What if it is recylced eps

Posted By benman on 24 Apr 2013 06:15 PM
Hi Dana, thanks for the advice. My plan is to add 4 inches of foam on the exterior walls. Are there any issues with putting poly-iso on the exterior of the main level (with furring strips and cladding) and eps on the basement level (with eifs)? Obviously with some kind of trim detail to separate the two claddings. Your roof insulation and air sealing recommendations were what I had in mind. Thanks!

The issues with a foam-over on an ICF are the lengths of the fastners.  Assuming you have 2.5" of EPS in the ICF, and 4" of polyiso, with 1x furring, and a requirement of at least 1.5-2"  into the concrete, you're looking at some pretty substantial ~10" masonry screws.  Mind you, adding R25 in exterior foam over an R22 ICF also puts you at near-PassiveHouse performance on the walls.  From a practical standpoint cutting down to 2" of additional iso is probably going to be more buildable, and even then finding the right screws may still be a challenge. 

It may be both easier & cheaper to go with an assymetric ICF than adding exterior iso when it's all said and done. The devil is in the details.

What if you have a way of attaching the foam to the foam without fasteners. The only thing that this would do is cut your exterior finish to stucco of some sort. If you used a standard icf you would have a web every 6 to 8" that is 5/8 to 3/4 below the surface for attaching the foam board, so you wouldn't need long screws to attach the foam board. My preferance would be no screws.

No screws is great, but it still needs to handle the weight of both the foam and siding over time, and under wind-loading.

Dana, maybe I missed this, but why do the screws need to penetrate 1.5-2" into the concrete? I would certainly agree that 7.5" of screw moment arm would be excessive. Most ICF systems have ribs/studs close to foam exterior for placing the siding.

Posted By sailawayrb on 25 Apr 2013 12:27 PM
Dana, maybe I missed this, but why do the screws need to penetrate 1.5-2" into the concrete? I would certainly agree that 7.5" of screw moment arm would be excessive. Most ICF systems have ribs/studs close to foam exterior for placing the siding.

Wind loading and dead-weight of "real" siding is going to be excessive.  SFAIK ICF system ribbing isn't designed to handle the moment arm of 4" of screw through the foam + 3/4" of furring pluse the heft of hard-stucco, fiber cement or even wood siding.  You might get there with vinyl, but no matter what you did it needs to be run by the manufacturer's engineers, and maybe even a local engineering company if the inspectors are sticklers.  The ribs are usually heavy piece of HDPE or nylon, no?

If you really think you need to increase the R value of ICF the simplest is to just apply eps with eifs cement the same way the acrylic stucco boys apply eps over wood strata and then just acrylic stucco it.
If you want to do siding a reasonably simple way is to adhere Nudura or similair panels to the existing ICF with eifs cement and then screw/nails siding into the Nudura ribs. (not necessarily the cheapest)

Posted By Dana1 on 25 Apr 2013 11:04 AM
What do you need a vapor barrier for, if it's on the inside of the ICF? 

By adding more insulation to the interior you're isolating the interior from the thermal mass benefit, but it's still a net win.  It's far better from a performance point of view to add the additional R to the exterior, and keep the interior-side R at a minimum.

The vapor barrier inside is to isolate that wall cavity from the human generated moisture inside the home.  In theory the inside ICF insolation should not drop below dewpoint, but if it does, say in extreme new york winter climate, the possibility exists that it will collect moisture if there is a source.  As I understand it, I don't think you will get much cure moisture thru the EPS.  The cavities are not vaporbarriered at the top, so would/should attain the same moisture level as the attic space if no additional moisture is added from the living space.

I agree that it would be a little more efficient with the insulation on the outside, it is just easier/less time consuming to fabricate on the inside IMO. 7"-10" sheething screws with 2+" of concrete penetration to meet wind and shear loading specs kinda goes along with this theory:) 

Posted By Ronmar on 25 Apr 2013 02:25 PM

Posted By Dana1 on 25 Apr 2013 11:04 AM
What do you need a vapor barrier for, if it's on the inside of the ICF? 

By adding more insulation to the interior you're isolating the interior from the thermal mass benefit, but it's still a net win.  It's far better from a performance point of view to add the additional R to the exterior, and keep the interior-side R at a minimum.

The vapor barrier inside is to isolate that wall cavity from the human generated moisture inside the home.  In theory the inside ICF insolation should not drop below dewpoint, but if it does, say in extreme new york winter climate, the possibility exists that it will collect moisture if there is a source.  As I understand it, I don't think you will get much cure moisture thru the EPS.  The cavities are not vaporbarriered at the top, so would/should attain the same moisture level as the attic space if no additional moisture is added from the living space.

I agree that it would be a little more efficient with the insulation on the outside, it is just easier/less time consuming to fabricate on the inside IMO. 7"-10" sheething screws with 2+" of concrete penetration to meet wind and shear loading specs kinda goes along with this theory:) 

This theory is all wet! Concrete is VERY tolerant of moisture, so WHAT if it takes on some human-activity moisture?

Wood is not tolerant of moisture.  But do the dumbed down dew-point math:

The dew point of 70F 35% RH air (comfortable healthy winter-time conditions) is about 40F.  The coldest wood in the assembly is the exterior stud edge, which is about R4 from the interior.  That's a difference of 30F degrees over R4, or about (30F/R4=)  7.5F per R.  For the edge of the stud to actually reach the 40F dew point with R20 on the exterior, it has to be (R20 x 7.5F= ) 150F colder outside than that 40F stud edge.  Are you expecting to see sustained sub -190 F outside conditions, a condition not even seen in NY during the last ice age!? 

But that's how cold it has to be for the stud edge to begin adsorbing moisture!

At center cavity you have about R13, or about (30F/R13=) 2.3F/R.  So for a moisture haze to form on the EPS/fiberglass interface requires a sustained outdoor temp of (2.3F x R20=) 46F below 40F, or -6F degrees.  Yes, it gets that cold sometimes, but even through 2-5 perm standard latex paint on the gypsum it would take WEEKS of temps below -6F for visible liquid to form.  If you had some real air leaks convecting from the inteior into the cavity you could conceivably get a visible-liquid forming on the foam in only a few days, but really, when was the last time the daily highs were below -6F for days on end?  And even if it did occur, both the foam and the fiberglass are undamaged by condensation, and it will dry quickly when it warms up.  When it's well into negative digits just the ventiliation air would normally cause the interior air dew points to drop too. Only if you're activaly adding humidity to the conditioned air would you ever get even these minor, not destructive wetting events.

Let the paint be your semi-permeable vapor retarder, since that allows moisture to get out, keeping the wood at the conditioned space moisture level, never trapping moisture.

Type II EPS is about 3-4 perms at 2.5" thickness, about the same as latex paint, so when the concrete is releasing moisture it still won't swamp the cavities with moisture at a rate faster than it can dry to the interior.  But kraft facers are about 0.4 perms, and 6 mil poly is under 0.1 perms, and you WILL see elevated moisture in the wood.  During the air conditioning season you can even get condensation on the interior of poly vapor retarders in an air conditioned house, even in Canada. Just don't do it!

The IRC spells out the amount of exterior R needed to avoid damaging levels of moisture on wood sheathing in studwalls by climate zone when using only class-III vapor retarders (like latex paint) as the interior vapor retarder, and even in Fairbanks AK a minimal R16 ICF would be more than sufficient.

Any time you add a low permeance layer like a vapor retarder, you magnify the moisture differences, and in this instance you'd be parking the susceptible wood on the wet-side of the vapor barrier where the concrete moisture-reservoir lives.

The notion that you can get sufficient vapor diffusion through the top plates of a studwall that the cavity would run at the same moisture levels at the attic is similarly off-base. The vapor permeance of 1.5" of wood is about 0.6 perms for single-thickness 2x plate, or 0.3 perms for a double plate, an order of magnitude more vapor-tight than that of the inner EPS of the ICF. And vapor diffusion is a vapor-pressure x surface area thing, and the area exposed to the concrete through the more permeable EPS is about 30x the area of the stud plate. The top plate might as well be true vapor barrier (and it almost is), for the amount of drying it is offering here.

Bottom line, a vapor barrier here only increases the likelihood of mold & moisture problems in your stackup.  A vapor barrier is a very sharp 2 edged sword, wield it carefully, and only when necessary.

So Dana, what do you think about a claim like this from build block:

All furring strips are 15" tall by 1.5" wide per 16" course and have two distinctive marked (sweet) spots on each furring strip at 8 inch centers vertically and 6 inch centers horizontally. These spots have a screw pull out strength that is roughly equivalent to wood or structural steel studs (close to 500lbs).

If this is true wouldn't it be able to hold 4inches of foam furring and siding just like wood?

there was a passive house in wisconsin that had 11 inches of eps on the outside of the icf's. they used eifs and it looked like they glued the foam on.

Interesting, thanks for the math lesson Dana

Posted By Ronmar on 25 Apr 2013 11:44 PM
Interesting, thanks for the math lesson Dana

Yeah, you're welcome! Didja catch the part where I subracted 150F from 40F and got -190F (instead of -110F)?  Not that it changes the conclusion...  

Vapor barriers are the bane of the construction industry, and should be designed out whenever possible, and in your stackup it's a no-brainer.    Even in the coldest NY climates it's possible to design it out with exterior foam, and in the warmer NY climates it can be designed out with back ventilated or moisture tolerant sheathing.  Viewing the ICF part of the stackup as "sheathing", it has both the moisture tolerance AND more than sufficient foam-R going for it.

But if you want to take the full-on building science approach on it rather than the 2-D simple arithmetic, modeling moisture movement & moisture levels within ICF stackups with interior side studwalls is fully within the capabilities of the Fraunhofer Institute's well vetted WUFI simulation tool. It is available as a freebie download from the Oak Ridge National Labs website.  You can get it to spit out the moisture content of the studs over time starting out with the freshly poured concrete, as well as over the historical weather data for your location and building orientation/shading factors, etc.  You'd have to simulate both with and without the vapor barrier, then compare.  It's not "Hygric Modeling for Dummies" though, and it's easy to fall into garbage-in-garbage-out syndrome if you're not careful.

My original statment refered to the ability to increase the R-value of an icf system by using more foam in the wall system itself as well as going to a different foam like iso board. This isn't a original idea, I got it from looking at a You Tube video of a British compay called Polar Wall. Sure liked their scaffolding set up like working on the sidewalk

I would at least look at putting as little of the building underground as possible and using a frost protected shallow foundation/slab (no footings). Sure, some of your building extends back into a hill, but that doesn't make FPSFs not work.

jonr, the king of fpsf. If you pour the footing, walls and slab in one step you are essentially doing the same thing at a far lower cost.

We need to recognize that foundations do more than provide frost protection. Their primary purpose is to support the building.
Consider that if you are building into a hillside, you have an unbalanced load on the foundation. Now if you put marbles (or drainage rock) under it and then put foam between the marbles and the slab to make sure there is no possibility of any friction at all and then backfill the over excavation on the hillside with loose fill, what is going to happen when the rain runs down the hill and saturates the loss fill?

Posted By Dana1 on 25 Apr 2013 12:36 PM

Posted By sailawayrb on 25 Apr 2013 12:27 PM
Dana, maybe I missed this, but why do the screws need to penetrate 1.5-2" into the concrete? I would certainly agree that 7.5" of screw moment arm would be excessive. Most ICF systems have ribs/studs close to foam exterior for placing the siding.

Wind loading and dead-weight of "real" siding is going to be excessive.  SFAIK ICF system ribbing isn't designed to handle the moment arm of 4" of screw through the foam + 3/4" of furring pluse the heft of hard-stucco, fiber cement or even wood siding.  You might get there with vinyl, but no matter what you did it needs to be run by the manufacturer's engineers, and maybe even a local engineering company if the inspectors are sticklers.  The ribs are usually heavy piece of HDPE or nylon, no?

Yes, the siding loading (weight, wind, seismic) must be considered.  I am not familiar with the SFAIK ICF system design strength.  However, most ICF system ribs are adequately strong to place most siding without issues.  Clearly, adding the additional insulation which results in having a 4” moment arm is an additional complexity and potential issue.  Having many steel fasteners providing a 0 R-factor path directly from the exterior siding to the concrete core might be something to consider too.  Some ICF systems will customize the insulation thickness and rib location to your requirements for relatively little additional cost.

It's true that the conductivity of steel fasteners is high, which is why most foam-overs on stick built are done with 24" o.c. spacing min, whenever it can meet the load. With ICF you get at least an R10 thermal break from the interior foam, and with judicious spacing specifications you'd get most of the benefit of the additional R.

Rather than adding extra insulation over your ICF and worrying about getting an exterior foam that will not absorb moisture, furring it for attachment, etc. Why not use a higher R-Value ICF? Take a look at the Quad-Lock R-30 system, it is minimally more expensive than our R-22 and R-28 options. It gives you 4" of eps on the outside and 2" on the inside.

You could also use a system like TF, which allows you to increase the depth of foam while leaving the furring just below the surface. You could do this at a cost of $3.50 to $4.00 per sq ft for an R-30 plus the cost of concrete and rebar.