So I made a little image to ask this question. 

In this image, the black lines represent the black plastic. 

The light brown stuff is gonna get pretty dry, right?  Since insulation is air pockets, won't the dry soil act as insulation?

I would think anything would have some insulating qualities. the shape of the soil particles would determine how much air was trapped providing insulating qualities or how dense is was to provide temperature stability. I think back filling with fluffy soil......idea 1 ....peat moss.....would provide the best insulating qualities. I have always had peat moss in the back of my head for this sort of application, it drains well, is lightweight but  I thinkwould be an excellent insulator. it is acidic and helps retard the growth of bacteria, which would be a good thing next to an underground wood structure.

Peat moss drains well  It holds water like a sponge!

Did you maybe mean Perlite?

Yes, dry soil acts as insulation, as long as you can keep it dry.  There was a site I posted a while back where the design incorporated a concrete apron around the house just for that purpose, to keep the soil dry.

Sue

you said the C word, Sue. 

Oehler talks about advantages of going underground.  One of those advantages is earthen temperature regulation.  Cooling in summer, warming in winter.  The soil insulates.

Now you are talking dry soil. The soil in between the polyethylene layers.  I would think that it does insulate.  There is probably something that insulates better, like insulation.  But you can't get insulation with a shovel. 

I would think that your soil on top of the second polyethylene layer will be doing most of the insulating, and you wouldn't have to worry about that 4 inch dry layer.  Nice stones in your top soil to absorb heat?

peatmoss may hold water once it gets in there but dry peat moss takes forever to get wet! I have filled a whole flower pot with dry peat moss and ran a hose in it. I had to plug the drain hole to have it actually float in water for a while because it shed the water so quickly.  I actually found teh process of trying to get it wet quite frustrating. those little peat pot planters that expand have to float in water for 15 minutes and that is if the water is hot. If you just hold it under water the water just skims over the top.  I have been at barns where they bedded the stalls with peat moss and they were dry dry dry. too dry and dusty to the point I thought it was unhealthy from a respiratory standpoint for the animals living there. it if it was in a protected area akin to the same plastic enclosed bundle that it comes in  adn was pure peat moss than any bit of water that happened to make it in there would just run through unless it managed to pool. but that is just my experience.

I know it is touted as increasing the water holding capacity of soil but it has to be mixed in to work. my peat moss experiment was dumped on the ground in a pile and it dried out and got "crunchy" way faster than any of the other dirt areas.

Mike?

Yes, I know it does take some effort to get it wet, but when it does get wet in an enclosed space, I think it might stay wet.  Esp if there was a problem that caused it to get wet in the first place.

Anyway, what about Perlite? If it does get wet, it drains quickly.  You can get it in different sizes (seven, I think).  I use it for making lighter-weight concrete.  Mutual Materials carries it in bags about three feet tall.  Here's some info:  http://www.perlite.net/

Vermiculite has that asbestos problem, so you probably wouldn't want to use that.

Sue

I had actually been thinking about using maybe straw or leaves in a 4/6 inch layer around the roof line and sealed against the tar paperunder the first layer.... i even wondered about putting straw BALES in next to the tar paper. And that would give a lot more insulation than soil in any climate, but in the cases of clerestories, sun scoops and such, would increase the thickness too much and make it difficult to get optimal light and ventilation. Also, dryness is absolutely CRTICAL in a situation such as that. So I don't know, I'm still pondering that idea.

What about harvesting one metric shit ton of cattail down and cottonwood down and other natural insulation type materials and using that in that 4-6 inch layer?  Of course one would have to account for compression, so you'd have to gather more.  A lot of work I know...

I like that idea. cattail down is notoriously resistant to moisture (don't know how it would do underground though) Worst case scenario, you have compost and then humus between layers. The only real issue with all this is maybe it doesnt settle uniformly and creates a place where water pockets.

Good point there.  You would have to put much more on the uphill side and as you go downhill put a little less. 

Another solution is to just add more dirt on top in places where it didn't settle uniformly. 

Cattail down does stay dry.  Even after a rain you pull that stuff off and the inner parts are bone dry.  It is abundant out here in WA, along with Cottonwood. 

Any other drawbacks I am not thinking of?

On that 4 inch layer any material that compresses will lose much of it's insulating value. On the ridge house we are now using two inches of rigid plastic insulation, EPDM, 4 inches earth, polyethlene and 14 inches earth, vegetation.

Earth is NOT a good insulator, Steve. For this reason many of the underground professionals, such as those at the former Underground Institute, or whatever it was, at the U of MN, were advocating total insulation around a U house. Nonsense! The earth at 8 foot depth stays the climate's yearly mean temperature (in Idaho mts./Canadian border 52 degrees) and will whisk away summer heat and pile in heat when the house temperature goes below 52 as during power outages, heat off at night and when owners are away. I think there has been an industry compromise and unstated agreement that insulating roofs and down to the frost line (to the depth the ground is likely to freeze during a severe winter) is OK.

P.S. Dry soil insulates MUCH better than wet.

So, Mike, are you saying that we don't need to concern ourselves with insulation?

I'm not saying that at all. I repeat, Im using it on the roof and certain walls. I'm saying don't use it all the way around the house -- lower walls, floors.

Hi all,

What an awesome forum.  I'm puzzled why it's taken me so long to find you.

I've poked around the forum some and haven't seen this mentioned anywhere, but is the PAHS (passive annual heat storage) concept relevant here? 

http://www.earthshelters.com/Ch_1.html

I've become quite passionate in the last couple of years about combining Mike's PSP with PAHS.

Excellent link!

That's the first I've seen the annualized thermal intertia mixed with something underground. 

Here is the stuff I'm familiar with:  http://greenershelter.org/index.php?pg=1

So ....  I've heard one report that dry soil has an r-value of about 0.1 ....  can anybody find some hard value for that?

I would think that a dry, sandy soil would have a higher r-value than a dry, clay soil ... 

How much do those insulation boards cost?

So if the total thickness of my roof is 36 inches, I have an overall r-value of 3.6.  Maybe less because more than half of that can be wet soil.  But maybe more, because that soil could contain a lot of air gaps.

Add to that, that the dry part is acting as a sort of thermal mass. 

But here's another weird thing, following the link that 9anda1f gave:  "It takes six months to conduct heat 20 feet (6 m) through the earth." --- so if I have three feet, do I have 1 month of buffer?  So if I heat the insides to 70 ... and let's say I've also heated a foot of dirt, does it take a month to bleed out? 

I'm looking around trying to think of what I might be able to use that is already lying around ....

Well, intensive googling tells me that  "Wood chips and other loose-fill wood products" have an r-value of 1 ....

The trouble with wood is that is can catch on fire, or get moldy ....  But it will be at least a foot of dry dirt away from me.  So maybe I don't care. 

I guess my idea is this:  I have my wood roof, and a layer of poly over that.  Then a layer of 12 inches of soil.  Then a layer of poly.  Then a layer of 4 inches of dry sawdust.  Then another layer of poly.  Then 20 inches of soil. 

BTW:  more stuff about PAHS:  http://www.norishouse.com/PAHS/UmbrellaHouse.html

So the PAHS folks have about double the house footprint in an umbrella with ...  R-8?  And they end up not having to heat their home at all.  So I wonder if I do something closer to 20% smaller and R-5 if I might have to fire up a wood stove on cold days and pay one tenth for the umbrella.  ??

So here it is.  From http://en.allexperts.com/q/Architecture-2369/home-insulation.htm

some very rough approximate r-values of materials per inch of thickness:

Dry earth: 0.33 per inch
Wet earth: 0.05 per inch
Wood: 1.25 per inch (assume the same for cardboard - not the corrugated kind but solid cardboard)

So if I have 12 inches of dry earth and 24 inches of wet earth, that would be 3.96 + 1.2:  about R-5. 

And some of my walls might be 10 feet thick - so that would be an r-value of ...  39.6 when dry? 

Hmmmm ....  I suppose if I put down a couple inches of newspaper, that might add some r-value too.

If I shift the top layer of poly up a foot, that would mean I would have two feet of dry soil and one foot of wet.  So, about R-8.5.

But the umbrella house has R-20 and it sounds like they need no heat at all.  So with R-8.5 perhaps I would need just a little bit of heat?

The umbrella house would need thousands of dollars of insulation.  Mine would not. 

In fact, with the roof, walls and floor, I think the total cost would be ....   About $400 for a three bedroom house?  All for the poly.

I talked to mike oehler about this yesterday.  I kinda hoped he would pop in here and give his opinion. 

He seems to be leaning, these days, toward using insulation a bit like what they do in the umbrella homes. 

I ran all of these theories by him and I didn't really get feedback on whether these ideas are sound or not. 

He did seem to like the idea that I would be using less materials.

I had thought of using (dry) moss in place of the four inches of dirt between layers of poly.  Actually, if I'd been able to get one of the lots in Alaska that I was hoping for, my plan was to build something like the old-timers used to build.  The house would be built like Mike's PSP for the under ground parts, with log for any exposed portions.  Waterproofing would be birch bark (one of the reasons I was looking at those particular lots was because birch is plentiful in the area), insulation would be dry moss removed from the future garden area and from the building sites (house and barn).  The whole thing would be partly dug into the ground, then bermed with a sod roof for the above-ground portions.  Just about the only purchased material that would be needed for the house itself would be glass for windows, and you could do without that if you had to.

Kathleen