Wood-fired earth ovens: experiments in DIY firebrick (aka “castable refractory”)

I’ve been experimenting with cheap ways to improve lo-cost wood-fired earthen ovens. How can I make mud denser, harder, and more durable? Without going to bricks and/or spending a lotta dough? Adding sand to mud reduces shrink and increases density. But clay and sand are generally still less dense (hold less heat) than a good, hi-fired dense firebrick. Hmmm…

mix exp.1

materials and lab: soapstone dust on left,
fine mason’s sand on right.

Experimental Goals:
1. to increase the density and toughness of a clay/sand thermal mix appropriate for building wood-fired ovens (and other wood-fired appliances?), 2. to fabricate a higher quality cast dome (“earth-oven”) style oven. Strategies: 1. adjust the mix of particle sizes to maximize density, and 2. amend/strengthen/improve the clay binder with sodium silicate (aka waterglass, a mineral solution which fuses at typical oven temps). Here’s three cups of packed sand, and a cup of packed dust. Three plus one equals 4, right? Wrong. Add the dust to the sand, mix it well, and you get three cups of mixed stuff.

I cup of soapstone plus three cups of sand equals...

I cup of soapstone plus three cups of sand equals…

mix exp.4

…three cups of “dusty sand”! The dust disappears into the gaps between sand particles…

mix exp.3

other cheap dust options might include crusher fines from a gravel pit/quarry, agricultural rock dust, mineral powders from a ceramic supplier, super-fine bagged sand, etc.

Each of these resulting three cups was about 15-19% heavier (and denser) than a cup of plain sand. (The dust is soapstone fines from an Oregon outfit that quarries and cuts its own stone; it’s extraordinarily fine, but other stone or brick dust — or super fine sand — will also work. Of course, if the particle size isn’t small enough, you’ll end up with greater volume — and more unfilled space. Engineers calculate surfaces and shapes and mix their aggregates accordingly.)

mix exp.6

I used my wife’s baking scales to measure weight, so take my figures with a grain of salty sourdough — or three…

OK, now I take my three cups of dense mix, and add a cup of wet clay. Three plus one equals four, right? Wrong again. Because clay particles are flat as well as fine, they will fill even smaller, tighter gaps between all those fine particles. And they’ll glue everything into a solid mass — like a dense, heavy brick. (Again, things will vary according to your clay, water content, and mixing, but in general, the increase in volume will be anywhere from negligible, to much less than the single unit of added material.)

Dust, sand, and clay would make a fine mix, but the shortcoming of a mix is just that — it’s a mix; different particles and different compounds behave differently. Over the long term, repeated heating and cooling can cause particles of a mix to separate and fall out — leaving you (or your customer) with bits of grit or sand in their pizza. Not what we want!

While it may be very noticeable in a clay-sand mix, the truth is that it can happen to the best firebrick as well, because outside surfaces get much hotter much faster than interior mass. As a result, surfaces can degrade and crumble — also called “spalling.”

To treat spalling oven surfaces, I often recommend a compound called sodium silicate, or waterglass, which binds everything together, even (especially) at high temps. Mix it with water (about half and half) and either spray or brush on. It soaks in quick and deep, and serves to “fix” a substantial layer of material.

Since sodium silicate is also used in the manufacture of hi-temp mortar, I’ve often wondered whether it would be worth adding to a regular clay-sand mix. So I made a clay:waterglass mix that was approximately 4 parts wet clay to 1 part waterglass. The clay had roughly the consistency of thick peanut butter or bean paste. After mixing, I packed the clay and dusty-sand mix into a brick form (leftovers were made into a ball).

mix tests2 (SIDE NOTE RE: WATER: the clay came out of the bottom of a bucket where it had been soaking for months. From previous experience, I figure it contained about a half gallon of water, so that means the liquid in the mix was roughly a 30% solution of waterglass. I was told by an industry rep that typical refractory mortar is made with a 40% solution, so I figure I’m in the ball park.)

RESULTS: both bricks appear to be roughly the same weight and density (and pretty close to firebrick densities!) But the really interesting result, so far, is shrinkage: the clay brick shrank a good quarter inch out of 7 inches total — almost 4%. In the photo, it’s the bottom brick, in the original mold it came out of. Above it is the waterglass brick, which had negligible shrink!

The surface hardness of both seemed relatively equal; both released grains of sand and grit when I abraded them (pretty hard!) with my thumb. Here, I think, compaction is very important, so all particles are in contact with each other, and you end up with a smooth, hard surface. This requires keeping the mix quite dry, and really whacking the material once the sand form is covered (I use a 2×4).


Around here there are a lot of gravel roads made with local crushed rock. Most of it seems basaltic: black, dense, and heavy, usually about 3/4″ bits and smaller (called “3/4 minus”); it also has plenty of fines so it compacts into a smooth, hard surface. I’ve used it for various mixes, in particular for adding thermal mass to a big oven. After I did the tests above, I added some road base to the sand-and-soapstone-dust mix, in a 3:2 ratio. The resulting brick was even heavier than standard firebrick, and significantly harder than the mix of pure sand and soapstone dust. Given that “road rock” or crushed gravel is probably easier for most folks to come by, it might be just the thing for this purpose. Here’s some pix:

L to R: standard commercial hard firebrick, homemade brick of sand & soapstone dust, homemade brick of roadbase, sand, and dust.

L to R: standard commercial hard firebrick, homemade brick of sand & soapstone dust, homemade brick of roadbase, sand, and dust.


L to R: standard commercial hard firebrick, homemade brick of sand & soapstone dust, homemade brick of roadbase, sand, and dust.

The pure sand and dust brick is a bit softer: you can see the dusty spots to the L of my thumb, where I rubbed off grains of sand.


The brick made with road base came out harder; less stuff came loose when I rubbed it w/my thumb.








The downside of waterglass is that it’s somewhat caustic, so glove your hands when you work with it.

Waterglass runs about $10-$15/gallon — a pretty cheap addition to an already cheap materials list.

Closing Story: A teacher presents his class with a large glass jar full of golf balls. “Is it full?” he asks. “Of course,” the students reply. He tips in a bag of marbles, shaking the jar as he pours so they fill the spaces between the golf balls. “OK, now it’s full, right?” “Right!” say the students. Then he pours in dry sand, again, shaking lightly to fill the (smaller) spaces. “Now it’s REALLY full, right?” “RIGHT!” yell the students. “The jar is your life,” he says. “The golf balls are the important stuff: family, kids, health, community. The pebbles are pretty important — earning money, keeping your boss happy, keeping your car maintained and your house painted. The sand is the little stuff: email, TV, the newspaper. If you fill up the jar with pebbles and sand, you won’t have room for golf balls, but if you start with the important stuff, there’s always room for the little things.” Then he grabbed his coffee pot, and poured in the whole thing, and a container of milk. The liquid didn’t even come up to the top… “That,” he said, “is just to demonstrate that even when think your life is really REALLY full, there’s always room for coffee with a friend….”

About Kiko Denzer

I live in western Oregon with my family and run Hand Print Press with help from friends Max and Eva. We are interested in restoring the arts of living to their rightful, traditional, public role, as cultural medium – and think the web is a poor substitute, but until we can fashion something better, we try to make the most of it.

8 Responses to Wood-fired earth ovens: experiments in DIY firebrick (aka “castable refractory”)

  1. Alexander says:

    Hey Kiko,

    Thanks for the post, very interesting to follow your evolution! I have some questions if you don’t mind, maybe others would be interested reading your answers too.. ;-)What do salt and sugar do to the clay mix and the role of horse manure? I’ve stumbled upon these two elsewhere. To avoid cracks in the oven, what do you think of adding some kind of fine fibre (hair, wool, fur etc.) but at the same time increasing the thickness of the thermal mass? Have you tried doing different layers in the oven, say the inner of grog mix e.g. 2 inches, then sand-clay, so that to avoid vitrification of the mass that is closest to the fire? Do you think it’s doable to layer materials of different densities in the oven? It’s just expensive to do all the thermal mass only in clay-grog mixture … Have you tried glazing the interior to prevent vitrification? In tandoor oven tradition they cook some kind of thick porridge with spinach and lots of oil and then brush the interior with it to achieve a non-stick surface for their flat breads. This prevents from grit falling and i guess protects the oven too!? In the central Asia they just apply cotton oil that burns into the surface…I have read elsewhere that it’s important to soak clay for a few days before making an oven- why is it important?

    Best regards,


  2. Steve Hecker says:

    Kiko- We’re just starting our oven next week. A couple of questions about waterglass. I assume if you’re putting it into your sand-clay mix you don’t want to be mixing it barefoot. Does it make sense to add after most of the mixing is done? Also I’m using purchased firebrick for the hearth. Would this best be treated with waterglass before putting the sand form on or after the oven is done and the sand removed? Thanks. Your book has been most helpful.

    • Kiko Denzer says:

      When I made those test bricks, the waterglass was part of the liquid for the mix. Since it’s a deflocculant, I don’t think I’d add it separately. And yes, I don’t like getting it on my skin, tho I suspect portland cement would be worse — and that’s not too bad. I don’t think it’s necessary to treat the firebrick w/waterglass. Keep me posted on your results! I’m teaching an oven workshop down here in Corvallis August 16/17. If you know anyone who might be interested, tell them to look up “cob oven workshop” on Corvallis Craigslist…

  3. Joe Battaion says:

    Hi Kiko,
    I took Bernhard’s class at Ananda last summer and picked up your oven book from him.I am finally getting around to building an oven and need to ask a question.I just put in my base insulation and to get it nice and level I have the highest bottle showing and the lowest bottle with less than an inch of well packed insulation covering it.Is this going to cause a settling problem?If so should I scrape some insulation off?By the way I am using perlite/slip.
    Thanks and take care, Joe.

    • kiko says:

      whoops, I missed this in February. Presumably, by now you’ve solved the problem and your oven is finished! But for others w/the same question, and assuming no more than about an inch of difference in bottle height, I would bring the (tamped) perlite up to the top of the lowest bottle, and then cover them all w/the next layer (dense mud), to insure good connection and bearing surfaces, so the hearth floor is well supported. More than an inch or so difference in height, I’d look for some other bottles…

  4. I love a good story.

    Thanks for doing this work.


    • Kiko Denzer says:

      Hey, Richard, thanks for the note. Good to be able to scan your site and see what you’re up to. Impressive. Congratulations! If I was closer, I’d join you for the mowing party. We need to set you up w/a Rocket mass heater…

  5. Garry Wood says:

    Thanks for this article on making a better brick. And, more importantly, thanks for the closing story. It is too easy to forget that there is always time to share coffee with a friend. I passed this on to our teacher friend and plan to use this example myself. Good Day

Leave a Reply

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

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong> <img alt="" align="" border="" height="" width="" hspace="" longdesc="" vspace="" src=""> <iframe title="" class="" type="" src="" height="" width="" allowfullscreen=""> <object src="" type="" wmode="" height="" width=""> <param name="" value="" wmode="" height="" width=""> <embed src="" type="" allowfullscreen="" allowscriptaccess="" wmode="" height="" width="">