tape the joints of that and seal the edges.

Anyway, then run full length 2x4's on top of the sheathing, running vertically, directly above and fastened into each rafter, and sheath that with Advantech, cover with a shingle underlayment and shingle. Vent the air space at the top and bottom.

The best products to use are hard to find, but are carried at the Small Planet Workshop, High Performance Building Supply (aka Maine Green Building) and 475 High Performance Building Products. All can help you with details.

How does this allow moisture out of the insulation space? Moisture can still condense on the underside of the first sealed sheathing layer and then rewet the interior of the rafter space, no?

Posted By FBBP on 05 Oct 2014 11:42 PM
How does this allow moisture out of the insulation space? Moisture can still condense on the underside of the first sealed sheathing layer and then rewet the interior of the rafter space, no?

Most sheathing materials will have vapor permeance of at least 1 perm when dry, but will rise to 3 perms or more if it accumulates enough moisture. The moisture accumulates as adsorb into the wood without condensing, unless it's painted with something vapor retardent- you won't see liquid moisture in the cavity until the roof deck is fully saturated, and can take on any more adsorb.




In Bob's stackup the sheathing would pass wintertime moisture into the dry air of the vented cavity established by the 2x furring above the structural roof deck. 

That's a reasonable approach in climate zone 5 or less, but becomes progressively less resilient the further north you go. When the vent space is above the fiber insulation the high vapor permeance of the fiber (>>25 perms) it's out & gone without delay into the low-low RH of the outdoor air in the vent space. But with a vented from above approach you have 1-3 perms of vapor redardency of roof deck, more than an order of magnitude slower drying rate, thus it it takes quite a bit more time to move the same amount of moisture out.  If the interior side is air-tight and has a polyethylene vapor retarder it's fine (even in Antarctica) but even fairly modest air leaks can deliver more moisture into the cavity than 1-3 perm OSB or polywood can pass on quickly.  That makes air sealing the interior of the assembly more critical than when the vent space is between the fiber & roof deck.

That's not to say you can't do it successfully, only that you have to execute it well.

So the double sheathing is not going to happen but I might go the metal roof route. I'm very last second scrambling with this idea...since I hadn't thought of it until now...but I need to get the roofing done this week. It would seem that a metal roof might provide the same vent space as the double sheathing idea...as they seem to be attached to strapping. Am I correct in this understanding? Again I'm scrambling to learn about them.

I'm guessing if I go this route I should still tape the plywood sheathing seams. I have SIGA Wigluv tape(expensive stuff). I think its its great for making things airtight but I think its vapour permeable. Shouldnt I be using something that is vapour impermeable...to keep moisture out?

Is there any other consideration I should be taking with a metal roof to keep moisture out of my unventilated spray foamed roof? Fasteners? Membranes?

Thanks!

Metal roof is installed on clips. The clips are screwed to the sheathing then the panels are clipped onto them. Depending on the type of "seam" the panel may be bent, screwed through, etc. If you get a panel with a flat bottom, there is not much chance for air to move under the metal roof. If you get a panel with raised ribs, there is some air movement but I doubt it would be much or considered enough for venting a roof assembly. If you are looking at metal roofs simply take a look at their installation details to see what is going on.

Thanks for the reply Surfsup

From my limited research... metal roofs are installed a bunch of different ways depending on who is doing it. The most hardcore way I have seen is to have vertical 1x3 or 1x4 nailed to the sheathing followed by horizontal ones on top of that creating a grid ....I guess with the goal of creating a gap that doesn't trap water in the horizontal strips. Less extreme methods are like you describe with the metal attached basically directly to the sheathing. So I guess it all depends on your goals and bank account.

I was going to go with standing seam because the fasteners are hidden. These have flat bottoms like you describe.

Im starting to wonder if I am getting insane about this. Maybe I should just spray foam between the rafters...tape the seams of the plywood...put on asphalt singles and call it a day. Ripping up the asphalt and the sheathing after 5 years is probably cheaper than some complicated setup to try and prevent leaking which is probably impossible.

If I go with the simple basic asphalt route does anybody have any advice on membranes on this unventilated assembly? Stick and peel the whole thing? Or leave it all alone so that some outside drying is possible(although probably not with the shingles and paper)

"Im starting to wonder if I am getting insane about this." uh, yea, I don't know what you're talking about... haha

Ok shingles are going on in a few days.

Should i cover the entire sheathing with a membrane like grace ice and water to reduce the chance of a leak but consequently create a vapour sandwhich with the spray foam on the inside...or..
Should i just use felt paper and in the potential case of moisture reaching the sheathing their is the slight possibility of drying to the exterior?

Im leaning towards covering the whole roof with a membrane because I'm not sure their is any significant drying potential to the exterior with asphalt shingles so i might as well do all i can to keep water out. Im also not sure what the rot potential is with a membrane on top of the sheathing cc sprayfoam beneath.

Any opinions would be much appreciated. I hope to put this all to rest this week. Nothing has caused me more anxiety than this roof.

I'd find out what Grace has to say about your proposed use. My guess is that felt would outperform under normal conditions (asphalt shingles are slightly permeable) but Grace would outperform once the shingles start to leak a little (holding water next to a permeable membrane isn't good).

How does one rip up rotted sheathing that is glued to the rafters with CC spray foam?

Well my roofing has been delayed again. This time by a corrupt roofer who has ripped me off. Now I need to find another one.

I'm considering going back to the city and convincing them to use Dana's idea.

Dana how well does the 2" ccSPF + cellulose /or fiberglass/or rockwool+ smart vapour barrier idea mitigate against rotting cause by leaks from the exterior? Would drying to the interior be sufficient in this setup?

Thanks

With 2" of ccSPF you would have about 0.5 perms between the roof deck and the interior, which isn't super fast, but good enough for most purposes. It is sufficient for managing rare bulk-water intrusions between the exterior roofing & roof deck, but not for chronic leaks.

Chronic bulk water leaks through the roofing materials to the roof deck are a problem whether the foam is there or not, but the foam will pretty much isolate the damage to the location of the leak (and down-slope, on the exterior), rather than creating a mold-filled rafter bay cavity and rotted rafters.

Leaks that penetrate fully into the fiber layer will have about 3-5 perms of drying capacity toward the interior, since the smart vapor retarder becomes highly permeable when the air in the cavity reaches 50% relative humidity or higher. The limiting drying rate factor then becomes the permeance of the paint, not the vapor retarder- a couple layers of latex paint on gypsum board typically tests at 3-5 perms.

Thank you Dana

This is going to be a tricky sell.

Previously this is the reply from the local building codes officer:

When spray foam is installed as the air/vapour barrier, no additional fibrous insulation is allowed within the assembly on the warm side of the foam. The installation of additional insulation will move the dew point to a colder portion of the assembly which will result in condensation of water vapour within the assembly

The code requirement here is 60ng which I think is about 1 perm. At 0.5 perms that puts the fiber on the warm side of the vapour barrier. Additionally when I first proposed the idea, the building codes officer said I can't have a vapour barrier on both sides of the fiber(he didn't really look into idea of a smart vapour barrier as it was secondary to his 1st argument.)

These aren't direct quotes from the code because the code is a little fuzzy on this stuff. Do you think he has a valid point about the dew point location?

Thanks

I'm not a native Canuckistani Code speaker, and I have a hard time even parsing...

"The installation of additional insulation will move the dew point to a colder portion of the assembly which will result in condensation of water vapour within the assembly"

...but it's an almost a nonsensical statement, that apparently misuses a commonly mis-used term.

The dew point is a TEMPERATURE, not a location.

The dew point doesn't "...move..." "...to a colder portion of the assembly..." It can't move- it's a function of the humidity of the body of air- it's the temperature at which the air is completely saturated, and removing more heat results in fog or dew formation before it's temperature can drop below that point. (Hence "dew point")

The dew point of the entrained air inside the rafter bay will track the temperature of the coldest surface that it's in contact with, which in this case would be the interior surface of the closed cell foam.

The relevant dew point for wintertime moisture accumulation in the assembly is the dew point of the interior conditioned space air. IIRC the National Building Code is predicated on a conditoined space average of 20C / 35% RH (a comfortable & healthy indoor humidity level) which has a dew point of about 4C. When conditioned space air comes into contact with a surface that is at 4C or colder, the moisture begins to condense onto (or adsorb into) that surface.

When the condensing surface is wooden roof deck, the moisture is adsorbed directly into the wood, raising it's moisture level. If that moisture level remains high enough long enough there is a risk of rot & mold.

When the condensing surface is closed cell foam, it condenses as liquid on the surface. If that surface is in direct contact with cellulose insulation, the cellulose wicks/adsorbs the moisture inside it's hollow fiber structures, redistributing it, and continues to insulate well, up until it is nearly saturated.

There is a finite amount of entrained air in the fiber insulation layer. As the temperature of the condensing varies over the course of day or winter season the average moisture in that air will vary with the amount of moisture that is condensing & re-evaporating. It's the AVERAGE temperature of the condensing surface that matters. As the average temp of the condensing surface drops below the dew point of the conditioned space air, the vapor pressure of the air in the fiber insulation drops, creating a vapor-pressure difference between the cavity-air and conditioned space air. When that happens water vapor will move via diffusion from the conditioned space into the cavity air.

The rate at which that moisture moves is a function of the vapor permeance of your vapor retarders (including the paint). If that vapor permeance is too high and the average temp at the surface of the closed cell foam is too high, it's possible for enough liquid moisture to get in there over the course of a winter to cause mold conditions, and that is what the inspector is (or should be) concerned about. The vapor permeance of the foam is low enough to protect the (even colder) roof deck, but since it's on the cold side of the fiber layer, it doesn't protect the fiber from moisture accumulation.

But that is why you want to have a "smart" vapor retarder. These materials are less than 1-perm @ 35%, which is about where you want to keep your conditioned space air in winter. When the condensing surface drops to 4C or below and condensing begins, the RH air inside the rafter bay right next to the smart vapor retarder drops below 35% RH, making it even more vapor-tight. The finish paint (about 3-5 perms) limits the rate at which the micro-gap of air between the membrane & gypsum rises. The colder it is outside, the lower the temperature of the condensing surface, and the lower the RH of the entrained air in the fiber, making the vapor retarder ever tighter. But when the roof deck temperatures rise (in spring, or a warmer than average winter day) the moisture is released into the fiber, raising the RH, opening the membrane's vapor retardency to release some of that moisture back to the conditioned space.

To do it without a vapor retarder, the average winter temp at the surface of the foam needs to be higher than 4C. The ratio of foam-R to fiber-R to make that happen is of course climate specific, but the notion that "... no additional fibrous insulation is allowed within the assembly on the warm side of the foam" is completely ridiculous, unless it's literally never warm enough for that surface to be above 4C over the course of an entire average year. (There are high-altitude locations in Antarctica where that might be the case, but not most places where humans actually live.) This has long been understood by code-drafters in Canada, if not by all local inspectors, which is why low permeance exterior foam is expressly allowed without interior vapor barriers, provided the R-value ratio is sufficient.

See: https://www.nrc-cnrc.gc.ca/ctu-sc/files/doc/ctu-sc/ctu-n41_eng.pdf

Our goal in your case is to provide a reasonable drying path for the roof deck, while severely limiting the amount of moisture that makes it into the fiber layers. The 2" shot of foam is sufficiently vapor open for releasing the miniscule amount of moisture that would accumulate in the roof deck over the winter, but insufficient R for accommodating a high-R fiber layer without enlisting a smart vapor retarder membrane. But those membranes DO work, and work pretty well. The places where smart vapor retarder would not work would be where the interior air is well above 35% RH much of the winter, say a hotel's pool & hot-tub facility or a sauna, steam bath, etc. Those situations aren't common in most houses, and where present those portions of the house would need special treatment on wall & roof stack-ups.

The primary vapor retarder in the stackup is the smart vapor retarder, even though the vapor retardency of the foam is important for protecting the roof deck from excessive moisture accumulation, but also for giving it a drying path. Both the accumulation rates and drying rates are important, but far more important for the roof deck than the fiber layers, especially if the fiber layer is cellulose, which can handle fairly substantial moisture loads without damage or loss of function.


It might also be useful to print out the Canadian short-sheet spec for MemBrain to present to the code enforcer, that notes in the fine print that it performs at less than 57ng/P·a·s·m (and thus ducking under the 60ng/P·a·s·m hurdle) using ASTM E96 dessicant method, while exceeding 572ng/P·a·s·m using water method, indicating just how fast it can dry should the assembly actually become wet:

http://www.certainteed.com/resources/30-26-074.pdf

It looks like you can even buy the stuff at Lowes in Canada:

http://www.lowes.ca/radiant-barrier...68192.html

Dana I'm only a new member but Ive been reading posts here and elsewhere for years and I have to say that the amount of thought and time your put into helping strangers admirable. Thank you very much!

Not that I want defend the statement about the foam and prohibited use of fiber, but I suspect it was just laziness. I suspect the thought is that because its so cold here that any typical fiber insulation worth installing between the foam and the interior in a typical rafter bay will increase the likelihood of the foam surface being below 4C causing excessive condensation due to a lack of vapour barrier which isn't permitted on both sides of the fiber(yes I know the smart vapour barrier fixes this problem). I think the real trick in convincing the officer is getting them to trust the concept of the smart vapour barrier which I suspect is something they have never had to consider when book-ended with ccSPF and fiber in between in an unvented assembly. Actually ill go one step further and say they have never seen a single request for this setup in their lives.

I'm going to try again...wish me luck!

I do have couple more questions if you don't mind...

Since asphalt shingles essentially make the roof vapour impermeable, should I consider covering the whole thing with a peel and stick membrane to reduce the risk of leaking? Under normal circumstances that is probably overkill but in this case with the very slow drying potential of the 2' of ccSPF ...shouldn't I take extra care in preventing leaks? It seems most people think its a bad idea but I think if the roof cant really dry to the exterior because of the shingles then why worry about the negative consequences of adding a vapour barrier(and hopefully a leak barrier) to the exterior of the sheathing? Its seems the fear is a little misplaced in an unvented assembly.

Secondly... does the fiber in your proposed setup have to be cellulose? I suspect it will much easier to insulate a cathedral ceiling with batts (rock wool or fiberglass).

Thanks again ...hopefully the next post from me will be a report back saying the city approves!

Posted By Dana1 on 17 Nov 2014 05:21 PM
I'm not a native Canuckistani Code speaker, .....

Post of the Day....!!! 

One more Canuckistani question

Dana if understand your previous post about leaks....would it be fair to say that leaking would be equally detrimental to both a straight 6" ccSPF setup and the one you are proposing? Which I guess brings me back to the question about membranes and their use on an unvented roof.

Thanks

A layup of #30 felt + asphalt shingles comes in at about 0.1 perms, which isn't much drying capacity even when it ISN'T covered with snow or raining.

Impermeable peel & stick membranes are commonly used with 1-2" of closed cell foam on the underside of the roof, and would not adversely affect anything, and WOULD limit the leak risk, since they are usually self-healing around nail penetrations, etc.

While cellulose would add some resilience due to it's moisture buffering capacity, high-density blown fiberglass (Optima, Spider, L77) , or rock wool batts also work fine with smart vapor retarders.

Most of WUFI simulations done in this study were done with JM Spider @ 1.8lbs density as the fiber, and NO smart vapor retarder- just standard latex paint:

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

If you look at the Table 3 results on p11 (p12, in PDF pagination) in the 2" ccSPF + spray fiberglass column for International Falls MN (US climate zone 7, comparable to cooler Ontario locations) it looks pretty encouraging with just the 2" of closed cell foam. Adding a smart vapor retarder does it one better.

In practice an air impermeable layer is required on the inside and outside of the air permeable insulation to control moisture movement.

This comment (from the link above) seems to support the use of a taped, interior side smart film in a fiber/CCSP roof design.

I wonder if you could keep the inspector happy and make a roof that can be repaired (ie, sheathing can be replaced) by applying CCSP from the roof side (ie, against wallboard).

Thanks again Dana

Just to be clear...the possibility of sheathing rot&mold caused either by interior or exterior moisture in either the straight 6" of ccSPF setup or the 2" ccSPF + fiber+smart barrier would be more or less the same. Correct? What has been achieved in the latter setup is a reduction in the amount of spray foam used but not a reduction in the likelihood of damage that might occur in the 6" ccSPF setup...correct? I don't want to make the argument that one affords more protection from rot and mold if that is not the case in any appreciable way.