FPSF using Monolithic Slab – House with Attached Garage Separation?
I have been searching this subject for a while now, including here at GBA. The topic has been brought up, but I have yet to see answers that allow me to really understand this. The subject is NAHB’s guide to frost-protected shallow foundations with a monolithic slab. Specifically, my lack of understanding comes with this scenario, a heated space (house) – attached to an unheated space (garage). The key word there being, “attached”.
The guide requires a heated-space foundation and floor to be one entity and an unheated-space foundation and floor to be a completely separate entity. Each has its own individual design requirements. The two meet each other separated by vertical insulation between them. From a thermal loss standpoint, I get it. But, to me, these are now two separate buildings butted against each other, separated by 3″ of foam. And yet, being attached, they share common walls and roof elements creating one overall structure.
My question. Is this scenario logical long-term? In other words, is there only a miniscule possibility of differential movement between the two foundations, which could have very expensive consequences.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
You want insulation between the heated and unheated buildings so you don't get a thermal short-circuit with cold air in the garage freezing the soil under the house. If everything is built properly, there should not be differential settling.
Where do you see this requirement spelled out in the NAHB guide? I have read and used it many times but skimming through it now I don't see where it is.
Pages 25-26 of the guide show figures and guidelines for two options, one where an unheated space, such as an attached garage, is designed following the FPSF guidelines given in pages 17-19 or that shown in Figure-15. The problem with Figure-15, is that they are not showing the foundation for the entire garage structure, which would be what?
Also, if I look at Figure-14 for a small unheated area, let's assume this small unheated area is a patio with a roof extending over it that connects directly to the heated house. So, some type of support for the roof would be required at its external end on the patio. Will this isolated small area (patio) be completely stable with respect to the house so the roof never moves?
There is a slight risk of differential settlement in figure 14, in my opinion. This is how I have handled a similar situation, albeit for a timber frame with a lot of point loads so we used a beefy perimeter beam. It's similar in concept to the detail shown in figure 15. It was reviewed and stamped by a licensed engineer. Involving an engineer is always a good idea, especially when doing something outside your comfort zone or typical practice.
Attached garages tend to end up somewhere between conditioned and unconditioned, maybe semi conditioned. Most people want them colder that the house but warmer than outside (in northern climates).
I agree that if the base is all well prepared at the same way at the same time the risk of movement between the two should be low, but I see your point.
I am thinking about doing this in zone 6b. Surely these FPSF guidelines have been followed many times in even colder zones. Also, I would guess virtually every house built these days comes with an attached garage. It must work, or surely people would have talked about it by now. Likely it is not a concern.
gerrha,
I would consider pinning through the foam at a minimum for Figs 14 & 15. Or alternate foundation choices to a slab to slab construction. It is possible to get fiberglass dowels to use for the pinning for less thermal bridging.
The cold bridge problems in Fig 9 are supposedly addressed in Figs 5 & 6, but I note that no edge insulation is provided for the interior slab in Fig 6. The foundation edge shown inside the envelope could easily be a condensing point in humid environs. At least the slab has insulation under it unlike Fig. 5's. The slab in Fig 5 might be acceptable in a warmer climate zone. The soil temps under a home are a vast heat sink. My skimming of the paper raises many other issues for me.
That said, I can relate to your concerns over future movement of an unheated and separate section. While my main foundation details are for a traditional basement set directly on sandstone, I do have a large apron for the 3 car garage and four post points (not set to sandstone) for porch roofs that are continuous with the roof areas over living space .
To ensure that the apron slab didn't wander or drop from the garage door openings, we elected to overlap and pin the slab edge to the foundation door openings. We did place 1" foam between the interior garage slab and the apron to temper the losses. Not perfect, but adequate despite our winters ranging down to -10F. The garage doors face south, so the apron tends to gain enough to melt back snow and ice about a foot. Seven years in, the apron is doing fine.
Sadly, I cannot say the same for one of my porch post points. Despite being set on undisturbed native soil 60" below surface grade, I believe the soil has proven just plastic enough to have settled over 1". I may have shrinkage in the support timbers that is adding to the slight droop my beam has developed. Either way, there is a noticeable shift in the roof line. I am hopeful that jacking the post will be practical response. The load on the post is small, as the roof area is only 14 x14 and just the one corner is not set on sandstone. The post top to sandstone length would have been over ten feet, hence choosing the standard below 42" frost line footing.
Without knowing a potential site's soil types and conditions it is questionable to assume that both sections will remain stable relative to each other. In my area, expansive soils are prevalent and wildly uneven in distribution, a fact created by the millions of years of alluvial plain deposits making up our landscape. The general construction around here is a short perimeter foundation with interior pier beams for posts allowing small spans for 9.5" TJIs. Most of the time it is adequate.
I do, however, know of two houses separated by less than 800 feet that show different expansive soil presence as well as different radon levels. One house displays very localized expansion by periodically cracking the drywall above an interior pier point. I would suspect it is quite possible for a floor slab to exhibit cracks from soil lifting. The Denver and Boulder areas are well known for similar problems and soil disparities. Floating basement slabs are not unheard of.
A pair of slab foundations, as you questioned, may not behave uniformly over the combined footprint of house and garage. The scenarios depicted in Figs. 14 & 15 leave me uneasy for this reason. The addition of two levels of drainage under the slab edges in Fig. 15 is a further question mark. Getting proper compaction of materials can be tricky. It would be wise to have a proper soil engineer sign off on new fill under a footing or slab edge. All this engineering is of course expensive. I would again suggest that if you can't live with the edge losses between house and garage incurred by pins, then reconsidering the slab method is in order.
There is no need for slab-edge insulation in an unheated building. The only thing that can freeze the soil is cold air. With horizontal insulation under the entire slab, the soil under the slab can't freeze. The frost wings need to extend farther in an unheated building (or in a heated building with high levels of sub-slab insulation) because the building perimeter isn't getting the benefit of heat loss from the building interior.
For anyone interested in learning more, ASCE 32-01 is the source document for the NAHB simplified guide and for what is in the IRC: https://webstore.ansi.org/preview-pages/ASCE/preview_9780784405642.pdf.
Thanks for your reply Onslow,
My question was a bit narrow, but I too see things throughout this guide that make me look up and wonder. I am not talking about quantitative questions like why 2" and not 3" of XPS insulation, but qualitative questions.
You mentioned Figures 5 & 6 and their differences. If you pore through the text, the guide offers two methods to design a FPSF. The first "Simplified" method requires that there is no under slab insulation in the heated space. That goes with Figure 5. The second "Detailed" method to design a FPSF, accounts for situations with under slab insulation as shown in Figure 6. I am more or less OK with the heated space design.
It is the unheated space section of the guide that makes me wonder what is going on; Specifically, Figure-7, which is their unheated space design model. I have questions about Figure 7 if I imagine a very reasonable scenario in which the weather surrounding this unheated space remains below freezing for an entire week and there is constant cloud cover during this week. In that situation, we should expect no solar gain and of course the outside air is below freezing.
In my simple thinking, the only way that foundation might be expected to remain above freezing, is if two things happen. One, heat loss to the outside cold air is minimized, and two, heat gain from the hopefully warmer ground, is maximized.
The external horizontal insulation in Figure-7 would help with number one, but why no external vertical insulation that would help to minimize heat loss from the foundation to the cold air?
With number two, I must be missing something big here, but what good is that horizontal insulation under the footing extending under the entire slab? To me, that is shooting yourself in the foot if you are trying to get some heat from the ground below.
I could see a design similar to that shown in Figure 5 being better for unheated spaces than that shown in Figure 7.
To answer your question, I think I will go back to a stem wall with footing design for the entire structure including the unheated space garage and patio combined with all floor slabs isolated from the stem walls by 3" of XPS
The only thing that freezes the soil is cold air. Otherwise the ground "wants" to be warm. If you protect cold air from reaching the soil, it won't freeze and cause problems. The frost wing needs to extend far enough from the building that the R-value of the soil keeps cold air from freezing the soil under the footing. The soil under the frost wing could freeze, but if you follow the guide, it won't make it to the footing. Here in CZ6, that often requires 5' wide frost wings.
There is nothing wrong with adding vertical perimeter insulation if you want but it's not necessary unless you're trying to heat the space.
The horizontal insulation under the entire slab is what keeps cold air inside the garage from freezing the soil. In reality the garage interior may be warm enough that the soil wouldn't freeze, but the design guidelines assume the interior air will be cold--I assume as cold as the ambient outdoor conditions.
In a heated building, the design guide factors in heat loss down through the slab to help keep the soil below warmer than it would be in an unheated building.
Thank you Mr. Maines,
I kept thinking I must be looking at this wrong and your response allowed me to finally see my error. I had a fixed idea in my mind that we need to protect both the footing and the floor slab from freezing. I finally now see that it is the footing alone we are focused on to ensure that it never freezes. We will accept the temperature of the floor slab and soil directly beneath it, even if it were to freeze.
I appreciate your explanation and now it all comes together!
You're welcome, and believe me, it took me much longer to understand the principles than it did for you to figure it out.
One key is in figure 7, that the sub-slab fill is to be "non-frost-susceptible." To make up that fill I have used gravel but it's a pain to work with. EPS works but it's a waste of a precious resource. Most recently I used Glavel, and that seems to be the best of all worlds--easy to work with and it provides about R-1.7/in of insulation as well. But plain old 3/4" crushed stone works too.
I am also still in the process of understanding the nuances of FPSFs and am looking to experts here to refine my understanding if it is incorrect.
gerrha,
My understanding is that we only care if the soil underneath our structure freezes, as it’s the soil that is “frost susceptible” and can expand with freezing and contract with thawing. All FPSF strategies aim to keep the soil below footings/foundations/slabs from freezing - either by slowing heat loss from the ground, allowing intentional heat loss from heated buildings, or both. The slab, stem walls, footings, and non frost susceptible fill (gravel or Glavel) can all freeze without consequence.
Maybe I still have it wrong and welcome corrections. Thanks for starting this discussion. I am currently mulling over FPSF vs. traditional foundation in CZ 6. Thanks to Michael for sharing your experiences, especially with Glavel.
Thanks! I will check Glavel, which I have never heard of.
gerrha and entropic,
Designing successful frost protection for a foundation does not necessarily mean sensible heat management for a home. As Entropic correctly sees, the paper is focused on controlling for potential frost heave under a slab's perimeter. There seems to be a surprising indifference to interior heat exiting to ground in some diagrams. Unheated structures of course don't look much beyond the frost prevention aspect. Still, the plan for Fig.7 and the other plans shown do not address expansive soils, which are significant problems unaffected by temperature. If your site is free of the problem great, but be sure the soils are suitable for your plans.
gerrha - in #6 comment you end by saying that you will fall back on standard stem wall/footing construction and isolate your interior slabs with 3" of XPS at the edges. Even without knowing what other insulation is planned, I can only urge you NOT to do this. There will be many headaches created for door thresholds, flooring and trims. Another poster some months back was having to deal with a similar wide gap between slab and stem wall. His problem revolved around a lack of support for a door threshold and ceramic tile flooring, as well as interior trim details. Slab edge only insulation, without under slab insulation generates sizable heat loss to the ground all year.
If your stem wall is not insulated on at least one side, the losses will be significantly higher at your perimeter. Exterior temperature profiles will be mirrored to the interior side of a bare stem wall. Fig. 6 addresses this by at least having exterior insulation. Unfortunately, termites and ants can be a problem in some regions. If you are getting ready to set plans in motion, I could make a few suggestions off post that might help. I imagine there are a host of parameters uncovered here.
Michael is right about high sub slab insulation affecting wing coverage, but if you are going with footing and walls then the frost line is already accounted for and the heat losses can be managed differently.