Information about design and construction of earth sheltered homes and a journal of my own progress

Radiant Basement Floor


Posted on May 30, 2015 by

Earth sheltered homes normally get very scaled down heating systems (some even skip them entirely).  My area requires a heating system, so rather than get an expensive furnace that I would hardly use, I decided to go with an inexpensive “on demand mini boiler” hot water radiant system.  I got quotes for install that were as high as $60,000, but figured I could do it for less than a 20th of that, so this went on the list of things that I would do myself.  I read a couple books and planned it out.  Then I bought the manifolds and supplies from (less than 400$ for the basement).

We got it all installed and inspected (our first mechanical inspection) and then had Dysert Concrete handle the actual pour of the floor.



Installing the radiant floor was easy, but some of the recordings didn’t work out, so the final video is shorter than usual.  You can read the story below for the details that wouldn’t fit in the narration.

The Video:

The Story:

I started with working out the layout on the computer.  Building code requires that no circuit be longer than 300 ft, and most experts recommend that you balance the lengths of the radiant tubes, so you definitely want to plan it out ahead of time.

I tried a number of different plans that ran the tubes thru the hall to the various rooms, but it was just too inefficient and cumbersome to get things “zoned” well that way.  In the end, I decided to drill some 5/8ths inch holes thru the base of the mechanical room wall to simplify the layout.  With the right tools (DeWalt hammer drill and a long 5/8ths inch bit), that was pretty easy.

We had leveled out the pea stone after the “underground inspection”, but David helped me do some final leveling of the peastone and then Zack helped get the 6 mil plastic down.  This plastic is important for keeping water vapor from the ground out of your concrete floor and is required by building code.  It also helps keep the radon out, etc.

Six MIL?

A mil is not a millimeter.  Six MIL is six thousands of an inch or roughly 0.152mm.  Before most English speaking countries switched from the imperial measurement system to metric, they would have called it a “thou”, based on the Germanic route word for “thousandth”, but for some reason, America decided to go “romantic” language based with this one and called it a “MIL” instead (based on the word for “thousandth” in languages like French or Italian).  This is a similar etymology to how the rest of the world got the word “milli” for the Metric system, hence the similarity.

HuskyWe don’t use “MIL” much in the USA, except for quantifying thin film thickness.

Since it is difficult to imagine things in thousands of an inch;

  • 5 MIL = grocery store bag
  • 5 MILS = Garbage Bag
  • 3 MILS = Husky Contractor Bag
  • 17 MILS = Pond Liner
  • 35 MILS = Credit Card


JigSaw Puzzle

David tossed us some sheets of insulation and we got started on the jigsaw puzzle.  My rooms are unusually shaped and since they didn’t actually stock those shapes at Home Depot, we cheated by cutting pieces.  We started with measuring, but usually ended up trimming each piece iteratively until it fit.  We taped all the pieces together and shoved trimmings into any gaps along the wall.  Not too hard, but certainly more time consuming than a square room might have been.  This probably wasted about 15$ worth of insulation, so not too bad.

Radiant tube

I marked the radiant tube layout directly o n the insulation based on that balanced plan I had carefully worked out on my computer.  I used piece of scrap wood marked with the right size increments and a can of upside down surveyors paint.  In addition to basic tic marks to follow, I also painted in the end loops so the whole plan would be pretty easy to follow.

Radiant_Layout_DrilledStapling the Pex tubes down was easy and fun, Sherri and I took care of most of it, but the boys were very eager to try it themselves.  I imagine it would have been quite a lot more difficult (and much less fun) without that commercial grade tool we used.  The tool cost quite a bit (~200$) but is very well built and I will use it a lot… I also plan to sell it and recoup most of the money at the end of the project anyway.



Connecting the pex to the manifold was straightforward and easy.  There are some simple little brass connector bits and you just tighten a nut to hold it all together.


Pex Stapler saved us a lot of timeI got the Manifold, Pex pipe, the Pex stapler, staples and the pressure tester from “”.  I had looked at lots of other sites (including sites that put it all together for you, such as, but this one had the best prices and the best hardware.  There are also easy to find “coupon codes”.

John (my brother-in-law) and Zack helped me finish off the third loop.

My sister Bonnie was in town and mostly helped me with the ICFs (another post/video), but she made it into this video by helping me to fill the tubes with water so they wouldn’t float in the concrete. I had been trying to pour it from the bucket into the funnel, but she had the idea to siphon it from the bucket, which was much easier and didn’t get us as wet.

Then we pressurized the system (according to building code) so we would know if anyone punctured the pipe before the concrete set.


Concrete day arrived and the guys started with putting down some six by six wire reinforcement.  This was left over from the garage floor and will help prevent cracks from growing.  It also helps protect the pipe and keep it all down under the concrete.

The concrete was pumped in from overhead (renting the pump truck cost ¼ of the job, but was well worth it in terms of making things go easier), and spread level.  They came back an hour later and hand troweled it smooth.



In all, I paid less than 1$/sft for the insulation, radiant tube, manifold and supplies, then 3$ for the concrete work plus an extra ~500$ for the pump truck and ~1100$ worth of concrete…  So, not bad.


I hope to get the “quad deck” in soon so we can put another concrete floor over this basement.


Cleaning out the slag


Posted on May 21, 2015 by

Now this was a tough long weekend.FamilyCleanup

Basically, the basement of our earth sheltered home was filled with approximately 11 cubic yards of concrete slag that needed to be broken up and removed so we could prep for pouring the basement floor.

It was something we have known we needed to do since last year, but were putting it off for obvious reasons.



Here is the video:

Why did this mess happen?

All this concrete was wasted shotcrete that wasn’t on the walls and should not have been on the floor either.

As you may recall, I had used steel studs to frame the basement and then placed metal lath on the inside to “catch” the shotcrete.  I had been told (by the shotcrete guys) that the lath would be enough to prevent much of the shotcrete (peastone) from blowing thru.  I was told to expect a thin layer of concrete on the inside, thin enough that it would break up into small fragments just by walking on it and that it would actually save me from needing to add as much pea stone later.

Watching the shotcrete being applied, it did appear that not much passed thru when it was applied at a downward angle onto the previous shotcrete.  They did do it this way for the first couple levels, and actually raised a scaffold jack platform twice as they went.  But then they got a bit tired and started shooting horizontally and even at an upward angle.  This allowed much more shotcrete to pass thru.

The effect was cumulative with blow thru coming from so many different angles, each adding its own layer of concrete.  The round central room was especially bad for this with at least 3 layers of 2 inch thick concrete across the floor.

And once the crew was working on the inner walls, there was also “rebound”, shotcrete that doesn’t stick to the wall, and “trimmings”, concrete that is cut off the wall because too much was applied in the first place.

All this concrete (that I paid for) ended up on the floor, but not in a good, “wow, you got bonus concrete floor along with your shotcrete” kind of way.  On average, I would say we had about 3 or 4 inches across most of the floor (in several layers), and up to 8 or more inches near the walls, especially in the corners.  It was uneven and lumpy and even had boot prints in it.  The whole feeling was somewhat “war torn” and more than a little depressing.

When I setup the main level, I plan to back the metal lath with fiberglass screen.  The metal lath will still provide the strength to catch the shotcrete, but the fiberglass screen will prevent any material from passing through.





Thought I would try to put some extra pics in here…

And the Story.

I like to include the text of the video, along with some extra info that doesn’t fit in a narration, so that the content is google searchable.

For this job, I had hired some teens, rented a jack hammer and taken the day off work to make the long weekend even longer.

The big question was, “How would I get this slag out of the basement?”  The final solution that I came up with was a Bagster dumpster that I got from Home Depot for 30$.


The plan was to load it up and use my trusty skid steer in to lift it up and out of the basement.

It took a bit of trial and error to figure out the best way to lift the bag and to empty it, but fortunately, we had lots of tries to get it right.  You can see how we did it in the video.

The bagster is supposed to be for only a single use, but it  held up very well, load after heavy load, for a number of days. The only tear was caused by dragging it up the rough wall in the first lift.

TeensThis first day, we were mostly focused on the edges where the thickest concrete was because I didn’t want to rent that 75 lb jack hammer for a second day.  The heavy jackhammer was actually very effective on the thick concrete, but kept getting stuck in the thinner stuff.  For that, the 11 lb breaker was much more effective.  My Dewalt hammer drill also got a work out.  At the start of the day, I couldn’t get the teens to touch the power tools, but by the end of the day, they were much more comfortable with me and the tools and were taking turns on the jack hammer.

On Saturday, my parents were in town, even though I warned them that we would be taking on the worst job of the build so far.  I also hired Zack again, he was one of the teens from the day before.

My father got to cutting a slot in the footings (doorway) for the radon tube while the rest of us got cracking on the concrete slag.  Our radon tube was made of a 4 inch corrugated drain pipe, wrapped in landscapers fabric to keep dirt out.  It just gives radon an easy way to escape so it won’t build up under the basement floor.

Then my father and I worked on the floor drains while the others just kept right on cracking up that concrete.  In order to get the slope correct from the floor drain in the central room all the way to the outer wall, we had to cut open the tops of the footings.



We had planned for holes in the footings to run these pipes, and I had even come prepared with 4″ PVC to use to form them.  However, the guys doing our footings told me they brought their own 4″ corrugated drain pipe, which they nailed in place very quickly.  The problem was that the flexible pipe “floated” up in the middle when the footings were poured. Instead of being a straight sloped hole thru the concrete, they bowed to the point that we couldn’t even get the 2″ pipes thru.  I guess they were not used to the footings being so wide.  Narrower footings probably wouldn’t have as much deformation due to “floating”.  You may recall this same issue cost me time and money during several other stages of the build. Hopefully this was the last of it.


Then we came back out again on the holiday Monday, just my wife and kids.  Sherri and I cracked things up with the 11 lb “breaker” and the boys scrambled to collect the pieces into the buckets.  When the buckets filled up, one of us would dump the bucket in the bagster.  The boys were motivated by being paid 1$ per 5 gallon bucket.  They worked for several hours before wearing out.

With the big chunks finally removed, we raked the smaller bits and then brought in some pea stone, which is required by code in my area.

I came back on another afternoon with Zack and my friend Aaron to get the second half of the pea stone down and rake it all level.  At one point in the video, you can see Aaron intentionally took a pea stone shower, just to see what it would feel like.  I don’t think he will do that again.

BuildingPermit_blurThe final product was a a peastone under-floor that meets building code.  The black pipes are to channel radon out of the home and the white pipes are plumbing or drains.  The inspector approved the work and we were able to rake the pea stone level and move on to the next step.

Next step is to get the vapor barrier, insulation and radiant floor tubes down here so we can pour the basement floor.


Precast Concrete Ribs


Posted on May 1, 2015 by

Tower_3D_2The core of my unusual earth sheltered design includes 10 precast concrete ribs.  I did try to get these done professionally and asked 4 concrete casting companies for quotes, only two got back to me and the average was about $80,000, plus shipping.  I asked what they planned to make the forms out of, and both companies said, “plywood”.

Since each of the 10 ribs only includes a few hundred dollars worth of materials and the forms are also relatively cheap to make, I decided to try it myself.

It was definitely more work than I expected, but I was able to keep the cost well below even my estimate by using a lot of the junk wood left over from forming the footings.   I think the cost for both forms was less than $300, and then I put in about $200 worth of rebar and 220$ worth of concrete into each.  That is less than one 10th the cost of having it done professionally.

The rebar cage was inspected before the pour (you can see the inspectors feet at one point in the video), and passed.

I think the first two ribs turned out well, and there were already some lessons learned, but I will make a few more before I add “Lessons Learned” to my Precast Concrete page.

In the mean time,

Here is the video;

And here is the story;

I like to include this text because it makes it “searchable”.  I also try to provide more info than I can provide talking thru the video.


If we get rid of all this peripheral stuff, at the core of this design are radial vaults around a central tower.  I need the load from the earth above the vaults to be carried to the footings.  I also wanted to keep the vaults small while allowing an open concept living space between them.  My solution was to support the vaults with these Euclidean-egg shaped concrete ribs.  I chose this shape, with its continuous curve, because of its compressive strength and drawability.  The inside end of these ribs will sit on this steel compression ring that will help distribute the load to the tower and steel posts and ultimately into the footings.  The egg is tilted so that there is lots of head room near the tower where doorways need to be placed.  On the outer edge, an integral column carries the loads directly into large 4’x4′ footing pads.

Eventually, I will span these ribs with steel arches that will form the shape of the vaults and hold up the metal lath that will catch the shotcrete, but I am getting ahead of myself…

First, I need to cast these ribs.

Rib assembly.  The brackets are screwed into the 2x# forms, which are bridged together with plywood scraps and all of that is screwed down to the melamine base... but the sidewalls are not actually screwed to the brackets...

I worked out the forming details in 3D, and it was well worth doing even though I didn’t stick to the plan exactly.  The original plan was to layout the wooden forms on a plywood base in the garage (this didn’t quite work out as planned).  I used shelf brackets and thin plywood to build the sides of the forms.  In the orginal plan, I imagined using 1/4 inch plywood, but In the end, I used multiple layers of 7mm lauan underlayment and a final layer of white board.

The loops are mostly spaced by 8 inches, but the engineer specified that the spacing be reduced to ~4" near the high end of the lower spandrel.

The specific rebar layout was designed by my engineer.  I really think it is probably “over designed” and working with that #5 rebar is a pain in the butt, but considering how important the strength of these are, I wasn’t going to short it…  In fact, based on my own calculations, I actually added additional rebar in the the one place where I felt the engineer had not gone overboard enough… I put an extra two 20ft pieces of #4 rebar along the central spine and belly of the arch.  I also integrated 3/8ths inch steel bearing plates at the high end of the arch.

The concrete volume is about 35 cubic feet, which, at about 150lbs per cubic foot, puts the weight of the arch at around 5000 lbs.  Maybe the crane operator can tell me the exact weight later.

Form Construction

For practice, I also made a quarter scale model a couple years ago, and practiced drawing out the full scale arch on my driveway.

Rib_01_ConstructionBut this year, it was time to try the real thing.  I laid out cheap particle board sheets, screwed them together, coated them with waterproofing and began to construct my curves.  This is really a scaled up version of the techniques that old-time graphic designers used to create fonts with tangential curves (before they had computers). The key to tangency is that when two arcs meet, the radial lines where the arcs meet must pass thru the center of both.  

Then it was time to cut boards to fit the profile of the arch.  This involved tracing them on the floor, then cutting them with a band saw and sanding them.  The curves from the band-saw were are not perfect, but I sanded off the main problems and the successive layers should smooth things out.

Then I realized that my particle-wood-base idea was not going to work… humidity and temperature changes during the day just caused too much warping and movement.  The waterproofing had been a waste of money and may actually have contributed to the problem because the edges were absorbing moisture and swelling differently to the rest of the boards and this led to curling. Hopefully I can reuse these waterproof boards somewhere else later.

Before removing them, I drilled thru to mark the key centers of the arcs on the floor, then I came back and drilled larger/deeper holes that could hold a nail and work as pivot points so I could quickly reform my egg-arch shapes without redoing the whole Euclidean process.

Then it was time to build up the sides of the forms…  I used shelving brackets as planned.  Each is supposed to be able to hold 250 lbs.  I spaced them less than a ft apart.  Even if they had to carry the full weight of the concrete laterally, they should have strength left over.  I had to back them with some plywood blocks (not considered in the original design) so my other layers could attach to something.

I needed to make sure that the forms could be easily taken apart.  I didn’t want any shifting at the joints, so I used a series of 4 inch offsets with the layers of thin plywood…  Where these “separable” layers overlapped, I had to be careful not to glue or screw.  Instead, I relied on notched pieces of 2×4 to keep the overlap closed tightly until the lateral load of the concrete was in place to keep it tight. Then I would remove the pegs.  (In the video, these were sticking up from the form with yellow tips)


Since I didn’t have the wood base anymore, I eventually needed to attach these form segments directly to the concrete slab with tapcon screws.

The forms for the spandrels (those cut out holes in the ribs) were a bit trickier.  I made the series of parts for them with a very carefully planned process of angled passes thru the table saw and router…  Then I nailed them all together with my air gun. I had included a fairly significant draft angle so I could get them out of the concrete at the end.  This made skinning them with the last white layer a bit more difficult and I had to trim to fit.

I had a similar problem with the tighter curve portion of the rib…  I guess my form walls were not perfectly vertical and I had some trouble with the 3rd layer and had to trim it off. After that I used smaller pieces so the angular error didn’t add up.

Eventually, I attached the remaining form pieces to the slab and caulked all the joints…  It used up three tubes of silicone.

Then my friend Dan showed up to help out…  He took on the job of creating the hangers across the form.  These will be used to hang the rebar skeleton so it is positioned in the middle of the concrete.  There was also some thought that they might help hold the form together against the pressure of the wet concrete.

Meanwhile, I got started on the Rebar…  I had ordered a pallet of pre-bent “Stirrups” and placed just the amount I would need on that pallet in the middle.  These were custom ordered and delivered for not much more than the cost of straight rebar.  Most places quoted me the price per lb, rather than per bend, so once you pick your supplier, it is pretty easy to estimate costs.


The majority of the rebar was the very thick “#5” rebar.  It is not easy to work with, and even more difficult to curve precisely.  Dan came up with this idea of clamping the pieces directly to the hangers and forming them in place…  It worked very well.


The#3 rebar stirrups need to be threaded on to the long rebar peices ahead of time for easy placement.

Tying rebar is not very fun (time-lapse did not capture many smiles, and this smiling pic above was early in the process.  I forget what Dan had just said).  You need to bend over a lot and (because I kept not wearing gloves) my hands were pretty cut up by the end of the process…  I was glad to be working in doors during all that rain though.

As we neared concrete day, another friend, Aaron, came out to help with the second rib.  We used the same process to bend the long pieces and were quite a bit more efficient with this second rib, but it still took me half a day to get all the rebar in and tied.

Pouring the Concrete

Finally, Concrete day arrived.  We poured the basement floor and ICF blocks first (another video), and then it was time for the ribs.

Sherri sprayed everything down with form release agent so the concrete wouldn’t stick to the floor or forms.  Actually, we used a bug sprayer, but it wasn’t able to spray finely enough (it should be more like a mist) so we wasted a lot of the release agent trying to get coverage. More is not better. The heavier application of release agent was running off the whiteboard instead of coating it. I plan to buy a proper sprayer for next time.

Then the concrete pipe was brought in and things got crazy for a while.

The pipe was very heavy and stiff and generally difficult to move.   The pump truck guy took pity on me and stepped in to show me how it was done…  Instead of trying to lift it, he was leaning into it. It reminded me of my windsurfing days as a kid.  I still never got the hang of moving the concrete pipe effectively, but we got the job done.  Next time, I am hopeful that the pump truck operator can park his truck in a slightly different location so that his articulating arm can better access the forms.

We used a concrete vibrator to consolidate everything and it really helped.  I had surveyed people on facebook about this and most told me not to worry about it, but these are so critical to the design, but structurally and as visual elements, that I didn’t want to skip on this part.  As soon as the vibrator entered the concrete, it liquefied (like sand in an earthquake) and flowed into all the crevasses.  Actually, when the vibrator touched the rebar cage, its vibrations were transferred and affected the concrete several feet away. I am glad I didn’t just try to hack it with a sawzall or something like that.  I don’t think it would have had the momentum to move all that form or rebar.

The concrete was poured very quickly.  I was exhausted and we took a 5 minute break for drinks, etc.

After the concrete setup for a bit, I cut the hanger wires and removed them (along with the hanger boards) and then we troweled over the surface so you don’t see any holes.  My wife, Sherri, actually liked the finishing part of the job.  We used buckets of water to clean the concrete off the hangers and tools.  And, of-course, it was all a race against the setting concrete.

Eventually the crew that had been finishing the basement came in and help with the final edging, etc.  This pic is immediately after the chaos, before the cleanup.  The wires were hanging the rebar cage.  This all cleaned up just fine.


Next Step… we need to take the forms off, and move the ribs out of the garage so we can reset the forms and pour 4 more times.


1 2 3 4 5 36 37