Side-by-Side Home Test Underway

Shortly after the Greenbuild 2010 tradeshow concluded in Chicago, two Habitat for Humanity homes were built by volunteer crews in the nearby suburb of Waukegan, Ill. One was built with LOGIX Insulating Concrete Forms (ICFs) having a nominal 6” core.  The other used 2x6 wood-frame walls (24” on center) with four inches of open-cell spray foam insulation, and an additional 1-inch of XPS foam board sheathing wrapped with an air infiltration barrier—well beyond what is required by code. 

Both homes currently are on track to achieve LEED for Homes Gold certification. Kouba-Cavallo Associates Inc. was hired to perform blower door tests on both homes.  They report that the ICF house registered 1.15 air changes per hour (ACH) at 50 Pascals, while the lumber-framed house registered 2.76 ACH.  The ICF home was 2.4 times tighter than the lumber-framed home. —even with its spray foam insulation and exterior air barrier.

The USGBC is monitoring actual energy performance data for both homes; the data will be made available after a full 12 months of data has been collected. When the 12-month study is complete and results are available, ICF Builder will revisit this story.

ICF Builder Magazine

What do you surmise the outcome will be???

PLEASE keep all of us posted. I really do hope that ICF Builder Magazine maintains a focus on this storyline and provides reports on the detailed analysis that I hope will come from the testing. Any chance we can arrange for a "side-by-side" comparison of ICF vs. CMU?

So a well sealed vs not so well sealed comparison? I'm more interested in why the builders of the wood frame weren't able to get it well sealed. Ie, where are the leaks and how could they have prevented them.

ya no kidding. It is already a one sided study from the start. There are too many variables. They need to make the wood framed sealed to match the ICF so then they can compare the performance of ICF vs the sticks especially come winter. 2.76 ACH with spray foam and exterior sheathing? I guess thats volunteers for ya

The Habitat volunteers who built the Logix walls came from a trade group next door in Skokie called the Portland Cement Association. http://www.concreteconstruction.net/portland-cement/helping-hands-build-icf-home.aspx Unless the volunteers on the stud comparison home happen to work for the American Wood Council, this comparison badly needs an asterisk. (* There are volunteers and then there are "volunteers.")

Before you guys get too carried away... 2.76 is pretty good for stick frame.

And 1.05 is not that good for an ICF home...


If you take away this ACH difference - what are you really testing - thermal mass? that would not be fair. ICFs by nature are tight. Wood is not.

2.76ACH/50 would barely meet IRC2012 code (3.0ACH/50 max, for US zone 5), but it does prove the folly of using cavity insulation as the primary air barrier (and the relative ease of air sealing SIP & ICF). (The article didn't specify the "exterior air barrier" or how it was detailed. House-wrap at "typical" installation method would not qualify, even though it can be when properly detail. At nearly 3ACH/50 you can bet details were missed.)

Timber framed construction has many leak points to be attended to that apparently weren't, in this case. But the Canadian R2000 program has proven that it's pretty easy to beat 1.5ACH/50 cost-effectively in timber framed construction even in production builds. The fact that a group of volunteer Habitat for Humanity labor & designers overshot that mark by nearly 2x means very little about what stick builders with good designs and trained crews will actually build. As IRC2012 becomes more widely adopted, the knowledge base among stick builders is bound to increase.

This does feel a bit like another straw-man test- the standard 6" Logix has 5.5" of EPS for a whole-wall R of about R22.5 , whereas the whole-wall R of 2x6 construction with o.c. cavity fill and 1" of XPS on the exterior comes in at R19, about 15% lower in steady-state R value.

So if you're matching your favored product against something more than 2x as leaky and running 15% lower in raw R-value from the get-go, is anybody going to be shocked (SHOCKED!) when <...yawn...> it edges out the "competition".

Assuming similar or identical shading factors and similar occupant behaviors, I'd estimate the ICF house will use 7-9% less energy overall, but there are many behavior, shading, and plug-load factors that can create deltas that large, so it might be closer than we think. The air leakage rate hobble is a bigger issue than wall-R, both on cooling and heating energy use. In the picture (http://www.icfmag.com/images/back_issues/2012-05/news_01.jpg ) it appears that one of the houses is on a corner lot the other wedged between two houses, so both shading factors and wind-washing will differ. Whether that's sufficient to make a measurable difference depends on the orientation, which isn't clear, given that it was taken on a shadowless cloudy day.

Being built by volunteer labor and a lot of donated material it's hard to get to the true cost basis of either house.

BTW: In Chicago putting only 1" of XPS on the exterior of the stick built means it still needs an interior vapor retarder, but had they gone to 1.5" they could have skipped the interior vapor retarder, which is a much more sustainable way to build. Had they done so the whole-wall R values between the stick build & ICF would have at least gotten within 5% of one another, even if the ICF was still slightly higher-R. Just sayin'...

Posted By Baldwin2014 on 07 Jun 2012 02:29 PM
Before you guys get too carried away... 2.76 is pretty good for stick frame.

And 1.05 is not that good for an ICF home...


If you take away this ACH difference - what are you really testing - thermal mass? that would not be fair. ICFs by nature are tight. Wood is not.

Pretty good for a stick frame on On Mars, you mean?

Production houses with fiberglass batts shooting for the paltry IRC2009 standard of 7ACH/50 (required by code in many areas) often come in under 3ACH/50, which is why that became the benchmark for IRC2012.  It means that even less experience people are likely to be able to hit it, with only few post-construction tweaks if they happen to miss. The foam should have made it easier to beat 3ACH/50 by quite a bit,if the designers & builders had any talent for it, or had set a more rigorous tightness goal. At 2.76ACH/50 it only means that they gave it at least a nod, and no, it's not "pretty good", it's (soon to be) pretty much code-min tightness. I suppose IRC code min maybe IS "pretty good" if compared to when there were no air tightness standards and nobody cared about performance, but it's dead-easy to hit.

The R2000 spec is 1.5ACH/50 and is beat every day in stick built without much trouble, but it requires some experience on the part of the designers & crew.  It's pretty easy to detail wooden sheathing as a primary air barrier, which is more robust than relying on housewraps. Details like beads of caulk or adhesive between band joists & subfloors, and between subfloors & stud plates are dead-easy and cheap to implement, but won't be retro-sealed by using foam as cavity insulation.  We don't really have information about what air-sealing attempts were made here, beyond the oblique reference to "exterior air barrier", whatever they meant by that. (Was the XPS detailed as an air barrier? The housewrap? The article provides no clue.)

Take away the ACH difference you're still left with the 15% deficit on R value, which could have nearly the same difference on energy use than any mass-effect of a 6" concrete wall in that climate, depending on the rest of the R/U values.  Add it all up (<1/2 the ACH + 15% more wall-R + mass effect) and it could hit 10%, but I'm guessing it'll be slightly less of a total delta.  I wouldn't be shocked if made a 12% difference, but I'd be looking for a more careful explanation with data to back it up if it hit 15%.

[edited to add]

The unoccupied  performance delta of this side by side test (with heating & cooling setpoints in lock-step) was about 9% in favor of ICF despite 29% lower whole-wall R on the stick-built, but they started out with nearly identical CFM/50 numbers.  The Chicago-land test will likely come in similarly, trading a high CFM/50 delta for a lower wall-R delta. 

Either way, it's a straw-man test of little significance. The fact that DOE2 is pretty good about predicting thermal performance in either/both wall systems means that a designer can model a wall system to determine the true thermal performance with pretty reasonable tolerances.  But enough with doing side-by-side comparisons with straw-men for the stick built. Compare what it takes to hit R2000 compliance with ICF vs. stick built (and include cost numbers) next time.  The ORNL test and the Chicago test are both too flawed to provide meaningful information to home buyers on a budget, but with green dreams.

Dana,

I think this is a fair test.

How many states have adopted IRC 2012? I think 2. Ach of 3 is pretty hard to do on a wood frame home unless you do some advanced stuff. Average new construction is between 4 and 6.

I think the higher performance insulation (XPS and sprayfoam) on the wood home combined with an air barrier should make things comparable to ICF. I think it just goes to say that even spray foam will not make your wood home airtight. It has little to do with the builder and much to do with the construction materials. ICFs are just that good.

I dont know who built the ICF home but @ 1.15 they did average.

and what about your wood frame home of 2.76 - what do you think this value will be 10 years down the road. It will double. And how about your ICF home? It will stay nearly the same.

The fact that DOE2 is pretty good about predicting thermal performance in either/both wall systems...

Does that include the "mass effect"? I thought that was still relatively uncharacterized as it pertains to ICF.

Unless the wood framed home is a CUSTOM built home, with a specialty contractor who is well versed in energy conservation, the average built wood home will NOT achieve superior air tightness. At the same time, an ICF home by nature is air tight, even if built by a "sloppy" contractor. A wood framed home by nature is not air tight.

Even with the 2012 IRC, most wood tract homes will just have fiberglass batt insulation stuffed between the studs and then maybe 1/2" - 1" of EPS on the exterior, that's it. It will pass 2012 IRC codes simply by stuffing R-19 batts in between the studs and throwing on less than an 1" of EPS on the exterior. 2012 IRC does not take into account the thermal bridging of wood. So even though 2012 IRC recognizes it as R-20 code minimum, with thermal bridging, you will be around a R-14.

Let's not forget that as a wood home ages, with constant wind loads and the resulting racking, a wood home will become less and less airtight over time. With ICF/concrete, the air tightness will remain stable over the years.

I think the comparison is fair and the results will be interesting. Of course the wood proponents will cry foul and conspiracy. 

From my research, a decently built, well insulated new wood frame home has around 2 and 3 air changes per hour and after a many years it's often between 5 and 10 ACH as the wood racks, dries, shrinks, swells and sealants deteriorate.

A well built ICF home can has as low as 0.5 and 1.0 air changes per hour. (Usually the WOOD ROOF on an ICF home causes the greatest air leakage, then the doors and windows)

So in this case study I believe they are on target with their ACH results. With the wood home at 2.76 ACH50 and the ICF home at 1.15 ACH50, that is about average for and ICF home and above average for a wood framed home. Especially for the wood framed home, as it used 2x6 wood-frame walls (24” on center) with four inches of open-cell spray foam insulation, and an additional 1-inch of XPS foam board sheathing wrapped with an air infiltration barrier. I can guarantee you that contractors will NOT be putting in spray foam insulation in the walls, even with the 2012 IRC. All the 2012 IRC requires is R-19 fiberglass batts in the walls. A contractor is not going to spend the extra money on spray foam when it's not required and the home buyer cannot see it.

On the thermal bridging issue, I can tell you that 24" on center would NOT be allowed in my area for an exterior shear wall. As 16" on center is standard. If the home were a 2-story, the 24" on center would not be allowed, that is for sure. They will even go 8" on center at times, especially near corners and windows. Although not mentioned now in the article, the thermal bridging issue will reveal itself in the latter test results.

I'm sure the USGBC will cry foul if they see anything askew, as they are monitoring the data.

2-3 ach and then it jumps to 5-10? pretty big numbers. I didnt realize they had data where they tested out fairly tight homes and retested them many years later. I would be interested in seeing this "research".

It IS easy for anyone, even a volunteer crew, to get 1.5 ACH 50 or less on a home. All it takes is a caulk gun, decent caulk, and time, no special or expensive measures. Why they used 4 inches of OC foam is a mystery to me when windowpane caulking, or even flash with 2# CC and fill with cellulose, rock wool or fiberglass would have been more economical and done a better job of sealing. And putting 1 inch of foam on the exterior is not impressive when the foam makers themselves are saying 1.5 is what is needed in such a climate to prevent condensation issues and provide an effective deterrence against thermal bridging.

-Rosalinda

Posted By Rosalinda on 10 Jun 2012 12:45 PM
It IS easy for anyone, even a volunteer crew, to get 1.5 ACH 50 or less on a home. All it takes is a caulk gun, decent caulk, and time, no special or expensive measures. Why they used 4 inches of OC foam is a mystery to me when windowpane caulking, or even flash with 2# CC and fill with cellulose, rock wool or fiberglass would have been more economical and done a better job of sealing. And putting 1 inch of foam on the exterior is not impressive when the foam makers themselves are saying 1.5 is what is needed in such a climate to prevent condensation issues and provide an effective deterrence against thermal bridging.

-Rosalinda

They probably used spray foam because it gave them a higher R-Value per inch and it provides a better air barrier over cellulose. All insulations have their pros and cons, even with cellulose their are drawbacks. Tests conducted by the Oak Ridge National Laboratory show chemical treatments used to treat cellulose can cause metal fasteners, plumbing pipes and electrical wires to corrode if left in contact with wet, treated cellulose insulation for extended periods of time. Fiberglass batts is probably the worst insulation you can put in a wall. Cellulose is way better.

Out in Phoenix AZ they put in 1/2" of EPS on the exterior. So 1" is above average, unless code dictates otherwise.

we are talking Chicago, so yes, 1/2" EPS (especially) does not meet IECC. Rosa is stating BS recommends 1.5" of XPS to stay safe above the dew point. While I do not agree 1.5 is EASY to hit, its definately not out of the question especially with the use of spay foams for volunteer crews. I think, like most contractors, they think spay foam is the end all solution to air sealing but it fails to address many leak prone areas in frame construction. I would venture a guess they did not caulk top and bottom plates, for example.
I do not think anyone will disagree that ICF compared to sticks in this situation should perform better. The problem is, for the cost inolved in the sticks (expanding and sheet foams), it should have better air leakage number. I think it would also be interesting if they compared construction/materials costs as well.

btw- I am still waiting on that "research" that shows a tested house at 3 ach tests out at 10 ach years later.

Posted By lzerarc on 11 Jun 2012 10:34 AM
we are talking Chicago, so yes, 1/2" EPS (especially) does not meet IECC. Rosa is stating BS recommends 1.5" of XPS to stay safe above the dew point. While I do not agree 1.5 is EASY to hit, its definately not out of the question especially with the use of spay foams for volunteer crews. I think, like most contractors, they think spay foam is the end all solution to air sealing but it fails to address many leak prone areas in frame construction. I would venture a guess they did not caulk top and bottom plates, for example.
I do not think anyone will disagree that ICF compared to sticks in this situation should perform better. The problem is, for the cost inolved in the sticks (expanding and sheet foams), it should have better air leakage number. I think it would also be interesting if they compared construction/materials costs as well.

btw- I am still waiting on that "research" that shows a tested house at 3 ach tests out at 10 ach years later.

ORNL STUDY of age and air leakage

This study shows that new homes with 3 ACH50, years later were around 8 -13 ACH50. There are other studies out there that back-up this data.

A wood framed home is constantly moving and expanding, albeit slightly but still moving with age and wind loads. A wood home racks in very high winds, you can hear it. Anyone living in a wood home knows when 40MPH + winds hit, the home will start to creak and make strange noises. Every wood home I lived in does this. Wood swells & shrinks with temperature and humidity level changes. All of these things eventually take their toll on air tightness of the home.

Maybe I am not finding that in that sutdy, but it appears they test a number of homes already at that age, comparing new, middle and old, and do not infact retest the same house later as you are implying.

I agree - the study refers to "age" but evidently "date of build" is a clearer term. I would expect that house wraps, some tapes and 100% silicon caulks would hold up well.

I find it interesting that ACH@50 is studied but there seems to be little available on just what pressure is actually present in houses (preferably on a room by room basis). For example, a furnace that draws intake air from the house will be constantly creating a vacuum - increasing the actual ACH. You need to know both ACH@50 and pressure to get an accurate ACH.

Marc Rosenbaum (see: http://www.energysmiths.com/ ) used to detail interior poly as the primary air barrier on his high performance stick built designs in VT back in the 80s. IIRC some have been recently re-tested and are within the measurement error of the original air tightness tests after more than 2 decades of service.

I s'pose if you do your air-sealing with silly-putty & chewing gum you deserve whatever happens to it. Acoustic sealant will tolerate quite a lot of flex & wood-drying compared to painters-caulk. Spray polyureathane foam is pretty good compared to tape-on-XPS too (but in double-layers with lapped seams the latter does OK too.) Duct mastic on plywood/OSB seams lasts pretty well too.

The air-retardency of dense-packed fiber and spray foam insulation holds up pretty well under stick-built hygro-thermal dynamics too.

Date of build is quite distinct from re-testing a house after X-years service. Trends toward higher-density and better-installed insulation, greater incorporation of low-expansion foam in window installations have grown quite a bit over the past 3 decades.

FWIW: I never hear my antique stick-built house creak in high wind, despite plank sheathing lack of rack-bracing of any type. (Full dimension 2x4, 16" o.c. circa 1923, 9" clapboards.)

I'm just an amateur, but just what is the "...advanced stuff..." you need to hit 3ACH/50 with stick-built Baldwin2012 referrs to? Even if the average new stick built is 4-6ACH/50, that's not the average for home built under IRC2009 code (fairly widely adopted, if not fully enforced), with a code-spec of 7ACH/50. For the IRC2012 it moved to 3ACH/50 simply because that's a number that most homes built by experienced builders under IRC2009 were already meeting. The 7ACH/50 number is dead-easy to hit even as a retrofit to already-painted new construction that somehow failed due to ignorance or ineptitude. For the builders who actually made almost ANY adjustments to the design or build for better air tightness will hit under 3ACH/50. If hitting any drilled holes for plumbing & electrical with can-foam and most of the seams with acoustic sealant (or actually detailing the housewrap) is "advanced stuff" in your book, it must be a different edition than my book.