Posted By Lbear on 31 Dec 2011 10:56 PM ICF forms typically do not have any buffering capacity of leakage, so even a small leak, that may occur undetected with no durability risks in a wood framed wall, may affect the interior of and ICF building.

My take, ability to soak up water. In a frame wall if you have a roof leak, for example, the water very well may soaked into the top plate before it ever reaches the interior drywall. With ICF the foam won't soak it up so any water leaking through the roof can just run on over and down the foam and into the drywall.

I'm inclined to agree it is a possible issue, but from a practical standpoint it is a very unlikely issue. I think it's mostly a case of excess caution on the part of the report writer. I would dismiss it as a genuine concern. Theoretical, yes, realistic, no.

Posted By Lbear on 31 Dec 2011 10:56 PM
They claim that ICF will require other insulation strategies in cold climates, why is that?

I don't recall without going back and digging through their book and other documents on the web site, but I believe BS recommends R values in excess of 30 for cold climates. Most ICFs are on the order of 20. If you want high R values, which the page you link to alludes to as R 40, you will need to add insulation to the typical ICF wall. If you add 4" of foam to the outside of an ICF wall with 2 1/2" foam, you'll have a lot of foam to go through to attach siding, trim, etc. That will complicate the construction process. There are ICF blocks available that will get you an R 40 wall now, more so than when that document was written.

To figure out just what R value you really should go for you need a competent HVAC designer to do a good heating/cooling load calculation. Part of that calculation can be a cost benefit analysis of increasing the R value of the walls above a typical minimum of 12 to 15. That calc is done for your climate using the weather data for your area to select the minimum and maximum design temperatures.

I don't recall what, if anything, you have said about working with an HVAC contractor, but if he even suggests the idea of "rule of thumb," cut his thumb off. You want a bona fide Manual J analysis. Pay for it separately from an HVAC bid if at all possible. It'll be more free of sales influence that way.

This is pure speculation, but it seems to me that if someone wanted R40 walls, something like ICF panels filled with cellulose and then concrete applied to the exterior (like a SCIP) would be interesting. As with any SCIP, would need some type of wire mesh on the outside and through the blocks. Concrete could be sprayed on or cast in place with forms. Unlike ICF, the interior and exterior finish might be usable as is.

Posted By dmaceld on 01 Jan 2012 12:51 AM

Posted By Lbear on 31 Dec 2011 10:56 PM ICF forms typically do not have any buffering capacity of leakage, so even a small leak, that may occur undetected with no durability risks in a wood framed wall, may affect the interior of and ICF building.

My take, ability to soak up water. In a frame wall if you have a roof leak, for example, the water very well may soaked into the top plate before it ever reaches the interior drywall. With ICF the foam won't soak it up so any water leaking through the roof can just run on over and down the foam and into the drywall.

I'm inclined to agree it is a possible issue, but from a practical standpoint it is a very unlikely issue. I think it's mostly a case of excess caution on the part of the report writer. I would dismiss it as a genuine concern. Theoretical, yes, realistic, no.

Lbear,  I'm going to research this a bit and get back to you.  First, however, you might want to start a new thread cause tacking this on is just going to further dilute your previous thread with more unrelated stuff and unrelated comments.

Anyway,  I took a quick look at the report http://www.buildingscience.com/docu...l-analysis

and it strikes me that these folks don't know a great deal about ICF.  But I also may not be looking at the correct report.  The first clue was the ICF they are using either doesn't exist or at least is not representative of 99% of the market.   They used an ICF with dimensions 2" foam, 4" concrete, 2" foam (8" overall thickness) and an R-value of 16.   To my knowledge this ICF doesn't exist.   Your standard ICF with a 4" core is 9 1/4 o 9" thick with an R-value of 22 and actually most of the market for ICFs is actually 6" core (also roughly R22).    

Regarding the buffering capacity -- I don't buy much of that at all.   The last thing you want in a wood wall is a water being absorbed into the wood that you know nothing about so that it doesn't inconvenience you.   Mold comes to mind but repeating this a number of times (like with a small leak you don't know about over time) will eventually cause failure due to rot.   This may not be the case for a non-wood wall.  Nevertheless, I'm from the school that if a leak exists it's better to know about it today than to find out after your wall or roof fails.   That said, with ICF, and I don't think it's somehow a "feature" if you put a dripping faucet on your ICF window ledge or top of wall the concrete will absord a huge amount of water (like a sidewalk) until it is saturated and startes to leak out. 

Water causes far more damage to conventional walls than ICF walls.   For example -- in some of the recent floods where both ICF and conventional homes were completely soaked (flooded) -- the ICF walls remained intact and sheetrock was replaced sometimes to 4-5 feet high -- the wood wall homes (as well as interior walls in ICF homes) were typically taken down completely to the slab.    Regards. 

P.S.  I did find a second report (more closely associated with your link) that showed 2.5 inch foam and R-20 continuous insulation.