Part of my earth sheltered home design included burying a Quonset hut. Actually, this was the easiest and cheapest part of our build and if we had done the whole house this way, we might have been done the first year.
This post is for phase 3 of building our earth sheltered Quonset hut. In phase one, we put up the steel Quonset (with a little help from our friends). Phase 2 was getting up the Fox Blocks ICF endwalls. That part was interesting, but took more time (and several posts). We chose ICFs for the endwalls because they made the transition from wall to parapet for retaining the earth very straight forward. Wood end walls would probably have been cheaper and faster, but not nearly as durable..
I have seen Quonset huts buried directly (with just a pond liner for waterproofing), but I wanted to put in a layer of reinforced concrete first. Phase 3 is to cover the quonset in reinforced concrete before we can bury it in phase 4. (actually, I guess there was a waterproofing step before we can bury it also). In this way, it is really the reinforced concrete that supports the earth load. The Quonset hub becomes fancy form-work and an interior finish.
Here is the video about phase 3.
Hot: The day we put the insulation and plastic on the roof was officially the hottest day on record for our area… The reflective Quonset hut was making it feel twice as hot and the heated metal was melting the duct tape that we used to tape down the Styrofoam. I edited the water breaks out of the timelapse, but they were about every 15 minutes. Fortunately, we got clouds and cooler weather for the second half.
Crew: Just a quick note that I appreciated all the help I got on this portion of the build, both amateur and professional.
Rating: My Quonset is rated for 75 lbs per square ft. I am sure the engineers were thinking snow load and not concrete, and there is probably a significant safety factor built into that. Either way, I am sure I significantly exceeded the official rating by about 50 lbs/sqft. Fortunately, my shotcrete was also sitting on the footings, shotcrete above that was really sitting on the shotcrete below, etc. The stiffening shotcrete really formed an arch from footing to footing and is probably not loading the Quonset hut significantly. Maybe if all the concrete were wet (no internal stiffness) at the same time, we might have had a problem, but instead, the shotcrete was curing and carrying load as it was applied. Spreading the shoot over the 3 days probably helped.
Crenelations: The Quonset crenelations (groves) were about 7.5 inches deep. Across the top sections, we decided to fill with polystyrene strips 4 inches thick, 10 inches wide and 24 ft long. Some quick math, and I can tell you that we placed more than 7 cubic ft of Styrofoam in each groove, which means we reduced the load on the top of the roof by about 1000 lbs per groove. There were 20 grooves, and I filled 16 with Styrofoam, so that is a 16000 lb reduction. I left 4 of the crenelations without Styrofoam (and added extra rebar) so they would have stronger hoop beams across. We filled the concrete to a depth at least 4 inches above the crenelations, so the concrete in the crenelations was nearly 12 inches thick.
Plastic: The plastic liner was primarily to keep concrete from directly contacting the Quonset, and secondarily to provide an extra waterproofing layer. I wasn’t fanatical about it. The plastic is tough, but not impervious to puncture. I have heard of people using pond liners with a heavy felt underlayment, or even grinding down all the bolt ends that could puncture the plastic. Pond liner and underlayment can easily cost more than 50 cents per square ft (and the Quonset surface is well over 2000 sqft). Our 6 mil plastic sheet cost less than 1/10th of that, so when it got a few little holes and tears I didn’t worry about it too much. I had thought that the plastic would make it more difficult to walk around on top of the Quonset (because it could slip), but actually, it improved the walking conditions quite a lot. I sagged the plastic to allow it to properly fall into each crenelation when the rebar and concrete were added.
Openings: The Quonset hut is a nice strong shape to work with (mine was an “S-Type”, the “Q” type would have been even better, but less practical without the straight stem walls). Loads are transferred around it like pressure on an egg. As you know, any crack in an egg weakens the overall structure tremendously. I planned to have 3 openings in my Quonset shell, two skylights and a side door (to the mudroom). The most stressful time for the Quonset would be when it was covered with 46 yards of wet concrete (186,300 lbs) and two tons of rebar. You do not want to have holes in the Quonset at that time. Instead, I just created bucks to keep the concrete out of these three areas. Later, I can come back and cut the Quonset steel with a grinder to make the openings.
Rebar: Each crenelation got 4 pieces of #4 rebar. One vertical piece was “inside” each crenelation, centered and about 4 inches from the inner surface. The other verticals were at a layer about an inch past the outer surface of the Quonset hut, spaced 8 inches apart. The horizontal rebar was mostly tied to the outside of the vertical rebar and spaced no more than 12 inches apart. Some horizontal pieces were placed first, against the surface. These first horizontal pieces made it easier to place the outer layer of vertical rebar without things falling inside the crenelations. They were kept off the Quonset hut by placing a few 1.25 inch rebar chairs (according to code). We wanted most of the horizontal rebar further out where it could help hold more shotcrete up.
The other important rebar code section to worry about is overlapping the pieces. We had a 40 ft long Quonset hut with 20ft long pieces of rebar. I just placed them end to end and then came back later and tied 4ft long segments along that seam. That was more than enough to overlap both sides by 40 diameters (code).
Electrical and Plumbing: I didn’t include it in the video, but we also ran plumbing and electrical before adding the shotcrete. These were then inspected. Part part of this was running white vent tubes up the side of the Quonset so they would come up the skylight curbs.
Balance: The Quonset hut shape holds even pressure very well. It will actually get stronger when I put the balanced earth load on top… But while applying dynamic load (such as the shotcrete) it is important to apply it evenly to both sides so the Quonset hut isn’t pushed flat. Our shotcrete was poured over 3 days. The first day got most of the short vertical walls done (our Quonset hut is an “S type”). This stabilized the base and got us ready for the second day where they added most of the shotcrete, including filling some crenulations all the way across. In the video, you can see them working one side and then the other, back and forth. This required moving the heavy hose, which is tiring. They made their lives a little bit easier by toeing the hose with the lift whenever they could. Then the crew left for the weekend (not ideal) and came back on Monday to finalize the top, add the shotcrete for the skylight curbs and do a final coat for smoothing. All this time was partially because they also shot the bedroom at the same time (upcoming post), but planning to shoot things in a balanced and paced way is a good idea even if you could shoot faster.
The Compressor: The shotcrete is moved to the wall in two ways. The mix truck dumps the concrete into a hopper where it gets pumped (by a very expensive concrete pump) thru the hose to the nozzle. In the nozzle, the “nozzle man” injects compressed air to blast the concrete at the wall. Part of the magic of shotcrete (the strength, lack of cold joints, etc.) comes from the way it impacts/compacts the wall particle-by-particle. Unfortunately, the compressor broke down soon after starting on the Quonset hut and somewhat spoiled the day (but somehow didn’t reduce the cost). Ironically, this same compressor had broken down when they did the basement shotcrete, which was the last time the crew had come out. They already had a concrete truck on site with 8 yards of concrete in it, so we decided to pump it out. I knew that this wasn’t quite ideal in terms of speed of application or strength, but it did work and I didn’t have to worry about cold joints, since it was the first bit to be applied. The next day, they rented a nice new compressor for only 60$ / day and it ran flawlessly.
Hard work: Pretty much everything about shooting shotcrete is hard work. Moving the hose, aiming it all day, keeping everything flowing, finishing the surface, repairing any issues with the equipment, cleaning up, loading up, moving scaffolding, etc. It is all hard work and I appreciate the effort of the crew. Certainly I would suggest anyone think twice before deciding to take on this part of the build themselves.
These are pictures taken from my cellphone or time-lapse screenshots… Enjoy.
Here you can see that we didn’t need to add rebar chairs very often to keep the rebar off the quonset
I used smurf tube (ENT conduit) to run my electrical.
Aerial view of the Quonset, ready for shotcrete
The lift was fun. Figured we would get some rides out of it.
Aerial view of the Rebar, ready for shotcrete
Shotcrete is hard work.
Here is closeup. You can see the rebar chair keeping the horizontal bar off the quonset.
Here you can see how the shotcrete pushed the plastic (peek behind the door frame)
This is pumpcrete around the plumbing vent stack that came out of the bathroom in the quonset hut. This is the mudroom, so it will also be enclosed later. Otherwise, I would have covered this pipe in concrete to protect it.
Here is the rebar on the other side of the door. These hoops were full depth all the way across, so I put an extra piece of inner vertical rebar.
Pumpcrete near the door. Note, I didn’t cut the doorway out of the quonset until the shotcrete had set.
Here Michael is playing with wasted concrete as they clean up the pumpcrete on the wall
Note, I didn’t cut the doorway out of the quonset until the shotcrete had set.
For my design, Rebar is spaced 8 inches horizontally and 12 inches vertically.
Getting going on the Shotcrete
The lift was indispensable
Ideally, they would have two lifts so half the crew could shoot while the other half finished.
Working on the roof
This is the end of the 2nd day (the first real day) on the quonset.
Aerial view at the end of the day. Some bands of concrete were completed across the whole arch and set before they came back to fill in the rest.
For some strange reason, I have always wanted to build a block wall. So of course, I had to work at least a bit of that into the plans. I chose to build the bathroom out of block because it was all internal walls and I liked the idea that it may slightly brace the quonset hut before we added concrete over it. We also had this idea that the garage would be finished first and it would be great to have a bathroom in there so we could use it during the rest of the build…
Also, I have often joked (only half kidding) that I would like the bathroom to be waterproof so we could clean it with a power washer. And that would require no drywall…
First, the Video…
Glass Bottle Wall
I have also wanted to make a glass bottle wall for a long time, but with a low R value, they are not practical as an exterior wall in Michigan. Instead, I plan to finish the gap between this bathroom wall and the Quonset ceiling with glass bottles… This way, it will be an internal wall and will let light in from the garage skylights while still finishing off the bathroom wall to the ceiling to provide privacy. Video to come later… much later. Maybe after we move in.
1) Buy half blocks, etc. When you go to price the blocks, you will find that the full 8″x8″x16″ blocks are pretty cheap. In my case, they were 86 cents each. But then if you look at half blocks, you will find that they cost almost twice as much for half as much block. At first, I said, “No way, those are for suckers, I’ll just split the full blocks.” And while it is true that some of the full blocks come designed to be split, it was not easy and I wrecked half the blocks that I tried. Eventually, I realized that those half blocks were expensive because they were worth it! I also bought 4x8x8 blocks and cut those in half for some quarter blocks.
2) Pay for delivery. Blocks are heavy and it took hundreds to make even my small bathroom. You could make a bunch of trips with a pick up truck (be careful not to overload it), but for the money you would pay in gas, you could just have them delivered all at once and without risking wear and tear on your vehicle. Delivery also saves you time, and time is money.
3) Cut the blocks for electrical as you go. I guess we were just so excited to be building that we forgot to do that. When I came back later, it looked pretty easy in the video, but… If cutting the block had cracked it, how would I have replaced it? It was also pretty challenging to get the ENT boxes to fit in the holes with the conduit attached to the tops, and then I had to run the conduit and wires all the way up to the top of the wall and back down again, which wasted a few extra dollars. Plus, you just look silly for forgetting.
4) Make sure your concrete dye is a nice liquid or powder so you can mix it accurately. Mine was a lumpy semi-solid mass that had been sitting on the shelf at Home Depot well past its expiry date and this made it very difficult to get a uniform color. I found that I could blend the color by wet sponging the wall later, but it still doesn’t look quite right.
Here are some pics from along the way…
This was the plan for the bathroom. The shower plumbing was supposed to come up in the wall, inside the blocks, but it was placed about 6 inches off… We discussed how to adjust the bathroom to fit and ended up moving the ind wall back another 8 inches, so things ended up a little bigger than this.
Get your blocks delivered. They are heavy and delivery is cheaper than all the trips it would take to carry these safely in my pickup truck.
Sherri waiting patiently for me to setup the camera
Blocks at the end of the second evening, before the bond beam went up.
Before the bond beam
View of the door buck and scaffold inside.
Laughing about something… Sherri has water in her bottle with a pin hole in the top that she was using to help smooth out the walls.
Michael helping out on the inside while Sherri and I finish up the outside.
Half done the inside, and I am leaning back to adjust the camera.
Sherri finishing up surface bonding the last wall
A view from the inside…
I had accidentally touched up some of the outside with lighter cement from the inside (see above the door), oops. I decided to try and smooth it out.
The surface after sponging to smooth out the color and texture.
I installed the toilet flange with a hammer drill and some tapcon screws. Easy.
Brand new toilet… Oh the horrors it will see.
After the walls and door were in, we could add a toilet…
These concrete ribs were designed to give me more of an open space feeling without needing to build a wide and tall vault. For more about the design of the ribs or how I made the forms, you can read this earlier post.
This was really an epic part of the earth sheltered home build project, spanning (no pun intended) much more time than I would have liked. The costs for the forms and concrete were pretty minimal, the majority of the cost was actually in hiring the crane to move them around and set them up.
Side note: since I am so far behind… I will probably go and improve the section on the timeline next. Thanks to those of you who wrote to inquire if I died. Nope, just really busy and didn’t have time to put a video together. Speaking of which…
The Pump truck did not make things easier. Pump trucks are expensive and I needed it to come out for the floors anyway, so the first few rib pours needed to be coordinated with other jobs. This complicated the planning and made the pour days harder. It also slowed down the progress on the ribs by delaying the pours. And after all that extra planning and delay and expense, it was just plain harder to fill the ribs from the pipe than from the chute because it was really difficult to move it around. I guess my advice here would be to talk to the concrete guys and ask them how they would recommend you handle it. It was probably obvious to everyone except me.
Originally, I used silicone caulk to seal the bottom edges of the forms against the floor. However, this was a pain to clean up later. For the 2nd set, we used play-dough that my wife got cheap on clearance somewhere. The play-dough came up easily enough at the end, but it was such a pain to roll out and put into place that we ended up going back to the silicone for the later ribs.
The first set of ribs took about 4 hours per side to polish. I would start with a diamond cup wheel and then follow up with successive polishing pads at 50, 100, and 200 grit. I also used a special wheel to put a 3/4 inch round on the edges. For the second half of the ribs, I got a larger, more aggressive diamond cup wheel. It worked so well for the first step, that I quickly did all the other ribs. It was only when I got to the 50 grit pads that I noticed the diamond cup wheel had made deeper scratches than the previous one and it was much more difficult to remove them. I even went back to try the less aggressive cup wheel once I realize that the 50 grit was not working well. Overall, this mistake cost me several extra hours for each rib.
The moving dollies were supposed to be able to handle 1000lbs each, but that was clearly an exaggeration. I used more than 10 for each 5000 lb rib and still we had crunching sounds as their ball bearings exploded out all over the floor. The tires on some of them shredded completely. Eventually, I learned that most of the damage happened as each wheel rotated into the correct position to roll forward. By the time I got to the 3rd set, I had learned to point the wheels all in the right direction before lowering the rib on to them. This increased the survival rate considerably. I also salvaged partially damaged dollies by consolidating the less damaged castor wheels on to other dollies.
The rubber form liner molds were an interesting part of the build for me and I like the final look on the ribs, but again, not the best idea. More details below.
Liquid Rubber Form Liners
Originally, I planned to use the Styrofoam ceiling tiles directly, but after handling them a bit, I was worried that they were too fragile and wouldn’t last thru multiple uses. I also thought it would be a bit tricky to place them in the form so they would be centered because they were a bit narrower than ideal. However, if I used the ceiling tiles to form durable rubber form liners, I could get longer pieces that would be reusable and would be the inverse of the tiles. I could carefully center them on a board of the right width so the full depth forms could be easily placed, etc. I did some math to find the volume that I would need and found that I could get the PolyTek 75-75 ingredients for about 175$. That seemed reasonable enough to me at the time. The box of foam tiles was about $40. However, I soon discovered that mixing carefully was critical. My first few attempts were mostly good, but 95% isn’t good enough to cast concrete with. I only ended up with enough decent panels to do a small section of the first few ribs.
For the living room ribs, I decided to make the panels a little thicker (these were the ones I showed in the video). I would need to order more liquid rubber. This time, it was more like 225$, so I was in for roughly $400 worth of liquid rubber. That would have been enough to buy new Styrofoam tiles each time, so not the smartest move in hind sight.
I didn’t quite use up all my liquid rubber ingredients on the panels because I decided to try the Styrofoam ceiling tiles directly on each alternate rib. This way the pattern inverts, positive/negative for each rib.
Cost and timing
See the other page about costs for the forms, but they were just a few hundred dollars and were reused for all the ribs. So divided by 11 ribs, that is just about 30$ each.
The rebar was fairly affordable also. We used about 60$ worth of #5 and about 50$ worth of stirrups, so about 110$ per rib. There were also some steel plates that I built into the ribs and I think I paid about 10$ each for those from the scrap yard. Tie wire and welding costs are hard to guesstimate, but lets say it is less than 5$ per rib.
The concrete was about 1.3 yards per rib, which would be less than $185, even after some waste (Concrete costs about $100 per yard, delivered, but there are a couple other charges).
The molds and ceiling tiles were about 450$ total, so about 40$ each.
Then I bought about 20 of those little moving dollies and some other miscellaneous stuff for about 220$ total, so 20$/rib to move them out of the garage.
The grinder and all the pads were under 220$, but I still have the grinder many of the pads, so I am just going to leave that stuff out.
So lets say the total was $400 per rib. Not bad considering the quote to have it done by someone else was about $8,000 each.
Unfortunately, the crane and welding the ribs to the ring just about doubled that cost. The guys from RTC were great to work with and I appreciated their help, but I had not budgeted enough in that area.
Timing varied as I got better at each task, but here is the rough break down in man hours.
Form prep took about 2 hours per rib and includes cleaning the form segments, fixing any damage, re-assembling the form and attaching it to the floor, and then caulking the bottom edge.
Rebar was next and was taking about 8 hours on average, including getting all the rebar in, tying and welding.
After the rebar was in, I needed to call for an inspection, which didn’t take much time, but did delay the next step.
Next was the pour. It actually took less than half an hour to pour each pair of ribs, but then we would spend at least an hour or so troweling and finishing it off. Lets be generous and say 2 hours per rib. Then there was some delay (several days to a week) as we let the concrete cure.
Then we would spend about 6 hours (3 hours each) unpacking and moving the ribs out of the way. This included a lot of clean up.
And then the cycle could repeat.
After the ribs were out of the garage, I could polish them. The first half took about 4 hours per side to polish, so 8 hours per rib. But, thanks to an overly aggressive cup wheel and adverse weather conditions (hose freezing, etc), the second set took about twice that long. Lets put in an average of 12 hours per rib.
The flip and setup time wasn’t too bad, but lets write it down as an hour each there.
This brings the total hours per rib to about 28. I had 11 ribs to make, so about 308 hours total. If I had done those in 40 hour work weeks, it would have been nearly 8 weeks of constant work (yes, some of those hours or days were worked by other people like Sherri, Bonnie, Aaron, Dan, John and Mark). However, I already had a full time job, and also had other things to work on at the house, so I ended up spreading this part of the build over a whole year. The video on this article covers from July 2015 thru April 2016.
Needless to say, I am glad the ribs are all done.
After a long and difficult project, looking over the pics feels pretty good.
To illustrate how you get tangential arcs, important here because the load paths are smooth and in compression
3D model showing the ribs and curved steel I beams.
Model showing the kitchen, I included it here so you could see where the kitchen door slides into the rib.
Euclidean geometry can be built with nothing but a string and nails…
At the corners, it is best to overlap the rebar like this so that neither side can be pulled thru
Ryan and Carter helping with the hose
Moving those 5000lb ribs tended to wreck a portion of the wheels on each cart. Rather than toss out the whole carts, I would consolidate the good wheels together so some of the carts could be used again.
First rib lifted up… I tiled the image so the rib would look upright
At one end, the rebar is welded to a steel plate and this still plate gets welded to the steel compression ring when we set it up.
On cold days like this, I placed a heater in the middle and also covered with rigid insulation before leaving for the night.
Love this pic of Michael playing on Ribs 7 and 8 while we are finishing up prep for 9 and 10.
Work often runs me past sunset, especially in the winter.
Ribs 7 and 8, ready to pour.
We used a wet grinder to polish the concrete. It had various different cups and polishing pads.
Rebar was fun, the first time, but got old fast. In these joint areas, it got a little messy
#5 rebar along the spine with #3 rebar stirrups to tie it all together.
Some times I welded in extra bars just to keep everything oriented and spaced correctly.
Ribs, forms removed and ready to lift.
The boys playing with balancing against concrete paving stones we made
Polishing the ribs as a snow storm rolled in
Another snowy day… It was above freezing, so I still polished, I just had to get the ice out of the hose first.
Sherri and I were pretty happy after curving the last of the #5 rebar… Just a few more hours of rebar tying left on the last rib.
Sherri’s father (Mark) came out to help pour the last rib.
By this 11th rib, Sherri and I pretty much know what we are doing.
Sherri pulling down the last bits of concrete from the chute
Finishing up the last rib and knowing I wouldn’t need to reset it felt pretty good.
This bottom end of the rib was broken off when the strap slipped and it was dropped. You can see the yellow target that I pained to help guide the crane in. There was also a #5 rebar peg grouted in the middle of the square.
The crane company said I should label this pic as “Mike, the RTC oompa loompa”. All good guys to work with.
We used these bolts to attach the steel brackets to the ribs
The steel attached to the concrete
The view from below after the ribs were setup. This is our “open concept” living room…