BuildingEnergy Systems

Rocket Oven Nitty Gritty

by Tim Barker, Koanga Institute

The first post on the rocket oven left many with more questions than they started with so this is a follow up to cover some aspects in more detail. It would probably help to re read the first article and my replies to comments as I’m just going to forge ahead with more detail on the design.

On my first design I was prepared, even expecting, to have to modify things to get it to work properly. One fundamental question I had was how small the rocket oven cross section could be and still do the job.

I consciously made the decision to start with the smallest cross section I thought would work which just happened to coincide with some square section of steel I had lying around. This was 90 mm square (3.5 inches). The plan was then to work my way up in size as needed. It’s a good idea and simplifies construction if you keep the cross section constant all the way through the rocket stove part of the design (feed tube, burn tunnel, heat riser) — this reduces turbulence and restrictions where you don’t want them.

Now the fundamental concept of a rocket stove is that you want to get complete combustion of the fuel before you even think about using the heat generated, so it’s important to insulate the feed tube, burn tunnel and heat riser. In my first stove I used perlite and or vermiculite. These are readily available from gardening shops and are used as water retaining additives, but being mineral based and light also make very good high temperature insulators. I then welded up my rocket stove elbow, or J tube, and made an enclosure out of blocks and old bits of tin to contain the perlite which I then poured in around the J tube. Please note that the dust from perlite, while not poisonous, is irritating to the lungs, so wear some sort of dust mask. On the latest design I used a 50/50 mix of fine sawdust and clay, and while not perfect, it’s a good low tech compromise. In that design I put on a thin layer and then wrapped everything in some old chicken mesh and added another couple of layers to build it up to about 50 mm thick all over, after which it helps to let the mix dry out a bit before firing up. Of course you won’t be able to restrain yourself, so just like us you will have a mini mount Vesuvius, blowing steam, hot mud, and smoke out.

Anyway back to the burn tunnel size of 90 mm. Much to my surprise this worked perfectly, straight away. Two things though — the smaller the burn tunnel cross section the harder it is to light (until of course you get the hang of it) and the second was that for cooking those things that required a lot more heat than normal (read pizza) it was a bit slow. What didn’t help either was that everyone was always opening the door to check out how it works so the poor old oven was struggling when we had pizza nights and people over. Where you are cooking for the average family these problems don’t occur, but I always said the next one would have a bigger tunnel. The current stove uses a 110 mm cross section and this is more than enough fire power for anything you could wish to cook (it does great pizza).

You will notice, in the diagram above, something called a diffuser plate. This plate evens out the heat as it comes into the oven. Without this the heat blasts up in a concentrated spot directly above the heat riser outlet, so anything you have in the oven will tend to burn on the bottom. In the institute stove the high-tech diffuser is an old cake tin sitting on the lowest rack. My old stove used a small steel plate propped up on some rocks (again, high-tech!).

As the institute oven is inside, all combustion gases leave via a flue and then exits from the bottom edge of the oven. In theory having the exit low will conserve more heat in the oven . If the exit was on top the gasses would rush straight through. Now we have some stratification of the heat with the cooler gasses being at the bottom.

In the photo below you can see our high-tech elbow, connecting the oven to the pre-existing flue from an old wood heater. This was constructed out of two home brew beer cans that just happened to be the perfect size.

While it’s not been added yet, I also have plans to install an adjustable baffle plate, as in the diagram above, so as to experiment with having more control of the gas flow rates through the oven. So far it seems that contrary to what you might think, slowing the fire down by baffling it seems to increase the temperature of the oven while reducing wood consumption. It’s my belief that this is due to a better air / fuel ratio as the oven draws in more air than it needs generally and anything over what is needed for combustion cools the fire. This is born out by the fact that today I went out to take some photos of the oven and saw that the burn tunnel was nearly completely blocked with ash to the point that there was only a 25 mm gap at the top for the hot gases to pass through. The interns hadn’t cleaned the ash out of the oven for nearly seven weeks! However, talking to them about it, they reported that the oven seemed to be getting too hot lately. So it will be interesting to see what a baffle can achieve.

The horizontal part of the J elbow is the burn tunnel and it’s good general policy to keep this as short as possible. The feed tube must be shorter than the heat riser or the rocket oven won’t draw well — generally the heat riser should be at least twice the length of the feed tube. The institute oven’s feed tube is 300 mm deep and the riser is 750 mm long, so the ratio is good . Although I haven’t tried it, for an oven connected to a flue I don’t think this ratio is as critical as the flue develops a lot of draw. This was apparent when we first tried the institute oven before the flue was connected and it drew well with the door open . When we connected it to the flue it smoked and carried on and I was starting to think "uh oh". However a quick trip onto the roof and a long piece of pipe down the flue pushed out a giant plug of grasses left there from old birds nests. When we relit the oven it took off with a soft roar, music to my ears, and drew a lot harder than previously with the door open and the flue disconnected .

A hot water cylinder to use the waste heat coming out the flue is a subject I’ll get into more detail on in my next post, as this is the basis of the hot water system I built for the PRI when I was there.


  1. I am officially in love with you. After standing on my head trying to work out how to make a rocket oven out of the last post I gave up sadly and left the process behind…but NOW its moved from the simmering back burner to the front page of my sustainable brain…”Steve…get the welder…we are making a rocket!” Thank you SO much for this wonderful heads up on rocket ovens. Should the mushroom cloud that results from our humble efforts here in Northern Tasmania be visible from the mainland… just whistle and pretend that you know nothing about it :)

  2. Amazing machinery!
    In your drawing the perlite insulation is not included in the riser. That could lead to some confusion.
    Do you think it might be possible to include a hole for cleaning out ashes? Or is it unnecessary?

  3. I`m so excited to read this update to your original post! Thanks so much for sharing. I`m curious, how do you regulate the temperature in the oven? I suppose you can keep an eye on it with a thermometer inside? Thanks again! This is great stuff!

  4. Hi guys thanks for the thanks . Herm , yes a bit lazy of me leaving out the insulation in the drawing , I kind of thought everybody in the universe knows its the insulation that makes the rocket , the burn tunnel is big enough you can get your hand in and scoop out the ashes ( make sure it’s cooled down) adding an ash draw is just more complexity and another place for a leak . Narf7 Steve doesn’t even need a welder , I’m officially in love with old red house bricks as the burn tunnel and heat riser, they are cheap, durable and to my mind heat up more rapidly and sustain a better burn due to a slower rate of heat transfer, you should be able to stack them so you can get up to a 7″ burn tunnel which will give you serious fire power and just plaster the outside in a 50/50 clay sawdust mix to seal the gaps.

    1. Hi, thanks for these infos. Don’t the red house bricks disintegrate over time in the intense heat of the burn chamber though?

  5. Hi guys me again . Herm just had a look at the drawing to check the confusion in the oven in the post I didn’t use perlite just clay sawdust mix . And to Narf7 in the drawing and post I’ve discussed and shown a baffle plate and said I was going to try it out . Short answer don’t do it , I knew it would make the fire less efficient but the decrease in velocity in the burn zone of the stove made it too smoky and yuck( technical term) . I’m considering something on the other end but regardless it works fine without air flow control.
    Cheers Tim

  6. Hello, regarding the oven it appears all the gases are directed into the oven rather than over some heat exchanger plates as seen in comercial wood fired ovens, is there any issue with smoke/ash contamination of the food? Or ingress of gases into the room when opening the door?

  7. Hi ippo
    See my previous post comments , but the short answer is yes the gasses get in around the food , no they don’t taint the food either savory or sweet, and no the gasses dont get into the room or should i say no more than if you had a comercial wood fireplace or oven, so yes a tiny bit depending on how you treat it but certainly not so your affected by it .You will get ash at the bottom of the oven that has to be periodically cleaned up but it doesn’t end up in the food as the baffle plate tends to drop it on the bottom.
    Cheers Tim

  8. Hi Steve, thanks for that extra post with more detailed info – fantastic work. I will give this a try at the next possible opportunity :)


  9. i am making a rocket oven using a ferrocement ‘barrel’ to contain the burn tube- i have made the oven dome out if castable refractory cement and perlite ( it came out well, like a skull!) but as i am making a sort of hybrid design, drawing on various ‘recipes’ for rocket ovens, i have run into trouble with the perlite insulation…i wanted to use a removable form for the burn tube but the perlite so far resists making the ‘cookie dough consistency’ required (where oh where was the website that suggested this was possible???)it just turns back into ‘cocoa pops’ which will obviously not hold the shape- i was under the impression that metal stovepipe would burn out eventually so i was gonna skip that part as i haven’t any..i have moved to the city so i no longer have access to an unlimited supply of glorious yellow clay or i would just go the clay’sawdust or other material toute as in the rechoroket site’s Haiti Rocket Ovens=. Has anyone used a recipe for
    perlite based insulation that will hold the shape of the burn tube? Maybe i just have to buy more fireclay but it was a bit expensive for not much- i was trying portland (which i know will burn out) sand, fireclay and perlite, maybe i should use lime instead of PC? Or find a supply of nice clay mud nearby? welcome any suggestions!

Leave a Reply

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

Related Articles

Back to top button