Central Indiana is US climate zone 5, and 3/4" of exterior poly iso is ~R4.5, ALMOST sufficient R value for dew point control on plywood/OSB/fiberboard sheathing to skip interior vapor retarders for full-depth 2x4 w/cellulose or fiberglass, and would have HUGE margin for 1.5" fiber insulation. With no structural sheathing between the cavity fill and the iso it doesn't much matter, since the foil facer on the iso isn't going to be degraded by periodic condensation, and the condensing events will be infrequent and not deep enough to damage the fiber insulation or drip liquid water down on the bottom plate, as long as it's air-tight. But there will be a small loss of performance of the outer inch or so of the fiber insulation.

Going with 3/4" of EPS would enhance exterior drying, but at R3, well below the IRC prescriptive R value for 2x4 framing for zone 5 (R5):

http://publicecodes.cyberregs.com/icod/irc/2012/icod_irc_2012_7_sec002_par025.htm

To bring up the existing exterior insulation to the IRC prescriptive levels without scrapping it, a layer of 1/4" or 3/8" fan-fold XPS siding on top of the iso will get you there or better yet cut'n'cobble 1/2" or 3/4" iso on the interior, sealing it to the framing at the edges with can-foam and compression-fit the batts into the 2.75-3" deep cavity. Compressing R13s increase the density and R/inch, but at 2.75" of depth even R4.5 iso would be sufficient for dew point control.

Rock wool is neither more nor less moisture friendly than fiberglass. It's somewhat better at rejecting liquid water on bulk incursions than fiberglass, but that wont' save you if it's really leaking liquid from the exterior. It's totally air and vapor permeable, but won't wick moisture away from the structural wood the way cellulose or cotton will. Cotton batts are expensive, but are more protective of the wood due to it's moisture buffering capacity. Cellulose is cheap, but the blower rental for a small job may make more expensive than cotton batts.

As long as you have sufficient R value for dew point control on the exterior you don't need an interior vapor retarder, and any batts should be UNFACED, since even kraft facers will impede drying toward the interior. If batts, make sure it's full fill- tuck it in the edges and corners to make sure it's touching the exterior sheathing, then tug the batt out to where it's slightly proud of the interior stud edges, to be pressed in by the gypsum. For the 1.5" deep sections it's OK to split an unfaced fiberglass R19 and compression-fit it in place, which will be easier to do without voids than trying to perfectly-split R13s. The density & R value of the compressed area will be the same whether you use split R19s or R13 (they're the same weight per square foot), but it's easier to get a full-fill with the fluffier batt. (This is one of the very FEW places where the crummy low density R19 batt is the preferred product.)

If it were my place (and it isn't) I'd add an inch of cut'n'cobble iso against what's there and fill with cellulose on the bigger cavites (since I own an insulation blower), and do a cut'n'cobble &/or FrothPak (1.75lb 2-part closed cell foam) fill on the really narrow cavities. If I didn't own the blower I'd go with 1" cut'n'cobble iso on the big cavities, and compress split R19s in place on the big cavities. A 12 board foot FrothPak would probably cover all of the edge sealing on the iso AND fill the narrow cavities.

As long as you have sufficient R value for dew point control on the exterior you don't need an interior vapor retarder,

To prevent condensation. Mold is another issue. I'd always block air flow with an interior side air barrier film. And consider MemBrain and a wall that can breath outwards.

I have some foil-faced (on one side) 1/2" EPS and R-19 craft faced batts: both leftover from another project.  Sounds like I could just take the facing off of each of those, install the EPS up against the existing polyiso, air seal, then put the unfaced FG into the cavity.  Am I correct with that?  Considering it's <18 ft^2, it'd be great to use my leftovers versus buying something else that I'm going to use such a small quantity of.

Don't bother taking the facer off the EPS just install it up against the polyiso and air seal. If you're concerned about somehow trapping moisture in the EPS (I'm not), put the facer side up against the facer of the iso. Half inch EPS is only R2, but that's just enough for dew point control where there is pre-existing 3/4" iso. But if you have enough to double up the EPS, an inch is definitely better. Where there's only fiberboard, no iso you DEFINITELY need to double-up the EPS, for a nominal R4 foam/R9-10 fiber. Also, it's slightly better on the fiberboard section to put the foil facer of the EPS on the inside, facing the fiberglass.

For purposes of the rest of the discussion I'm going to assume you're adding 1" EPS in everywhere, leaving a 2.5" nominal gap for the fiberglass.

Split the ~6" nominal R19s into two ~3" thick versions for compressing in place. Cut the taping tabs off the edges, but install the batts with the facers with the facer tight up against the EPS, to guarantee maximal drying toward the interior. The kraft facer isn't any more susceptible to mold than OSB or fiberboard.

The worst-case sections with the most condensing hours would be the fiberboard section, but with R4 foam/R10fiber, (28.6% of the total R as exterior foam) that's already a more favorable ratio than the IRC prescriptive R5/R13 (27.8% of the total R as exterior foam) for 2x4 construction in zone 5. The areas that have the pre-existing 3/4" iso are even better, with an R8.5/R10 ratio, or 46% of the total R as exterior foam.

There's really no point to using MemBrain in this wall since you have enough exterior R for dew point control. If you were a bit shy of the IRC prescriptive levels it would be worth it from a technical point of view, but you're already building with more margin than the IRC requires for being able to use latex paint as the interior vapor retarder. And even if you weren't, a sheet of 1" iso to buy you even more margin is a heluva lot less money than an 800 square foot roll of MemBrain (for an 18 square foot project?) Air tight wallboard is more than adequate here.

I have plenty of EPS to do a double layer: maybe a triple layer for no more area than I have to cover.  I also have some 2" XPS remains in the garage... not sure if it's enough to do it all; but I could prioritize it to the fiberboard areas if that is worthwhile. 

... I also currently have 50 4x8 sheets of 2" foam control EPS stacked up in the old part of the house; but that's for a different project

http://i3.photobucket.com/albums/y8...46c011.jpg

If you put the 2" stuff over the fiberboard sections and half-inch over the pre-existing iso and seal it all up with foam you'd be just fine.

I checked and actually have more than enough 2" XPS do fill the necessary space.  Should I cut 2" XPS to a loose fit and then 3 layers (1.5") of EPS to completely fill the cavity (canned spray foam around the edges); or is that unnecessary?  I don't see much benefit in adding 1.5" thick FG over the XPS; but I may be wrong.

If you split an R13 batt in to two ~2" thick batts (or split an R19 into two ~3" thick batts) and compress it into the 1.5", at that higher density it will have the same R/inch as EPS, and with a compression fit it would eliminate unimpeded convection currents, and perform quite well. If you did it with EPS it might stick out a bit proud of the studs and make a less-flat surface. But it's your call.

Air sealing the edges of the foam to the framing is key, either way you do it.

I didn't realize you could compress FG and still maintain (or enhance???) its effectiveness.  I had always been taught in the past that FG relies on the air space between the fibers to work properly; which is why it's not "better" to put 2x6 batts in a 2x4 wall.  Is that thinking wrong; or does the cavity being air sealed change things?

Posted By strawmyers on 05 Jun 2013 07:58 AM
I didn't realize you could compress FG and still maintain (or enhance???) its effectiveness.  I had always been taught in the past that FG relies on the air space between the fibers to work properly; which is why it's not "better" to put 2x6 batts in a 2x4 wall.  Is that thinking wrong; or does the cavity being air sealed change things?

The compressed batt charts from the manufacturers tell a more nuanced story. Yes, fiber insulation relies upon the air for insulation but with increasing density comes lower air-redardency, lower convective losses, and higher performance, up to a point.  An R19 batt is so low density that even though it's R19 at it's manufactured 6.25" nominal loft, it performs at only R18 when compressed into the 5.5" cavity for which it is designed, and it performs at R13 in a 2x4 cavity (just like an R13 batt- go figure.)  That means it's R/inch goes from (R19/6"=) R3.2/inch at full loft but rises to (R13/3.5"=) R3.7/inch at the compressed density.
 
But an R13 batt (or a full R19 batt) stuffed in to a 2x3 cavity (2.5" instead of 3.5") performs at R10, improving it's R/inch numbers from R3.7/inch to R4.0/inch.

At some density the R-value begins to decline, but not until it's the hard-rolled shipping density.

If you read the manufacturers' recommendations about not compressing batts clearly, what they're really saying is that if you compress the batt (even a little bit), the total R  will be less than the labeled R value.  But that is due to the reduced thickness, not reduced R/inch performance. An R19 stuffed into a 2x3 cavity won't perform at R19, but it'll be more than half that, despite being compressed by more than half. 

Labeling a 6.25" thick R19 is bordering on fraud, IMHO, since it doesn't perform at that level when installed in 2x6 cavities perfectly, per the manufacturers' instructions. (It may perform at that level installed at 6.25" of loft in an attic floor condition, provided there is an air barrier on the top side.)  An R19 is just a fluffed-out R13- it's exactly the same amount of material, just spun for a higher loft.

Interesting information!  Well then that brings a question to mind:  I had planned to air seal each cavity of my future garage really well and then have wet-blown cellulose installed.  It's 2x6 studs.  The wet-blown cellulose would be ~R-20... looks like I can stuff "R-30" batts into the same cavity for an R-21.  They have the unfaced 25' long R-30 batts on sale at the big box store for dirt cheap sometimes (even less expensive than the kraft-faced R-13 batts).  I bought 48 rolls to add to my "R-19" attic a couple of years back.  If I'm air sealing anyway, could get a similar R-value with the compressed FG as the cellulose for a good savings in cost.  Any reason that wouldn't be suggested?  Will be 5/8" OSB on inside of walls: not drywall.  Unfortunately I can't add a 1.5-2" layer of EPS on the outside of the garage because it has to line up with the existing structure for continuation of the siding.  (I guess I could if I took all of the side off of two sides of the house, removed the windows, added the same to those walls, reinstalled the windows and siding... got new siding for the ends because the original pieces would now be too short, etc... but that's not going to happen).

The difference between R21 cavity fill and R20 cavity fill has almost no effect on the "whole wall-R" after thermal bridging is factored in. The more even-fill and higher air retardency of the cellulose vs. mid-density fiberglass more than erases any advantage the nominal R1 at center cavity gives you. But if it's substantially cheaper and you take the time to obsessively tuck every corner/edge, split & trim the batts over wiring & plumbing etc for as perfect a fit as can be done, you CAN get the performance out of them.

Without exterior foam you'd be looking at adding an interior vapor retarder, and going full-on with air sealing the interior gypsum as well. A layer of 1/4" fan-fold XPS siding underlayment with facers, seams taped and edges sealed under the gypsum would be about the right permeance (0.6-0.8 perms depending on manufacturer and facer) and is easier to air seal than the gypsum itself. Before installing the batts foam-seal the back side of electrical boxes with can-foam, or they'll leak air even after you've caulked the face to the XPS or gypsum.

The difference between R21 cavity fill and R20 cavity fill has almost no effect on the "whole wall-R" after thermal bridging is factored in. The more even-fill and higher air retardency of the cellulose vs. mid-density fiberglass more than erases any advantage the nominal R1 at center cavity gives you. But if it's substantially cheaper and you take the time to obsessively tuck every corner/edge, split & trim the batts over wiring & plumbing etc for as perfect a fit as can be done, you CAN get the performance out of them.

Without exterior foam you'd be looking at adding an interior vapor retarder, and going full-on with air sealing the interior gypsum as well. A layer of 1/4" fan-fold XPS siding underlayment with facers, seams taped and edges sealed under the gypsum would be about the right permeance (0.6-0.8 perms depending on manufacturer and facer) and is easier to air seal than the gypsum itself. Before installing the batts foam-seal the back side of electrical boxes with can-foam, or they'll leak air even after you've caulked the face to the XPS or gypsum.

No gypsum: 5/8" OSB for the lower 8' and corrugated metal (white pole barn siding) for the upper 7' and ceiling.  The wall cavities at the exterior 7/16" OSB, top, and bottom plates will be fully air sealed with with acoustic sealant.  Fully air sealing the interior metal siding is likely an unreasonable goal.  At ~1500 ft^2 of wall space, the fan fold stuff might get rather pricey... any other options that would be more palatable to the budget?  Asphalt-coated craft paper with seams Tyvek taped?  I know the interior 5/8" OSB acts as some level of vapor retarder. 

Would I still require/want an interior vapor retarder with the wet-blown cellulose?  I had read at some point in the past that an interior vapor barrier should never be used with cellulose; but I realize that generalized comments such as that probably are not true in all situation.  There will be Tyvek or similar house wrap over the exterior 7/16" OSB.  I know that is not a vapor retarder; but thought it should be noted in case it factors into the equation.

With the interior sheathed as OSB it's already about 1- perm when dry, a minimal class-II vapor retarder, but becomes more vapor open should the OSB become damp, making it something of a "smart" vapor retarder. As long as it's carefully air-sealed and you don't have big moisture drives from the garage (got a hot tub in there? ;-) ), it'll be fine.

You'd probably still want something under part finished with the corrugated metal. The smallest rolls of Certainteed MemBrain (designed as a smart vapor retarder) are 8' x 100'. Poly would be a lot cheaper (and easier to find), but whether that's the right way to go depends on the exterior stackup.

Fan-fold XPS costs between 20-25 cent per square foot at box stores (comparable to sale-priced box-store 7/16" OSB, cheaper than full-retail OSB), and you'd also need a few rolls of housewrap tape for sealing it:

http://www.homedepot.com/p/Owens-Corning-Foamular-1-4-in-x-4-ft-x-50-ft-Fanfold-10UM/100320301#.UbCbE5ywW

The limited number of seams to seal and the inherent rigidity/ruggedness make it the preferred air & vapor barrier/retarder under the corrugated if it's in the budget.

It's true that you don't usually want to use a true Class-I vapor barrier like poly or foil with cellulose, but something in the 1 perm range is fine. Cellulose can buffer a lot of moisture, but you also want it to be able to dry in a reasonable amount of time when conditions are right. Under air-tight OSB or ~0.6-0.8 perm fan-fold it'll be fine.