Foam Under Monolithic Slab/Footing
Inspired by a recent blog post by Martin, I’m wondering if anyone has attempted to form and pour a monolithic slab/footing with foam at every facet, as depicted in the section sketch below. Pieces A and E are straightforward enough. Piece B calls for a lot of fussy smoothing at the bottom of the trench. Piece C looks like a serious pain in the neck to fit and hold in place. Piece D is a fantasy.
How much energy is saved by insulating under and to the inside of the footing in this situation? Surely one could get the same net energy savings by thickening the perimeter and underslab insulation (pieces A and E) while skipping the other, vexing pieces?
Ray
Zone 4C
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Why not just go with an insulated raft foundation... You'll be requiring structural engineering either way you look at it.
Ray,
In case you didn't see it, this article discusses other (easier) approaches: Foam Under Footings.
Martin -- Yes, it was that very article that got me thinking about incorporating under-footing insulation in the addition I'm planning. All of the methods outlined seem to require one or more of the following: extra excavation and disposal, an exceptionally thick slab, two separate concrete pours. I'm reluctant to opt for additional excavation or a second pour, as my site has access issues, and I'm concerned that a thick slab would be more costly to heat with a radiant system.
John -- Good point. The engineering might be enough in itself to curb my enthusiasm.
I'm tempted just to go with the perimeter and underslab insulation and call it good (or pretty good, as the saying goes).
If you use thicker foam right at the edge beam on the inside, you can get a descent bevel to replace piece D. I am planning on something very similar, and don't see why securing the foam to wood (some of which you'll never see again) and/or some rebar spreader-bars inside won't do the trick. Or, thicken the foam on the inside of the edge beam as well and it will be more of a structural item that you could pin to the dirt. I think the concrete company would be happy to bring a pumper truck and fill the edge beam first, then just go to town like normal. Just my thoughts on it.
Ray,
Here's a detail from Fine Homebuilding. In theory, you could modify the detail by adding a layer of horizontal foam under the thickened edge -- as long as your engineer and building inspector agree.
The problem with Ray's proposed approach, and the FHB article Martin linked to, if you've ever attempted a monolithic pour, is keeping all those pieces of foam where they're supposed to be. Not easy. Doing it in two pours is one answer, but a pump truck adds as much as $800 to the job and part of the efficiency of slab construction is the idea of a monopour. And Ray, Pretty Good doesn't mean cheating on insulation!
What I do think is a Pretty Good system is something this guy has been refining for twenty years: http://dcchomes.com/. I've been to a couple of his presentations and as designer/builder/owner he has thought things through. Check out his photo gallery and you can see his system, which is based on the idea that a variable thickness slab makes dealing with the foam easier. The idea gives engineers fits (always fun in its own right) because the slab will not be perfectly flat, as the different thicknesses of concrete will shrink differently. But he says it ends up flat enough for a house.
Michael,
I've visited the photo gallery, and I don't see a photo that clearly shows what you are describing.
Can you provide a link to the page you are talking about?
Martin, his site does not allow links, but check out "Westport Island House", starting with the 12th photo down. It's hard to see but you might think is photographic parallax is really the haunch of his slab. In his presentations Greg says that 2" foam readily conforms to a slight curve.
He owns a steamroller (or whatever the correct term is--rotary compactor?) which helps him get a nice base.
Michael,
I'm not sure why you say his site doesn't allow links. I hope I'm not breaking any rules to provide one to the page you're referring to:
http://www.dcchomes.com/westportislandphotopage.html
I'll reproduce two of the photos below. [Photo credit for both photos: Design Concepts Company].
Huh. I usually cut and paste links from the browser window, but his url didn't change for each page. I didn't think of cutting and pasting photos. I don't know Greg personally but have long thought his approach would fit in at GBA. He presents every year at the Common Ground Fair in Unity, Maine.
Martin -- The FHB detail is similar to my sketch, except for the neat right angle formed by the earth at the inside corner of the footing/slab. In practice, the earth (or more appropriately, gravel) at that corner is going to crumble off and change from 90 degrees to 135 degrees, hence "piece D" in my sketch.
Question: how would the elimination of piece D affect the energy performance of the slab as a whole? Would some disproportionate quantity of energy "rush" or be "funneled" to the one narrow uninsulated part of the slab?
Generalizing the question to any component of a building enclosure, if 50% of an area is, say, R-20 and 50% is R-10, is the net R-value of the entire area R-15, or is it smaller?
Michael -- Thanks for the link and the thoughts. What Greg Roberts is building seems like a hybrid of a raft foundation and a traditional monolithic. My purely intuitive sense is that the load bearing capacity of a raft foundation depends on the slab being thick enough and therefore stiff enough that the exterior walls don't rely exclusively on the earth directly below them; there's a sort of cantilever effect, like a waiter bearing a tray on his upright fingers. Roberts' slabs don't appear that thick, so he's beefed up the edge somewhat. But I suspect that it's not deep enough for the west coast, i.e. earthquake country. I like his creativity, though, and plan to spend some more time perusing his website.
Ray,
If 50% of an area is R-20 and the other 50% is R-10, then the R-value for the whole assembly is R-13.3.
If you're shooting for "pretty good", I would consider a stem wall and floating slab (see photo #2 from Martin's "Foam Under Footings" article). Yes, it will require two pours, but it will give you the best performance for the least fuss. You'll get better performance with a stem wall and floating slab than with a monolithic slab with pieces B, C and D eliminated....and you shouldn't have to hire a structural engineer or fight with a code official.
John S. --
Thank you for your reply. The floating slab is definitely under consideration.
How did you arrive at the overall R-value? What principles are involved? Do you have a rough and ready way of conceptualizing and responding to situations involving variations in R-value? As a concrete example, how do you approach the question of letting a cabinet into an exterior wall, thereby significantly reducing the R-value of that portion of the wall?
Ray
Ray: It appears that John S took a weighted average of the corresponding U-values for the R's given: (.5)(1/20) + (.5)(1/10) = .075 for a weighted-average U value. 1/.075 = 13.3, U being the reciprocal of R.
For what it's worth, my structural engineer doesn't want to see any gravel anywhere under the foam or the slab. Some time with a jumping jack should get the soil as flat as you need it.
The engineer's reasoning is that the gravel ensures that if there is some water below the slab, then all the soil under the slab will get wet. If there is some expansive soil, the gravel distributes the water to it, and differential expansion would then occur. Since it's frost protected, you don't need the gravel to take the water away anyway.
Good surface water management ensures that the compacted soil under the foam and slab remains dry.
Kevin,
I disagree with your engineer. A layer of crushed stone under the rigid foam allows for sub-slab depressurization and radon remediation, among other benefits.
I can see the logic in being concerned about water getting under the crushed stone, if the soil below is graded concavely. As long as the soil is perfectly level or, even better, crowned slightly and well-compacted, I doubt there would be a problem.
"The engineer's reasoning is that the gravel ensures that if there is some water below the slab, then all the soil under the slab will get wet." So, if there is water below the slab and no gravel, how is the soil that is there going to repel this water so that it does not get wet? It seems there is a fault in the logic here. Doesn't water under the slab go hand in hand w/ "wet soil"?
Lucas Durand deals with some of these issues in his blog
http://ourhouseuponmoosehill.blogspot.ca/2011/12/footings.html#more
Lucas sloped the excavation slightly before installing crushed fill which was compacted in layers. He installed a modified monolithic slab by laying down rigid foam before pouring his footings. Then he poured the slab over a layer of foam. Rebar ties the slab and the footings together. He has some excellent diagrams and photos of the process.
My structural engineer has extensive experience with bentonite soils in the Front Range of Colorado. From the post mortem of these heaved slabs he's concluded that the gravel under the slabs added to the problem by allowing the water to flow across the entire area. Compacted soil allows much less water flow.
Remember we are talking about a slab on grade, not a basement slab, or a walkout. That means zero water pressure. The water table is 20 feet down, roof water is directed at least 5 feet away, and the lot is sloped and swaled as required. Even if you were to lay the poly and foam on damp soil after a soaking rain, the soil under the slab will dry out quickly and stay dry.
Martin,
The taped poly under the slab should keep radon from entering the house through the slab.