!!!!!!!!!!!!!! Lessons from a Rocket Stove Disaster

I'm generally a big advocate of DIY ideas and implementation, but a fire last year has reminded me of the benefits of conventional codes and safety protocols, and I feel compelled to speak a bit of lessons learned from disaster.



I urge all fire builders to consider:
1. The flash point of combustibles changes over time. The walnut on the enclosure wasn't going to smolder when it was a living tree. Nor when it was first installed. Nor during a foggy day. Nor when the stove is first lit. But after a thousand burns, it had been dried out over years. Couple that with a super long fire, it was dry, and it was hot, and it was ready.
2. Relatedly, it's not enough for fire appliances to perform safely for 1,000 burns. Stoves need to perform safely in perpetuity. As someone put it, this fire wasn't so much as an accident as an eventuality.
3. The nature of fire is that when there is failure, it is immediate. When you build a cut rate frame, the building might lean over time. A cut rate roof will leak, and eventually this leak will rot the rafters, and eventually cave in. Even a cut rate foundation will crack slowly and usually not be catastrophic overnight. But build a cut rate fire appliance, and when that appliance fails, people can die within minutes. I just wrote an IG post on the economics of building; you can find my handle at the bottom of this post.
4. You can't know who will use your stove, or how they'll use it. Talk to enough stove builders, and they all have a story about a stove they built that worked well... until the client had a party and fired it nonstop for three days straight. Or who burned green wood. Uncle Mud calls these users "Gilligans"-- well meaning folks who have less than ideal experience. But ultimately it's our responsibility as builders to construct appliances that factor in this misuse, and which should be able to withstand extreme usage outside how we might use it. This extends to even stoves we build for ourselves, because we don't know who might use them after us. Another way of putting this is -- a design that relies on specific use to burn safely is a bad design.
5. I've come to understand the function of insulation differently from this experience. I had thought that by heavily insulating between the fire and the combustible, I would shield the combustible adequately. There's of course some truth to this. But all materials (no matter how light and fluffy) have mass. In extended firings, with enough energy, this heat can build up in insulation. Stoves need to allow for heat dissipation between them and combustibles. Usually, this is an airspace. There are, of course, exceptions, but generally, all fire appliances need an airspace to dissipate heat before this heat can affect combustibles. For example, keeping a 2" airspace inside a ceiling box around a Class A chimney is safer than filling that space with perlite.



The fire occurred in the Critter Kitchen, a three season outdoor kitchen at Dancing Rabbit Ecovillage. It served a co-op of about ten Rabbits, as well as our seasonal influx of visitors and work exchangers. We pride ourselves in being off grid, and fossil fuel free, so we cook a lot with solar and (primarily) wood. We actually had about 5 woodburning appliances in the kitchen -- a cob oven, an old traditional wood stove, two purchased rocket stoves(a), and a modified La Rena rocket cookstove. My neighbor Dan published a pretty good video of these appliances on his blog, Hardcore Sustainable.




The modified La Rena was the culprit for the fire. I built it to burn very hot and efficiently, and to be capable of making lots of food for lots of people (we had the other appliances because it was less practical for small meals). It had a small insulated batch feed, leading into a highly insulated riser, which came up under the first burner. From there, the flue did a horseshoe around the remaining burners. The cooktop was a half inch steel top with two trivets fabricated under the first and second burner(b). Once around this horseshoe, the hot gasses continued up the Class A chimney.

Air was preheated by traveling below a steel plate under the burn tunnel and batch feed(c), where it then came up two air intake channels on either side of the firebox. By placing a cake pan over the top of the firebox(d), you could force the fire to burn this preheated air. The firebox, riser, and flue under the cookplate were lined with insulating firebrick (the super light kind). Outside these bricks I used a perlite cement (a super light mixture, roughly 1 portland::18 perlite) as additional refractory insulation(e). Outside this perlite later was a walnut facing. You can guess how the fire happened.


This stove served us for about five years. By my estimates, we made about 1,000 fires in it. Then, during a cool November night, it burned the kitchen down.

What happened was this -- a couple of folks had harvested a bunch of garden produce before a hard frost, and used the stove to can a bunch of jars for winter. I remember going over there that day, and I remember how cheery a scene it was, the heat from the stove feeling good in the crisp air, with multiple massive pots filled with scalding water and an abundance of food sitting atop the vibrant cooktop. They spent the bulk of the day there, and eventually concluded their canning sometime between 8 and 9. Long after they went home, an insidious smolder from inside the stove kept its breath. This smolder, once tiny, continued to grow over the next three hours. Eventually, around 12:30am, a neighbor woke up and looked out their window, to see flames creeping up the walls. I was among the first on the scene, and within a few minutes, the core of the building was engulfed in flame. The building was completely lost (except for, of course, the cob oven).

Very fortunately, it was a a still night, and our local fire department made it to the scene quickly (it helps when the fire chief lives in the village), so no other buildings were burned, and no one was hurt. I was also grateful that the fire I was ultimately responsible for burned a building down that I also built. But this was small consolation when I consider the buildings and lives that were at risk that night.

The folks who did the canning felt terrible, but it wasn't their fault. The stove was the culprit, and I was accountable. I'm writing this post to impart considerations for others based on the lessons I learned here rather than focus on accountability or blame. I've expressed my extreme regret, sorrow, and apology in the appropriate channels, but it's worth saying again -- in my opinion, I was accountable for this fire, and I feel very sorry and pretty terrible about it.

I had built the enclosure for the stove out of wood for a number of reasons.
- It was fast and easy to use wood first as an enclosure, which then could be filled with the crumbly perlite/cement insulation.
- It was cheap. The entire kitchen was built for next to nothing (labor not considered...), I estimate the cost of all materials was less than $500. We try to make eating in the kitchen very financially accessible, which means we didn't want to transfer a high building cost to the kitchen users.
- It was light. The stove was built on a wood deck (see more below), and a masonry facing's weight was more than the deck could handle.
- Because the appliance was a cookstove, I reasoned that any use of the stove would have people (or at least the cook) around (versus a home appliance where people might be absent or sleeping), and any smolder would create smoke, which would be quickly discovered.

In hindsight, of course, none of these reasons justify the failure and danger that culminated that fateful night. All of them would have been seen differently when considered against the lessons learned at the top of the article.

I don't like building codes, and I'm not advocating for the strict adherence to them in every context (because those contexts can vary so widely), but if ever there is a sector of building in which the builder should consider following code, it is building that contains fire. In the very least, fire builders should be familiar with code and understand why it says what it does.

I am still in love with fire, and will continue to build appliances for it. But I have regret from this experience, and an increased understanding and respect for due safety protocols. When we build appliances without appropriate safety design, we put property and lives at risk. Furthermore, we turn folks away from fire by making it a danger and liability, when we all know what it can and should remain -- a powerful tool capable of unleashing renewable energy amidst beauty, warmth, and joy.

Please consider these points as you go forth and bring fire to our world.

alis

P.S. An upshot to living in community is you've got a lot of friends to help clean up your messes.



Footnotes:

(a) I think these rockets may have been donated. They're nice because they're compact and movable, and insulated so they burn pretty well. They're also extremely intuitive for first time users. They also cover everything in soot and smoke gets everywhere, which was one reason to build the La Rena in the first place. You can see in the picture the wood floor; also a fire hazard, as burning sticks will sometimes fall out. The next floor will be brick.


(b) Some more about the stove -- originally the top was solid, as shown in this pic and the Hardcore Sustainable video. We later added the trivets for increased versatility. Someday maybe I'll make another post about things I'd do differently about the design (other than safety stuff), but for now I'll say that the half inch cooktop was too thick. One reason for half inch was out of concern for buckling/flexing. Another was for heat stability. I had previously received advice from another stove builder to make it "as thick as possible" (he went on to say that a manhole cover would be awesome...). Of course, the half inch had too much mass and took too long to heat up. So we fabricated it to have two trivets holes, and two trivets for each hole (one solid and one with holes). These trivets helped a lot, but even with them the half inch steel was an issue because they were heavy to use.

Additionally, because the top was one piece, heat from the first burner spread very quickly to the back burners, making the whole top less responsive (the fourth burner area (near chimney) was hotter than the area off the right. If I rebuild using this top, I'll cut a groove between the first burner area and fourth burner area, so that the first burner heats up faster and loses less heat to the back.

Here is a video showing the cooktop with trivets.

(c) I'm always interested in finding ways to preheat intake air for increased efficiency of combustion. However, in this case, I'm aware that air being preheated under the firebox in this way is also cooling the fire, and therefore negating a lot (all?) of the benefits. I still don't understand this relationship exactly. My thoughts are that perhaps it is worth warming the air at the cost of cooling the burning embers from below, because the air is directly involved at the point of combustion (where it reacts with the fuel), and therefore more important for this air to be hot than it is for the sides/bottom of the firebox. But perhaps this is wishful logic. Ideally, the air would be heated by the flue downstream of the combustion, which wouldn't adversely affect the efficiency of the combustion, past the point where you're using the heat (above the cooktop in this case). The only consideration as I see it from preheating air in the downstream flue would be to not cool it to the point of reducing draw (which seems like a pretty small issue if the flue is appropriately designed aka straight). Here is a video that shows the base / air intake, in probably more detail than anyone wants.



(d) I liked this cake pan design and will likely incorporate into future designs. You can always leave the pan off and burn longer wood, or attend the fire more easily. This cake pan also helped reflect heat back into the fire for a cleaner/hotter burn. The cake pan was also a safety feature, as burning long sticks in an open rocket feed can allow flames to creep "backward" along the sticks, and lead to fires. The cake pan prevented air/flame/smoke from burning backwards. I told people they could leave the fire unattended (to got get stuff from the garden, etc.) if they had the cake pan on. You can see the cake pan on the left side in this pic (the bricks at the top of the fire were just temporary while I was trying out something momentarily).


(e) I also used polyiso insulation in a few places, mainly underneath the base of the cooktop flue. Of all foam based insulations, polyiso is highest in temp rating (at around 400F iirc). I wouldn't do this again. I don't think this was the issue in this fire, but the gains of insulation don't outweigh the risks of burning. Even if you're able to isolate this polyiso inside a masonry facing, it could still melt eventually and cause the whole unit to settle. There's also the issue of off gassing which I don't know much about, but imagine it's enough of an issue to avoid using. Here you can see the foil faced polyiso under the flue path.


All this insulation made for an extremely hot, and therefore efficient burn. After the first few minutes of firing up the stove, no visible smoke came out of the chimney.  My logic (very faulty in hindsight) was that by the same stroke of keeping the heat inside the fire would keep heat from the combustibles, but of course it (eventually) wasn't enough.

The point remains that insulation makes for a great burn where you want it. I continue to wonder at the general lack of insulation implemented in a lot of rocket stoves (or any fire appliance), but that's another topic for another day. (There is a case against insulating inside longer burner appliances that already reach high temps, like masonry heaters, because after a point (~1800F), there becomes undue stress on masonry/flues/etc and the returns from efficiency diminish. And fly ash is created which has a very negative effect on air quality.)

Lastly, here's a video that Uncle Mud recorded with me explaining on site many of the things listed in this article. (Thanks, Uncle Mud!)

Thanks for sharing this Alis.
We had something similar happen to a business near us - a pottery. They had a beam over their kiln that started to combust after twenty odd years. The whole building again was gutted and needed complete rebuilding, but luckily no one was hurt.

alis said, " Long after they went home, an insidious smolder from inside the stove kept its breath. This smolder, once tiny, continued to grow over the next three hours. Eventually, around 12:30am, a neighbor woke up and looked out their window, to see flames creeping up the walls. I was among the first on the scene, and within a few minutes, the core of the building was engulfed in flame. The building was completely lost (except for, of course, the cob oven).

I am sorry to hear about your loss.

Did the folks that were canning fail to make sure the fire was out?

This thread suggests using sand if the fire needs to be put out quickly:

https://permies.com/t/143776/practice-putting-fire-rocket-stove

That looks like the building was a lovely structure.  Do you have plans to rebuild?

Anne Miller wrote:

Did the folks that were canning fail to make sure the fire was out?

That looks like the building was a lovely structure.  Do you have plans to rebuild?

I'm not sure, but this wasn't the issue. With the cake pan on, the firebox is contained. The issue wasn't with coals in the firebox, but with a smolder in the wood enclosure. I can't be sure, but I suspect the smolder began in the wood around the cooktop (which would have been the case even with a cold firebox.

This actually happened November 2021 (it's taken me a year to process...), so yes, we are in the middle of the rebuild now. And I'm happy to say the new structure is going to be quite a bit more splendid than the first!

Hi Alis;
Thank you for sharing, we all need reminders not to become complacent around high heat and potential fires!
Please post a few pictures after you rebuild.

I built a new outdoor kitchen last year and I used clay brick rather than wood.
Three-foot tall walls with corner support columns.  A tall steep wood frame roof lets all the smoke out.
Very fire resistant and quite good looking, although tiny compared to your build.

Thank you for explaining.

I hope you will share when the new build is completed.

Maybe do a "Project" thread during the build.

https://permies.com/f/69/projects