Here's the dilemma. I want to build a small 500 sq foot place in Montana where there are about 6 weeks hot days that are hard to take. I'm a wimp when it comes to heat. I have a house on the same property I built 18 years ago that has a bottom story 4 feet in the ground with no insulation under the slab but the 4 ft sidewalls are insulated. Stays wonderfully cool in the summer and not hard to heat with hydronic radiators in the winter. Somewhere I read some research where the temperature of a basement slab was measured with and without insulation underneath and the difference was just a few degrees. Surprising results that made me think that the additional heat requirements for an uninsulated slab might be worth the tradeoff for a cooler house in the summer. New small house project is very low budget so I am just thinking of simple slab on grade. Walls and ceiling would be fairly well insulated. Also thinking of berming sidewalls up a couple of feet. Possibly mini-split with added wood stove for heating/cooling. Any and all opinions are gratefully received. Randy

If you do the math, in an MT climate you'll be using more heating season energy to heat the place than any air-conditioning savings you get in the summer. With slab on grade there's a wintertime comfort factor to consider too, particularly at the slab-edge. "A few degrees" difference in slab temp adds up over the course of a heating season.

For every degree below room temp it's sucking about 2 BTU/hr per square foot out of the room, so for a 500' house it's about 1000BTU/hr per degree F below the average room temp. If over the course of your ~8 month heating season the slab average 3F below room temp, that's 3000BTU/hr. Figuring about 35 weeks of heating season that's (35 x 7 x 24=) ~6000 hours, times 3000BTU/yr is ~18 MBTU of parasitic heat loss.

If heating with propane in a 95% condensing propane boiler you get about 87,000BTU out of every gallon, which adds up to about 200 gallons.

If heating with an 80% efficiency woodstove that's more than a cord.

If heating with a mini-split (probably your lowest heating cost option) running at a heating-seasonal average COP of 2.7, you get about 9200 BTU/kwh, so that 18 MBTU number is about 195 kwh. At 12 cents/kwh that's ~$25, and there's no WAY the "free" cooling factor of parasitic loss through the slab is going to save you that much during the cooling season.

The reason crawlspaces stay cool & comfortable is they are fully shaded (for very low direct solar gain), and have a conditioned, often COOLED space above them, which isn't quite the same as an above-grade windowed room, with a hot attic above. Manage summertime solar gains well by limiting or eliminating west facing windows, including the appropriate overhangs (or awnings) over south facing windows, and putting R50 or better in the attic will do far more for cooling season comfort and annual energy use than any passive-cooling you could get out of a slab.

The cost of putting 2" of Type-II EPS under the slab would run about $400, and cut that loss down to nearly-nuthin', and your bare toes would be comfortable every day of the year. If you have access to factory seconds or reclaimed roofing foam you can probably get the material cost under $150 for 3" EPS or XPS ( but don't use polyiso under slabs- it can take on water.)

Assuming a simple 20' x 25' with 9' ceilings you're looking at ~800 gross square feet of wall area, less maybe 100 square feet for windows & doors for about 700' of wall. At a code-min 2x6 w/R20 cavity insulation your wall losses are about 0.075 BTU per square foot per degree F. Assuming a design temp of -10F and an interior temp of +70F, that's 80F x U 0.075 x 700'= 4200 BTU/hr. Assuming code-max U0.35 for the windows and doors that's U 0.35 x 100' x 80F=2800 BTU/hr for the windows & doors. With an R50 attic that's about U0.02 x 80F x 500'= 800BTU/hr out the roof. Add it all up and you're at ~7800 BTU/hr, plus whatever is leaking out the slab, and whatever infiltration factors you have. If you insulate the slab the slab losses are "in the noise", but if not, it's going to be a significant fraction of the total. Either way, even with infiltration losses you're unlikely to come in over ~13-14KBTU/hr, and with better than code windows/doors/walls you'd be under 10K.

FWIW: You can probably heat & cool the place with a 3/4 ton mini-split if you insulate the slab, but would likely need a 1-ton if you don't, and the difference in equipment cost can be a large fraction of the cost of insulating. It'll be oversized for your cooling loads no matter what, but smaller to get it under 2x oversizing is going to be better for cooling efficiency, as long as it can still cover the heating load.

Also, with heat loads under 10KBTU/hr it can be hard to find woodstoves that won't turn it into a sauna even at the minimum clean-burn firing level. If going wood stove you'd probably be better off with a high-mass ceramic or soapstone stove, and fire it intermittently rather than keeping a long slow burn.

If a cool slab moves you from needing AC + ducts or heat pumps to not needing ducts or AC at all, then leaving the center of the slab uninsulated may well be the right move. Ie, sometimes it's more than just btus.

Posted By jonr on 26 Apr 2013 05:15 PM
If a cool slab moves you from needing AC + ducts or heat pumps to not needing AC at all, then leaving the center of the slab uninsulated may well be the right move. Ie, sometimes it's more than just btus.

Small house, slab on grade, in Montana? I think Dana's calc's will work out for most anywhere under that slab. If you consider frost drive is normally considered at about a 45º angle, the cooling effect will reach a long way under an on grade slab.
Personally I would not do a slab on grade in this climate, but a small house like this might be a good experiment for a frost protected foundation. Even better if op gives us a play by play for the next couple of years.
The other thing to consider is how near surface water aquifers are and how moist the soils are. With moist soils, slab on grade won't work without some heavy underslab insulation.

> With moist soils, slab on grade won't work without some heavy underslab insulation.

I believe that the code requirements for FPSFs make them work with all soil conditions, moist or not. Typically with 2" of insulation (more if you want improved energy performance). Would probably save enough to buy a nice heat pump.

FPSF came from Sweden and popular in the permafrost of Alaska. They are good/best anywhere it is cold. Basements are for roots. http://www.huduser.org/Publications/PDF/frost.pdf

Posted By BadgerBoilerMN on 26 Apr 2013 10:27 PM
FPSF came from Sweden and popular in the permafrost of Alaska. They are good/best anywhere it is cold. Basements are for roots. http://www.huduser.org/Publications/PDF/frost.pdf

Interesting document. Thanks Badger.
It pretty much says what one would expect of fpsf. Stab temps of around average 52-64ºf. Some corners experienced freezing. I trust we all agree this is not acceptable in an energy efficient home.
They claim a construction saving of 1 to 4% of the homes value. That really not much and would be much less if you need to provide for more limited heat loss from the slab. They don't say how much it cost extra to heat. Their only point is that they were able to keep the frost from penetrating the substrata under the slab in most cases. If one foot down is under 32ºf, I would think you are losing a considerable amount of heat to keep it there.
My understanding of fpsf is that you must do one of two things. Extend wing insulation far enough out from the house to allow geothermal heat to offset frost action or you must lose enough heat from the house to do the same. In most cases it is a combination of the two. If you really insulate the slab to stop a lot of heat loss, you need to go much further out to capture the geo. If anyone has a better or different opinion please voice it. Education is good!
Building on perma frost is a different cat all together. Because it is perma, I don't believe you will have any issues with frost heave as long as you don't allow any melting. Therefore you insulate to very high levels under the complete slab and don't worry about the wing because there is no geo to capture. Of course you must insulate up the side of the slab but not in plane with the slab. Whether it is a house , water line or any other utility, you must not allow enough heat to escape to cause any degradation of the perma frost. Especially on muskeg.
My apologies to OP if I have taken this thread to far around the corner!

I don't know if this would/could happen in Montana, but both of the homes we have lived in in FL had uninsulated slabs. Under certain weather conditions, the floors sweat like crazy- they actually get wet. What causes it is a cold spell followed by suddenly warm and humid weather. The cold slab sweats like a glass of iced tea.

No idea how FBBP came up with that conclusion. Right out of the document: "Experience has shown that the frost-protected shallow foundation technology may be used to increase the energy efficiency of new houses at a minimal or reduced construction cost." and "The demonstration homes performed well". There was NO freezing of the correctly designed foundations and while a FPSF is designed to protect from frost (hence the name), if you want even better energy performance, you just add more insulation (mostly near the edges, it is not needed in the wings). But then you have to do that with ANY foundation. In the better FPSF designs (there are several), building heat is not used to prevent freezing of the soil and you can reduce heat flow to any arbitrary level. The 1-4% less expensive FPS foundations also had less energy usage than the conventional foundations - at equal energy loss, the savings would be even HIGHER.

Under certain weather conditions, the floors sweat like crazy- they actually get wet. What causes it is a cold spell followed by suddenly warm and humid weather.

Insulated slabs do the same thing, although probably not quite as often.

If both un-insulated and insulated slabs sweat, then is there no answer to this problem?  I would think that an insulated slab would not sweat, or if it did, it would stop as soon as the slab warmed which should be quicker than an un-insulated slab.

Slabs never warm up quickly. Dehumidification is a solution to the problem of internal mass getting cool and then causing condensation when the weather changes. Or avoid significant internal mass and use external mass.

Well jonr it might be that I'm miss interpreting the article but it says in its Design Performance (viii) "The frost heave potential at all of the sites was successfully mitigated for both the heated and unheated areas of the homes and none of the homes experienced freezing of the subgrade soil supporting the foundation." In other words they only concern is to prevent frost heave.
In the backgound section (pages are not numbered) associated with Figure 3 it says "Energy efficiency is increased by limited heat loss through the above ground portion of the foundation wall where the most sever losses occur in conventional foundations." The claimed energy efficiency comes from comparing a concrete wall (not slab) to an FPSF. Who today would have an low insulated concrete wall?

Yes the demonstration homes performed well. There was no FROST HEAVE. They still had cold slabs with huge heat losses right??

"The coldest slab surface temperatures were recorded at 1 foot (0.3 meters) in from the slab edge at the corner, and they ranged from 37 to 48°F (3 to 9°C) for all but the Alaska site."

http://www.cs.arizona.edu/people/jcropper/desguide.html This is the recommendations from the original document. Look to - Step 2: Calculate the R-value for the Floor Slab Cross Section, Rf. This clearly indicates that if you do not allow heatloss from the slab to protect it from frost action, you must treat it as an unheated building.

Posted By jonr on 27 Apr 2013 07:36 AM

Under certain weather conditions, the floors sweat like crazy- they actually get wet. What causes it is a cold spell followed by suddenly warm and humid weather.

Insulated slabs do the same thing, although probably not quite as often.

Okay so why insulate a toilet tank. Same principle. This is what all of Dana's talk of where the dew point in a building is about.

This clearly indicates that if you do not allow heatloss from the slab to protect it from frost action, you must treat it as an unheated building.

Yep - the better FPSF designs were originally designed for unheated buildings. But no problem with using them with heated buildings for better energy performance. With thicker insulation for low energy use.

If your primary concern is energy use, then the study simply isn't applicable - it's primary focus was frost protection and the insulation thicknesses were the minimums needed to achieve that. The numbers there do not represent what a low energy FPSF would achieve.