The following email from Andrew explains an idea he has for a way to arrange earthbags into a fairly rigid, reinforced configuration. It seems interesting enough for me to post here and solicit comments from you dear readers. What do you think about this concept?

Having read the article you posted on Machu Picchu, I began thinking very seriously about seeking a fairly reliable method that might meet code approval by a somewhat comparably lenient code district, and possibly streamlining part of the tamping process.

This is in line with the comment I submitted about the no-crete (or non-crete? lol) bond beam methods. I wonder if you have tried the sketched design method attached? Could this be viable as a streamlining method?


I want to use straight and right angle walls to make my first roof framing as basic as possible, and would like to drop a 2×4 / 2×3 roof rafter truss (connected to top plate at side walls and the ridge beam) to allow space for R-30 or R-45.

As I think about these sketches, the slumped shapes on the end should probably be considered an exaggeration. Main idea being just walk down the rows checking for vertical plumb and whack down on the edges wherever bags need to be trued up, they’ll give and slide a bit without any barbed wire, and then lay your 2x3x8’0″ plate down and screw as shown. Then a quick tamping down either side, the bags already (hopefully) squished out to the sides a bit due to pressure from screws. Quick run down each wall face whacking in the odd lumpy spot, or checking for bags needing to be diddled. But the bags should definitely be squared off between courses to resist an unending slumping outcome.

I am primarily interested in making a transition from a 1st floor 8′ Earthbag wall to a loft and roof connection using a lot of small fasteners, without concrete. I’m considering an off-grid cabin in areas where water caching, or re-connecting a low yield well from a condemned trailer home, for example, might be the only option. For that case, if the screws cost a bit, but can be toted on the worksite in a single bag, that would make erecting a basic shell much easier in many ways. For the potential of high winds, I would like to be able to cinch the roof connection points down into something to resist lifting forces on the wide overhangs, and the nailers I’ve seen used on many earthbag projects without a bond beam look like they would only offer lateral resistance and rely on roof weight to keep things in place.

I may not have a good grasp on the weights involved, the reactions of earthbags and the forces that can occur during extreme earthquakes, but have thought vertical pinning down through the wall to the foundation (like a wooden sill plate to concrete, or rebar from concrete footing to concrete masonry blocks) is a good idea that engineers and code inspectors might want to see. I suppose a rolling, bouncing earthquake motion might jostle masonry or earthbags up and out of alignment, but considering the way the shapes at Machu Picchu are a bit Tetris-like, and have managed to shake back into alignment during earthquakes, many vertical threaded pins, combined with earthbag weight, might be enough to ensure the wall retains a vertical position, even if it’s riding a bucking bronco.

I know that wood in direct contact with cement is a big no-no, especially for anything structural, unless the wood is pressure treated. Do you think the interior of an earthbag wall would contribute more or less to rot, once the bag has completely dried out and set? PT wood drives more cost and toxins. What about painting or staining? Could the wood be coated with lime plaster to help it stay dry?

If you have any thoughts on this, please let me know. Would enjoy some feedback if the idea seems worthy, or horribly flawed in some way I missed. lol. better before building. I definitely would like to hear other ideas from the community.

Here is my response:

This is an idea I have never seen presented before. I can imagine that it would create an extremely strong, almost monolithic wall that would resist most kinds of stresses placed upon it. The downside would be the amount of wood and special screws needed to accomplish it. But then, if compared to the standard barbed wire, rebar and concrete used to secure most earthbag wall systems, the carbon footprint and amount of work involved may be similar.

As for your question about wood rotting if embedded in an earthbag wall, I am  not sure. Pressure treated wood would certainly be the safer option, but I understand your reluctance to use it. Lime plaster can hold moisture, so that might not be a good solution. Paint might provide enough protection; stain probably would still allow moisture to migrate. But once the bag fill material dries, there shouldn’t be much moisture to deal with. This concept would take some testing and experimenting to refine.


A New Concept for Securing Earthbags — 6 Comments

  1. Creative thinking, Andrew. I’m glad you’re trying for better reinforcement. I like the idea of connecting embedded blocks with long screws. I wonder how this would compare to velcro blocks as attachment points for doors, interior partitions, or shelves.
    And could this be used as a type of ring beam? Use linear wood pieces on the edges of the course, with short plastic or pressure treated blocks embedded in the middle. Then the wood could be inspected and replaced when needed?
    Yes, wood embedded within earthen walls may sadly eventually be eaten by termites in many climates. And since it cannot be checked for strength, some anti-insect treatment would be absolutely necessary.
    Horizontal embedded wood should not be used instead of barbed wire in areas with any seismic risk because the barbed wire barbs provide a unique flexing between the slippery surfaces of woven container courses. Barbed wire also delivers a fine-grained (every 5″ in height and two per course) axial reinforcement along the wall that I suspect may be stronger than embedded wood because of its grip into the dried soil masses.
    Most importantly embedded wood connected by long screws would not provide nearly as much vertical reinforcement as a common 1/2″ dia. rebar. The rebar textured surface bonds tightly with dried soil creating almost a very important trussed situation spreading force from steel into the earthen walls, similar to but of lower strength than reinforced concrete. Steel strength in wall reinforcement all depends on the diameter of the steel. Screws are so much thinner than rebar. And the multiple connection points are all vulnerable spots where the wood will be weaker than the straight steel continuous through that course would have been.
    I really appreciate your sharing these ideas, and know it takes a lot of time to mull over innovations like this. I really want to know what you think would be the best way to do a horizontal ring-beam using this interesting combination of long screws to embedded insect-resistant blocks. This would be helpful at window-sill levels or lintel levels in some buildings in higher risk areas.

    • Patti,

      Really long answer here. I’ve been having fun puzzling about this.

      Yes, I think after more discussion with Kelly that this is most likely not the best solution.

      The idea was originally to use a large number of small, “pins,” (screws) to create a grid of vertical reinforcing that could transfer all the way down to the foundation, where they could perhaps be secured to a sole plate anchored to a grade beam.

      It sounds like the weight of the earthbags and the finished density/hardness would not allow such short and relatively weak pins to provide what is needed. And beyond this, as you point out, the above idea would create a very risky failure point if rotting could occur (and it sounds like Earthbags retain enough moisture for this to be a serious potential).

      This is also not a very cost effective method. That’s a ton of 2×3’s or 2×4’s over an entire length of wall.

      I have some sketches that revolve around using a variety of rivet-like connections (thinking about connecting vertical reinforcement all the way down to the foundation) using 15″ log screws and two 4′ or an 8′ long pieces of rebar as a full-wall height rivet device that I’d be happy to send your way if you think they might be of interest.

      The above ideas would be more cost effective, and would also probably perform better.

      Years ago I discussed some ideas about Earthbags with a structural engineer (a former colleague and friend) who mentioned the importance of securing vertical reinforcement to a foundation in blockwork like concrete masonry, for example, as a complete chain of connected components. I may have been asking about Hurricane potential, and securing a roof. If I understand correctly, the windward wall of a structure acts like a fulcrum if placed in tension (wind blowing up under eaves trying to lift the roof and driving horizontally), and the force is transferred to a giant lever that is the span of the roof, which in turn puts leeward walls into compression. For many areas where I have been looking for a building site, Tornado, Hurricane or Coastal extreme MPH are real potentials, so I am hoping to come up with a connection method that would guarantee a level of resistance to uplifting (from a roof that deep down yearns to be a kite.)

      This seems a challenge with Earthbags, because grade beams are going to be hardened before the next course is laid, and making a connection from a top bond beam down to the grade beam is then restricted by the entire wall height and the solidified grade beam at the bottom.

      I sketched a few ideas to attempt a way around this, and I think they might be decent and buildable. The log screws seem really easy to install, just pound through until you hit the plate, then screw them down.

      I’ve seen some Earthbag examples (from areas where hurricane and coastal force winds are not a concern) where roof plates are just hammered into 1-2 top courses of earthbags, and for highest windspeeds I think those nails would just zipper out of the wall and go for a ride with the rest of the roof. lol.

      Log screws might achieve some resistance to both tension and lateral movement, if the weight of several courses of Earthbags would not completely overpower the screw’s bite into a 2×4 chunk of cedar or PT wood (If they bounced vertically in an Earthquake, for example). They would also look like vertical reinforcement to building officials if they’re on the fence. More importantly, they would pin Earthbags to Earthbags. I think this points out another flaw in the design pictured with the original post. If the wood did fail, 15″ log screws would still be doing something of structural value (I think), Where as above, they’d be short of achieving this.

      This topic was all in line with avoiding a concrete bond beam at top entirely. For an off-grid, uninsured dwelling, this seems like the type of cost I would want to avoid if it could be done safely. And if a slightly remote build site, not dragging concrete bags in would also cut a lot of work out.

      I am missing a good feel for the characteristics of Earthbags and this discussion has provided some better insight.

      From your experience, do you think log screws (.220″ diameter) would be likely to tear through the sides of a dried Earthbag? Is there a minimum required diameter size to resist tear-outs?

      I appreciate your description of the dried earth and rebar connection, as well as the barbed wire / polypropylene connection.

      The wealth of Natural Building info online is very helpful as I look through different design potentials, so I appreciate all comments.

  2. Hi Kelly. I´ve read the whole thing twice. Unfortunately, I am not sure I understood. (I´ve tried google Spanish translator but it was not much help) Is the idea to substitute the barbed wire between each row with a long piece of wood and then use press the bags down using long screws? I will be following this one. Saludos Kelly.

    • Yes, Cato, each course of bags would be connected to the one below with long screws that go through the top wood and screw into the wood beneath it. This would tie the whole wall into a fairly solid unit and use no barbed wire.

  3. They use a similar system in Peru. The concrete foundation is connected via rods to the roof structure, through the earth bags or compressed earth blocks. Because of termite activity, I would consider using pressure treated timber everywhere. The weight of a roof is not enough to hold it down with high winds, I have seen them fly 60 ft.

  4. One approach for the timber would be to consider scorched/yakisugi treatment. It would extend the life of the wood but can’t say by how much in that scenario.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.