Exterior polyiso or XPS for wall and roof assembly?
I have plans for construction of a 15 x15 foot addition to my home in climate zone 5. The walls are currently designed as 6-5/8 inch Murus OSB polyurethane SIPs with adhesive spray foam in the panel joints, All Weather Flashing tape on the inner seams and Grace Vycor adhesive on the outer seams. Tyvek Drain wrap is over the SIPs to act as a small air gap. The drain wrap is covered by 2 inches of foil faced polyiso with 1×4 furring added to support UNA-CLAD steel siding. The entire wall assembly is roughly R 53 or so.
The sloped roof is also 6-5/8 inch SIP again with adhesive spray foam in the panel joints, All Weather tape on inner seams but with Grace Vycor covering the entirety of the external SIPs as per a PERSIST design. I then have 4 inches of foil faced poly iso on top of the Grace Vycor adhesive with 2×4 furring over the 4 inch poly iso then 5/8 inch CDX plywood, #30 felt and UNA -CLAD standing seam metal roof. Roughly R 66 for assembly.
My architect is now telling me that she has reservations about having poly iso exposed to possible water in both wall and roof assembly and thinks that it would be wiser to use XPS for both assemblies as she believes XPS to be 10 times less absorptive of water compared to poly iso. She also is making a case that poly iso is less stable than XPS, degrades in R value, and shrinks more than XPS. She is also making the argument that the permeance of foil faced poly iso is so low that the walls will not be able to dry to the outside and that XPS would have higher permeance and therefore better for that reason as well.
As for the roof she also now has concerns that nailing of insulation and furring strips through the Grace Vycor adhesive will compromise the water proofing function of the Grace Vycor. She is now suggesting for the roof the usage of 4 inch wide metal Z clips to hold the insulation over the Grace Vycor and to mount the furring strips onto the Z clips so as to keep penetration through the water proof adhesive to a minimum.
My questions are the following: If her statements about the inferiority of poly iso to XPS are true with regards to moisture and durability issues why would one ever use poly iso as an exterior insulation? Not with standing her concerns, I have seen many pictures ,and read many articles on GBA , in which poly iso has been used as the exterior insulation with furring strips added on both wall and roof assemblies. Are her concerns about the poly iso legitimate and should I be using Z clips for the roof?
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
First, is there any way to design out the polyurethane SIPs?
Almost all 2lbs/cubic foot or higher density polyurethane in the US is blown with HFC245fa, which is a very powerful greenhouse gas (more than 1000x CO2). Similarly, XPS is blown with a mix of HFCs, dominated by it's HFC134a component, (the most common automotive AC refrigerant) which runs about 1400x CO2. (In Europe XPS is blown with CO2, making it comparatively benign, but it has the same R/inch as EPS, which would be a marketing problem in the US.)
By comparison, polyiso is blown with pentane, at about 7x CO2, as is EPS. From a total verde-tude point of view XPS, and mid to high density polyurethane are on a completely different planet than EPS and polyiso. When all environmental factors get factored in, open cell polyurethane (blown with water) is generally considered the greenest per unit-R, followed by polyiso, followed by EPS. But they're all over on the green end of the scale (at least in foam-terms.) XPS and 2lb polyurethane are the stuff of brown fields, to be designed out or minimized. There are lower impact blowing agents out there (Honeywell Solstice) and some 2lb foam is blown with water (Aloha Energy, Icynene MD-R-210), but unless Murus will stipulate that their SIPs are made with foam blown with a blowing agent with a GWP less than 10x C02, you aren't doing the planet any favors in a high-R stackup.
It's true that polyiso is far more hygroscopic and vapor-permeable than XPS (that's why they sell it with facers), but it has a long and successful history as both roofing insulation and as insulating wall sheathing. It will take on moisture slowly in a flood, but will eventually dry out after the tide goes out, the rate dependent on the vapor permeance of the facers. Most roofing iso is sold with asphalted paper or fiberglass facers rated crudely at under 1-perm, but 1 perm is still almost two orders of magnitude higher than foil or 6 mil poly, EPDM, etc. Roofing iso is typically 2lbs density (unless you specify higher density) and is more rugged than the more commonly seen 1.5lb density foil faced goods. (It's usually about 25% cheaper per unit-R than XPS too.) A single layer (any thickness) of asphalted paper faced iso is about the same vapor permeance as 2" of XPS. It's fine to use 2, even 3 layers of roofing iso if needed- the decking or sheathing can still dry over the course of a season. At 4+ layers drying times potentially overlap with annual periods. With any rigid foam product it's better to use at least two overlapping layers to thermally break any gaps that open up at the seams with age/temperature/humidity.
It is also true that with foil facers there is no drying through the foam, and you have to pay attention to the stackup & climate to keep the OSB dry enough. Assuming the PU SIPs have about 5.5" of foam in them, the permeance of the SIPs is only about 0.2 perms, which is both a long loading time as well as a long-ish drying time. As long as you have sufficient exterior R relative to the ~R35 of PU in the OSB sandwich to keep the exterior skin's mid-winter average temp above ~35F you're golden. For a zone-5 location that will take more than 2" of polyiso. But it doesn't take a lot more- 3" will get you there.
I'm not sure how you get to R53. Murus specifies R41 for the center-panel R in your wall stackup, and even without accounting for the thermal bridging that seems optimistic. With only about 5.5" of foam in the core that would need R7.3/inch foam, which simply is not an aged-value of any density polyurethane. It probably performs at that level for the first 10 months or so, but it declines to about R6/inch fully depleted, maybe R6.5/inch as an average on a presumed (fairly short) 25 year lifecycle basis, and if they were selling insulation rather than a wall-system they could be called on the carpet legally with that claim. There is also far more thermal bridging in a SIP than some of the marketing numbers let on (they all have top & bottom plates, window & door framing, structural splines, etc.) It may be academic, but if your "whole wall" R is over R45 even with the 2" of polyiso after carefully accounting for the structural timbers I'd be a bit surprised. An R45 wall is pretty good wall, but there are both cheaper and greener ways to get there.
Dana, thank you very much for your very thoughtful and informative reply. I am in a position to reconsider the SIPs and based on your information will definitely consider that option. You have done your duty for the planet today
Based on your response it seems to me that using poly iso as exterior sheathing for my walls and for my roof is very feasible, workable, and safe particularly with asphalt or fiberglass facing rather than foil so that drying can take place to the outside. As stated my plans also call for a 3/4 inch air space behind the wall veneer and 1 and 3/4 inch for the roof which will help greatly in that regard.
Should I use asphalt faced poly for both walls and roof? Based on permeance issues alone (ability to dry to the outside) I don't see any situation in which foil faced poly iso should ever be used externally. Am I correct in this assumption?
Timothy,
Q. "If her statements about the inferiority of polyiso to XPS are true with regards to moisture and durability issues, why would one ever use polyiso as an exterior insulation?"
A. Polyiso is the most common type of rigid foam insulation used for roofing installations. It has an excellent track record, going back decades. It is the most environmentally friendly type of rigid foam available in the U.S.
Q. "Should I be using Z clips for the roof?"
A. Grace Vycor seals around fastener penetrations, and you don't have to worry about the integrity of the Vycor as an air barrier, even with many fasteners penetrating the Vycor.
My main concern with your roof design is that it doesn't really follow PERSIST principles. When a peel-and-stick membrane is attached to the sheathing of a PERSIST roof, the roof sheathing is able to dry inward if it ever gets wet (because the rafter bays of a PERSIST building have no insulation).
You are choosing to install peel-and-stick on top of a SIP roof, and then install more foam above the peel-and-stick. This isn't really a PERSIST approach. The upper OSB facing on the SIP will be trapped between two vapor-impermeable layers. I know that this is commonly done with SIP roofs, but it isn't quite as resilient as a PERSIST design. You'll need to cross your fingers and hope that the upper OSB of the SIP never gets damp.
It's fine to use foil-facers as the exterior polyiso layer, as long as any susceptible materials (like OSB or plywood, fiber insulation) in the stackup have a path for drying at reasonable rates to the interior, and are located at a layer where they won't take on excessive wintertime moisture.
On roofs there are several ways to deal with it- some installations have an un-vented non-structural OSB nailer layer between the iso and the roofing (through-screwed to the structural deck and rafters), others will mount the nailer on vented purlins or furring parallel to the rafters. With the unvented approach it's presumed that when you re-roof a few decades at least some of the OSB may need to be replaced.
I'm not sure if all of the fastener retention specs would work with a SIP roof unless there are regularly space splined in the SIP to timber-screw to. There are a number of vendors of both vented and unvented nailbase polyiso for roofing applications where the iso is bonded to the nailer deck layer (sort of a half-SIP, in 4x8' or 4x10' panels). They go up quicker, but have a higher material cost than, a separate roofing-iso + nailer approach, eg:
http://www.hpanels.com/index.php/flat-products-sp-714077411
http://www.atlasroofing.com/tabbed.php?section_url=58