PINEHOWL PROJECT: Simple Foundations

I am about to make my second attempt to build a Yurt (Yupee) with a Rocket Mass Heater inside.  The first version was on an area levelled about even with surrounding surface, with the RMH placed upon a few inches of sand, and the rest of the floor as pallets screwed together, cut to match the circumference, and covered in a layer of diagonal boards to hold structure. Sand was filled around the outer edge of the pallets to provide thermal insulation. This has worked about as well as intended- no sign of frost heaving in the first winter, no drafts greater than what the makeshift doors allow. The top of the pallet floor was not warm, but didn't seem any colder than might be expected.

The second attempt will be to improve on keeping and spreading heat. I would like to have the ducts from the RMH run in a loop under floor around the centre of the Yupee. The first RMH has run so strong and flawlessly I'm confident I can create a design to do that with a few modifications. The question I'd like to address here is how to provide the best thermal break around the perimeter of the foundation using Wheaton Labs standards for building materials. More specifically, if I wish to abandon or move this structure in 5-10-50 years, could I easily remove the foundation elements and use the space for a healthy garden?

The ground is made of sand. Sand, sand, some glacial outwash rocks in the sand, sand, sand, and 4-10 inches of soil on top. Of the simple structures I have built so far, all have been on levelled ground with pallets screwed together for a floor. There have been spots where the floor seems to have grown uneven by a fraction of an inch, but it looks more like rodent activity than frost heave. Everything drains so quick and thoroughly (except where the soil has been compacted by heavy vehicle traffic) that I can't imagine that frost heave would ever be a problem here, but this is far from any area of my experience apart from these few buildings.

I had been thinking of digging down a few feet, placing a layer of pumice, some sand, then cob/masonry/dirt for the fill/mass around the ducts. I would probably have to live with a dirt floor the first winter, and finish it in cob next year. But I am wondering if the existing sand provides enough thermal break and drainage that the pumice would be superfluous? Perhaps just a French drain around the perimeter to deal with the spring thaw would be enough (or needed in addition to the pumice)?

When I first mentioned this to a few PTJ type 'ringers', they were insistent that I not try to "heat the Earth". But I never made clear the sand content of my area. My current arrangement with the Bamm-Bamm style RMH has no thermal break between the mass and the Earth. I did not even clear out all of the soil when levelling the area under the Yupee, I basically stacked the sod to the side and mixed the rest of the soil in with the sand for the base, and made an empty trench around the perimeter. I did run the RMH most of the day in the winter as the Yupee only had regular wool blankets and quilts for insulation, but the mass retained enough heat to feel comfortably warm in the morning, even when I woke to the Yupee being 40* and the outdoor temp was below 0. This is similar to results I have read of the Teepee on the lab. I know for a while they had part of the duct in the Teepee running under ground and working, but I have not seen it documented why they ended up removing that.

Thoughts on how to best make this work?

Hi Coydon, sounds like an interesting project. I'm not sure why we don't hear more of floor ducted RMH so I wil be interested to hear how you get on.
My understanding is that sand makes for a poor thermal mass because the air gaps between the particles acts as insulation. If you found just a layer of sand was good enough for your previous heater, then it seems that you ought to be OK.
If I were doing it here, I would be worried about damp - water will conduct heat and turn to steam, neither of which sound desirable in a floor! But if your climate is drier, or you have a method of keeping the floor dry then that won't be an issue.
The other thought I have is that only the top of the mass will be radiating into the tipi, so you will lose some of the comfort factor from the sides of the mass. Maybe the positioning of the ducting in the cob will affect the heating and cooling of the mass, but I don't have any experience to suggest what might be better.

Just how flat and level is your building area? If there is not a place to run the perimeter ditch to, I would be more comfortable adding sand to get the mass above surrounding grade. Otherwise, it sounds like (with your natural conditions focus) you would be best off just removing all dirt so the sand can be maximally insulative. I would concentrate any insulation materials you do decide to use around the perimeter, as if the interior is continually heated above ambient there will be a much bigger gradient horizontally than vertically.

Nancy Reading wrote:Hi Coydon, sounds like an interesting project. I'm not sure why we don't hear more of floor ducted RMH so I wil be interested to hear how you get on.
My understanding is that sand makes for a poor thermal mass because the air gaps between the particles acts as insulation. If you found just a layer of sand was good enough for your previous heater, then it seems that you ought to be OK.
If I were doing it here, I would be worried about damp - water will conduct heat and turn to steam, neither of which sound desirable in a floor! But if your climate is drier, or you have a method of keeping the floor dry then that won't be an issue.
The other thought I have is that only the top of the mass will be radiating into the tipi, so you will lose some of the comfort factor from the sides of the mass. Maybe the positioning of the ducting in the cob will affect the heating and cooling of the mass, but I don't have any experience to suggest what might be better.

Hi Nancy~
I'm using the sand as an insulative element in the design. I have no way to measure how effective the current construction is, just that the ground contact through a thin layer of sand was not disastrous. Of course it is still mostly sand below that also.

My proposed design is to have dirt/cob/masonry around the ducts filling the centre of the Yurpee footprint, then some sand underneath and around the perimeter, then some insulative rocky material like pumice (or perlite or vermiculite if better). Unless the rocky perimeter would be redundant since the sand is insulative and drains well, meaning I just need to make sure the dirt/soil is cleared out in a perimeter extending beyond the edge of the yurt.

The whole in-floor mass idea has mostly been beyond the experience of the experts I've unloaded my zany proposal on. What I've been told is that heat travels equally in all directions through solids, rather than rising like it does in gases or fluids. But my hypothesis is that if the bottom and sides of the floor mass contact insulative material, it will retain most heat and radiate it out of the surface, sort of a 'wicking' process I'd speculate, where it will rise to fill the room. The main problem I noticed while staying in the teepee on the lab is that you could stand there with no shirt and have a warm torso while having pants and boots on and have cold legs and feet. I never did heat it all day to see how the bench mass radiates when fully heated. The idea here is that the barrel radiation will keep the area warm while running (I will make it as low and unobstructed as practical), and the in ground mass will provide whole area warmth after it has warmed.

Of course this design will be prone to serious cold plug problems, but my usage plan is for extended stays, definitely not a weekend warrior hang out. A priming port is an intended feature.

Coydon Wallham wrote:My proposed design is to have dirt/cob/masonry around the ducts filling the centre of the Yurpee footprint, then some sand underneath and around the perimeter, then some insulative rocky material like pumice (or perlite or vermiculite if better). Unless the rocky perimeter would be redundant since the sand is insulative and drains well, meaning I just need to make sure the dirt/soil is cleared out in a perimeter extending beyond the edge of the yurt.

It looks like the conductivity of vermiculite, pumice and perlite is about 10 times that of dry sand, vermiculite being rather more than the other 2. so you would need a much thicker layer of sand to get the same effect as with the porous materials. Even a couple of inches of perlite would make quite a bit of difference.
from engineering toolbox:
W/(mK) at 25oC
pumice : 0.02
perlite: 0.031
vermiculite: 0.065
dry sand: 0.15-0.25
dry soil (low organic matter): 0.33  
gravel 0.7

What I've been told is that heat travels equally in all directions through solids, rather than rising like it does in gases or fluids. But my hypothesis is that if the bottom and sides of the floor mass contact insulative material, it will retain most heat and radiate it out of the surface,

The insulation will make the difference, heat escapes where the path is 'easiest'. You want that to be upwards, not sideways or down.

The idea here is that the barrel radiation will keep the area warm while running (I will make it as low and unobstructed as practical), and the in ground mass will provide whole area warmth after it has warmed.

That makes sense!

Glenn Herbert wrote:Just how flat and level is your building area? If there is not a place to run the perimeter ditch to, I would be more comfortable adding sand to get the mass above surrounding grade. Otherwise, it sounds like (with your natural conditions focus) you would be best off just removing all dirt so the sand can be maximally insulative. I would concentrate any insulation materials you do decide to use around the perimeter, as if the interior is continually heated above ambient there will be a much bigger gradient horizontally than vertically.

The internet crashed when I submitted this last time, let's try again.

It is about 6" from the high side to the low side. I'm clearing a 16' circle for the 12' Yupee footprint. The site is about 20' from a drop off of about 2', so a ditch could be easily dug. However, it is next to the Yupee 1.0 and that site has seen no hint of flooding, only erosion of sand around the drip line. I have a source for more sand right by the site and will be raising the floor base at least back around original height with that fill once I've dug out all of the dirt/soil.

I had to come back and read your last sentence for a third time before I think I got the point. At first I thought you were saying a warm tent would resist allowing the heat to escape up and send it out sideways, but now I'm thinking you are referring to the bottom layer of insulation. I think you are suggesting a thick wall to the cylinder around the floor, and a thin or absent bottom? Better to take the depth of the insulative sides down below the frost line and leave it exposed to the buried thermal inertia than to have shallow walls and an insulative bottom?

Insulating under AND around the mass would be best, but if you are not going to do that, insulating the sides would give the most bang for the buck.

Coydon Wallham wrote: I think you are suggesting a thick wall to the cylinder around the floor, and a thin or absent bottom? Better to take the depth of the insulative sides down below the frost line and leave it exposed to the buried thermal inertia than to have shallow walls and an insulative bottom?

If you dig down, eventually the soil temperature is an annual average so will be slightly warmer than the colder ambient air in winter - so the top part of the soil is more likely to reflect the (cold) air temperatures. Hence there will be a greater gradient sideways than down.

I had another thought. On the toolbox site about insulating/conducting materials, it mentioned that soil containing organic matter was about as insulative as sand. (peat is as low as 0.08, compressed straw 0.09). Obviously you wouldn't want it where it would get too hot if there was a lot of organic matter, but then cob contains straw....Just a thought that may help depending on what resources you have locally.

Nancy Reading wrote:It looks like the conductivity of vermiculite, pumice and perlite is about 10 times that of dry sand, vermiculite being rather more than the other 2. so you would need a much thicker layer of sand to get the same effect as with the porous materials. Even a couple of inches of perlite would make quite a bit of difference.
from engineering toolbox:
W/(mK) at 25oC
pumice : 0.02
perlite: 0.031
vermiculite: 0.065
dry sand: 0.15-0.25
dry soil (low organic matter): 0.33  
gravel 0.7

Thank you, I had no idea what terms to use searching. That is a nice compilation of relevant information. Where did you get the data number for pumice from? It isn't in the list that I pull up under that name, and nothing that sorts into the 0.02 level matches.

I'll have to dig into that site more as I'm also interested in using wood for insulating structures. I'm confused though because the Red Pine I'd like to use is listed 3 or 4 times with different numbers ascribed to it. I'm wondering how exactly they came up with the numbers for perlite and vermiculite. Is that just one solid rock, and how would size/configuration of multiple rocks affect those numbers? I'm assuming they are only talking about "expanded" perlite and processed vermiculite...

Depending on that number, pumice looks like the preferred substance, with perlite a close second. I can find mention of them shedding water or floating, whereas I'm reading about vermiculite helping to retain water.

I'm having trouble looking for pumice. Seems I can only find it in bricks for personal care or small amounts for horticultural use. Landscaping places I'm finding carry "lava rocks", which turn out to be scoria or cinder. Since pumice floats and scoria sinks, I'm thinking pumice is going to work much better as a layer in the ground by not becoming saturated. I'll have to call some landscaping places directly to see if it would be affordable to have available.

Coydon Wallham wrote:
Thank you, I had no idea what terms to use searching. That is a nice compilation of relevant information. Where did you get the data number for pumice from? It isn't in the list that I pull up under that name, and nothing that sorts into the 0.02 level matches.

I'll have to dig into that site more as I'm also interested in using wood for insulating structures. I'm confused though because the Red Pine I'd like to use is listed 3 or 4 times with different numbers ascribed to it. I'm wondering how exactly they came up with the numbers for perlite and vermiculite. Is that just one solid rock, and how would size/configuration of multiple rocks affect those numbers? I'm assuming they are only talking about "expanded" perlite and processed vermiculite...

Depending on that number, pumice looks like the preferred substance, with perlite a close second. I can find mention of them shedding water or floating, whereas I'm reading about vermiculite helping to retain water.

I'm having trouble looking for pumice. Seems I can only find it in bricks for personal care or small amounts for horticultural use. Landscaping places I'm finding carry "lava rocks", which turn out to be scoria or cinder. Since pumice floats and scoria sinks, I'm thinking pumice is going to work much better as a layer in the ground by not becoming saturated. I'll have to call some landscaping places directly to see if it would be affordable to have available.

I'm sorrry Coydon - that table I made was a bit misleading; the pumice number came from a different site..it was about insulating floors. I'll have it on my browser history on my tablet and will post it later. You might find pumice in Oregon this site has some discussion about using it and the poster was Oregon based.

Yeah, supposing that number is at least in the ballpark (I have to wonder if the same agency tested all these materials under identical conditions), pumice looks like the standout material for this application. Unfortunately, in my area (upper midwest) it looks like it would cost 5 to 10 times more than Scoria (Lava Rock to the landscaping trade). I'm finding Perlite costs even less per volume bulk. It may be worth my while to take my pickup truck to Wheaton Labs next visit if they have affordable pumice out west, and haul a load back. For now I'm planning to go with the Scoria (3") for the bottom layer because of ease of installation and support underneath, and Perlite (6") around the perimeter where the compaction won't be as much of an issue.

Just working on the numbers in that table, it doesn't include Scoria (Cinder) either. It does have "Rock, pourous volcanic Tuff" listed at .5-2.5, which is worse than sand or even soil. I didn't find a direct comparison between Tuff and the others, but it does sound like a much different form of volcanic rock than Pumice or Scoria?

Re: your link, wowzers! One of the comments from 2013 quotes a price for pumice as $9/yard. The best I've seen looking at what would be available to me was coming in around $1500/yard, a bit of a price hike over the decade. If only I had my TARDIS working properly...

Doh, can't find it now:
matmake.com give values of 0.21 - 0.5 W/mK
a pumice quarry in Ireland give a value of 0.32 for their pumice
Ha!
I think this is where I got it from:

image source
This was from a paper on the effect of moisture content on thermal properties of building materials. I just approximated it to 0.2. This may not have been a loose fill material however, which may increase the insulative properties slightly.

Nancy Reading wrote:Doh, can't find it now:
matmake.com give values of 0.21 - 0.5 W/mK
a pumice quarry in Ireland give a value of 0.32 for their pumice
Ha!
I think this is where I got it from:

image source
This was from a paper on the effect of moisture content on thermal properties of building materials. I just approximated it to 0.2. This may not have been a loose fill material however, which may increase the insulative properties slightly.

Somebody is dropping a decimal place somewhere. That chart is showing .02, the other places are quoting .2. I see them switching between C and K, but that usually doesn't affect ratios. I would expect the higher number to come from concrete with pumice aggregate, but the matmake site seems pretty clearly to be just about the pure rock.

Maybe it isn't worth displacing sand for Pumice or Scoria? I'm thinking perlite is the clear leader now in price, availability, and insulative capability in a loose fill application like I'm considering.The main drawbacks are wasting a bunch of plastic bags, the dust during application, and it's soft, spongy consistency. I would just have to compact it for the floor space like in this thread, but that would likely take away some of the insulative advantage...

Coydon Wallham wrote:Somebody is dropping a decimal place somewhere. That chart is showing .02, the other places are quoting .2. I see them switching between C and K, but that usually doesn't affect ratios. I would expect the higher number to come from concrete with pumice aggregate, but the matmake site seems pretty clearly to be just about the pure rock...

Oh well spotted! No idea then. The quarry seemed to be blocks of the pumice too rather than concrete aggregate and the way I read the article implied that it was for a solid material. It doesn't appear pumice is worth going out of your way for though. My experience of perlite is that it is very fluid, and that is probably how to think of handling it.