Energy Systems

Rocket Stove Water Heater Redux

Rocket stoves are awesome, experimental, and a knowledge stream in flux. Or ours is, at any rate. Our rocket stove water heater has been doing its thing for nearly three years now, so we decided to take it apart and do a full examination of how it had fared.

So Nick and our current permaculture interns set to work completely dismantling the rocket stove water heater and examining all its components. We made new discoveries and adjustments, put it all back together, and then covered the whole thing with mud.

Diagram of our rocket stove water heater’s internal workings

The rocket stove water heater prior to redux, after 2.5 years hard labour.
Working, but not as well as it might.

For a short history on our much beloved rocket stove water heater, see the original article here and our 2.5 year assessment here. This setup has definitely done good service, but it wasn’t functioning as efficiently as it used to. Time to see what was going on.

After taking the heat riser and the water jacket off, two things became clear. The first was that the vermiculite that we used as insulation in the heat riser had settled, leaving a 10cm gap at the top of the heat riser chamber. So that was decreasing the efficiency of heat transfer.

We chose to switch from vermiculite (which would continue to settle over time) to some left over earthwool insulation we had from the tinyhouse build. As earthwool is made from spun rock, it’s a good choice as it won’t burn or shift under the temperatures employed in our rocket stove.

Settled vermiculite insulating the heat riser. 2.5 years ago, this was full.

Adam suits up to install earthwool as the new insulative layer in the head riser.

Installing earthwool into the heat riser. Non-settling insulation that won’t burn,
made from offcuts of another project.

The second thing we looked at was the water jacket heat exchanger. The internals of it were filthy. While you would expect any chimney to be filthy (in a wholesomely sooty way), there was so much creosote build-up on the walls that we could scrub it off in massive flakes.

This meant that there was a thick layer of stuff between the hot air from the fire and the metal lining in which the water waited to be heated — which is not what you want for efficient hot-air-to-hot-water transfer. So we scrubbed it out.

This creosote is coming from the eucalyptus sticks we burn in the rocket stove. It’s a natural byproduct that you’re going to have to live with, if your available wood source contains large amounts of it. So until we get our super year-round-willow-coppice-sapling-stickwood supply sorted, we’ve got lots of creosote. Fair enough.

The junction of the burn chamber and the heat riser. Quite a bit of creosote here!

Olivier tests the toot-ability of the water outlet pipe on the heat exchanger

Inside the heat exchanger. Lots more crud-like creosote
preventing good heat transfer.

We tested we still had the desired ‘suck’ within the burn chamber,
before cobbing back over it.

Once we’d cleaned everything out, we made a couple of other small improvements, one of which involved sacrificing one of my salvaged stainless steel bucket-things for a new and improved feed chamber surround.

I wasn’t entirely happy about this but I suppose all those future hot showers will be worth it.

Then it was time to cob. We had previously cobbed around the firebricks and feed barrel, but due to inadequate roofing (long since fixed) the cob eroded early in our rocket stove’s life. Now it was time to get serious. A layer of cob from top to tail, all in the name of mud.

Jurgen mixing up some mud for the cob

Nick tests the consistency of the straw-and-mud cob mixture.
If you can juggle with it, it’s ready….

Making a surround to prevent air bleed into the bottom rim of the heat riser.

Heat exchanger and heat riser re-installed, Ashley starts in on the first layer of cob.

Nearly ready to put the feed chamber back on

New feed chamber, made from one of my prized steel containers. Sigh. Oh well.

Feed chamber in place, and ready to cob

The maiden burn upon completion. And it’s working better than ever!

All finished and time to fire it up for a nice hot shower. And, our rocket stove water heater worked! Much better than it had for a long time previously! Huzzah!

Finally, I think the creosote issue is something that we’ll just have to address on a yearly or bi-yearly basis. I’m ok with that, especially as this is an outdoor, experimental system, and so we may be looking at breaking down the system every 1-2 years to clean it.

Hey, two years of zero-footprint hot showers for one day of tinkering and slapping mud on stuff? Sounds ok with me.

Rocket stove water heater all cobbed and ready
for another 2 years of active service.

Sabina showing off her muddy farm shirt

Thanks to our fabulous interns for doing a great job on this project and for taking so many silly pictures of each other smeared in soot and mud. Cheers to Adam, Olivier and Claire for the pictures.


  1. I had a combustion fireplace in my last house. Bought a chimmney scrubber for clearing out the creosote. Basically a large version of a bottle cleaning brush. It is just a “head” and a string, drop the string down the chimmney and pull the brush thru. Does a great job and saves having to dismantle the chimmney.

  2. Hey Milkwood Permaculture people!

    I also use eucalyptus (obliqua) to heat my water and house plus the addition of solar hot water panels. The old timers around here have told me to season the timber for at least 12, preferably 24 months prior to use and this will sort the creosote build up problem in the flue.

    Because it’s so densely timbered up this way, I’ve got about 5 years timber seasoning in nice neat piles. By seasoning, it means leaving the cut logs out in the rain, but off the ground using scrap metal I get at the local tip. A month before it’s burnt, the logs are split (using a solar powered electric splitter) and left to dry in a stack out of the rain. As a bonus, the local reptiles get a man made high rise apartment complex!

    The flue is clean as even after a couple of years. The creosote is a result of not letting the sap dry out of the cut logs. Also as an additional note burning wet timber wastes about 40% of the energy in that log. You can hear the timber hissing as it burns as the water turns to steam. If you can hear that, then the timber is too wet and quite possibly also green.

    I think the creosote is also a very toxic residue too. Chimney cleaners in the 19th century didn’t tend live that long.

    Top work on the rocket stove, it’s a great design and much respect for connecting up the water jacket! Maybe you could also try and incorporate a solar hot water panel in as well (as a suggestion).

    Timber is truly the most sustainable fuel that we have access too.



  3. Hi Chris, thanks for the tips on the creosote… yeah as said in an ideal situation we’d have a year’s worth of seasoned sticks, but we’re still getting there on that one! It would be quite a few sticks, i think… ah, yet another thing to aim for…

  4. hmmm… Impressive, however not impressed by your use of insulation.
    What you refer to as “Earthwool” seem to be what generally is referred to as “Mineral Wool”, right? Well despite your claim that it’s made from rock I’ve seen fire burn straight through it, as a matter of fact it only gives a fire rating of 30min in wall conventional drywall/studs wall constructions (Straw with earth plaster gets 60 min which is Max., but actually holds up for 220 minutes!).
    Now, I’m not saying you should use straw, however I am saying that your article promotes something which you’re very likely to find will have failed once you redo it in a couple of years.
    Personally I use the white kiln-grade mineral wool whenever I make ovens, be it mass or rockets. It is graded to some 1200 degree celcius. Costs a bit more, but you typically only need 1 or 2 m2. Cob on, mates!

Leave a Reply

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

Related Articles

Check Also

Back to top button