3D Analysis of Integrated R-Value of Wall
Three years ago, there was a discussion between Martin Holliday and a subscriber named Ethan about the integrated R value of Nexcem/Durisol ICF walls with poured concrete inside. There was a lack of complete data or at least data that was easily available. I’m wondering, after three years,if Green Building Advisor is aware of more documentation publicly available that includes a three dimensional analysis of this particular product as urged by Martin Holliday. Has Oak Ridge National Laboratory, e.g. ever done an analysis since?
Martin Holliday in his discussion quoted the following from the Nexcem literature. “The Nexcem material has an R-value of 1.75 per inch. This means that all Wall Forms regardless of thickness have a basic R-value of R-8.2.” The math of this statement implies a total wall thickness of 4.7 inches. But there is not a total thickness of 4.7 inches of Nexcem material from one outer surface of the block (the inside of the wall) to the other outer surface 0f the block (the outside of the wall) all along the block from one end to the other end. I am assuming that the concrete poured into the air gap in the middle of the form does not contribute to the total R value of the filled block at all, and I am also ignoring for the moment the value of the insulation that is inserted into the inside of the form before the concrete is poured. So at least the two dimensional question that was asked but unavailable three years ago was: what would be the integrated or average R value across the entirety of a row of blocks that are filled with concrete, ignoring the additional insulation. The followup question is what is the integrated or average R value be across the entirety of a row that is both filled with concrete and which also has the insulation inserts. The next question would add the third dimension – integrating the R value from the bottom of the wall to the top of the wall to get an integrated or average R value for the entire wall. The answer for the entire wall would have to be something less than the maximum R value that the wall could have at any given point, just like the integrated value for a whole wall that has a conventional window in it is less than it is for any single point in the wall where there is no window.
This is the kind of information that the public really needs in order to fairly compare products. I hope that some analytical organization like Oak Ridge National Laboratory or a resource center like Green Building Advisor can follow up with an answer. I’d like to use the Nexcem product but I want to make an informed decision.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
To make a numerical solution for the 3D case I would run a time simulation on a simple 3D matrix of temperatures, using a configuration matrix of thermal resistances.
I tried WUFI, but didn't manage to crack the interface.
I think you could get a fairly good estimate here, considering that the "insulation" concrete has moderate thermal resistance, and the block itself has higher thermal resistance, by simply using the field estimate in a 1-dimensional analysis. That's a peculiar property of this particular system, and I think that 1D will be closer to 3D+time with this than with any other wall assembly.
Their 8" block is filled with 5" of concrete grout. If the block itself is R-1.75/in, that means (1.5" + 1.5") * R1.75 for the block plus 5" * R0.07/in for 150pcf concrete grout, for a total of R-5.6.
Since I left my initial message, I realized that, if I have all the cross sectional dimensions of the block itself from one end to the other and from top to bottom, and the per inch R values of each of the materials involved (the wood/cement slurry of the block itself, the poured cement, and the rock wool insulation inserts), I could use a spreadsheet to calculate a weighted average of all the combinations of the three materials across an entire block to estimate the net integrated R value of a block. This would be a crude estimate of what a true 3 dimensional nonlinear time simulation computer model would produce, but I would think it would still give a good idea of the real R value of an integrated wall (absent window, electrical, and plumbing inserts, of course). Unless someone comes up with either the sophisticated or a crude analysis that has already been done, I would try the spreadsheet weighted average approach myself if I can find a good cross sectional view of a block with proper dimensions.
David, that's the basic process to calculate whole-wall R-values (or U-factors) for any assembly. Just add up the weighted R-values. It's not perfect but gets pretty close.
If you do a couple of the online tutorials, Therm is not too hard to figure out. It won't do 3d but does a great job of 2d heat flow.
https://windows.lbl.gov/software/therm
Their specs do seem aspirational as it does not include the thermal bridging from the wood/cement walls or the internal concrete beams.
It does look like a nice product without a lot of the issues of regular ICF.
I just did my simple spreadsheet weighted average calculation of an R28 Nexcem block. According to Nexcem specs, this block is 14" thick, has 5" rock wool inserts, and 5.5 inches of concrete thickness in the middle of the block. I got the exact dimensions from the Nexcem Wall Form Technical and Installation Guide.
Although they refer to the block as an R28 block, I calculate a max R value of only 24.5 at any given point given these dimensions and the assumption of an R value of 1.75 per inch. An R value of 1.75 per inch is what Martin Holliday quoted from Nexcem's own literature. The maximum R value of 24.5 would occur at that point in the cross section of the block (the ends) where you have a full 14 " thickness of nothing but the block form R 1.75 material from inside to outside (i.e. where there is no concrete or rock wool in the intermediate space between the inside and outside of the wall). At other points of the cross section along the length of the block where there is rock wool and cement inside, I calculate the R value from inside to outside to be less than the maximum of 24.5. Taking the 3 dimensional weighted average R of a single whole block, I calculate a value of 22.8. This is pretty good, i.e. the weighed average is close to the max value of 24.5.
I can't explain why I calculate a max R of only 24.5 for a block that Nexcem labels nominally as a R28 block unless the assumption of an R value of 1.75 per inch is inaccurate. But that is a number that supposedly came from Nexcem's own literature. The R value for that section of the block where there is a 5" insert of rock wool and 5.5" of concrete is only 22.5 - less than the maximum of 24.5 for the section where you have only 14 full inches of Necem block material. Could they be assuming some additional R value from the materials that will cover the exterior of the Nexcem block on the inside and outside??? Perhaps this explained somewhere in their literature that I haven't read yet.
I am having a conversation directly with Nexcem about all of this, but wanted to share the fact that one can do one's own calculations with a simple spreadsheet to get a handle on the matter. I recognized the ability to self educate after I had already submitted questions to Nexcem yesterday. I will update if I get more definitive info directly from Nexcem.
+1 on using THERM over a weighted average. THERM is easier to use if you import the drawing from some other program.
I think the r28 comes from doing a complete assembly R value calculation including drywall on the inside, siding on the outside and associated air films.
So if you take an R25 material, with drywall/cdx+siding/air flims it works out to R28.
Can the NexCem block accept a peel and stick membrane or would it melt the block face and require a non-VOC type of glue, similar to what's used on EPS?
Does a Durisol/NexCem block need a WRB?
I passed your questions on to Nexcem. Their reply is as follows:
"Yes – Peel n Stick membranes are used with Nexcem all the time. Our material will not “melt”. Any membrane used on cement/concrete, etc., can be used with Nexcem.
And Yes – you need a WRB barrier."