During the first good cold snap of 2018, it had become excruciatingly clear just how ineffective the minimal insulation I’d installed was at keeping a giant metal box covered in single pane windows warm. Any sort of electric heating was completely off the table due to the limited power budget. I briefly considered a propane heater or a diesel heater, but neither was a particularly compelling solution, especially since one of the design goals of the bus is to be able to operate off-grid as much as reasonably possible.

Which brings us to wood heating. I love heating with wood. There is nothing quite like basking in the radiant heat of a fire, or the soft heat of warm stonework. Small wood stoves are available and would be a great fit for the bus, but tend to be pretty spendy, and most of them want a stovepipe that goes straight up some not-insignificant distance. This would have necessitated either snaking a stovepipe out one of the windows and up the side of the bus, or putting another hole in the roof, which I was not about to do if I could avoid it.

So, instead of cutting holes in the roof for a conventional wood stove, I decided to wander off into uncharted territory and engineer a rocket mass heater!

Advantages of Rocket Mass Heaters

  • Generally quite efficient
  • Extremely low particulate emissions- when tuned right a rocket stove burns so completely that exhaust is mostly CO2 and water vapor
  • Needs much less stovepipe rise to draw properly than a traditional stove
  • Capable of supporting long horizontal exhaust runs, which translates to much higher heat transfer to the interior

Disadvantages of Rocket Mass Heaters

  • Prior to this undertaking, I had absolutely no experience whatsoever designing rocket stoves

Key design criteria

  • Combustion chamber needs to be insulated and impervious to temperatures in excess of 1000 Celsius. This requires some form of masonry
  • Rocket stoves burn hot and fast. Thermal mass is needed to act as a “battery” and slowly release captured heat
  • This rocket mass heater needs to be sturdy and relatively impervious to vibration and shock
  • This rocket mass heater must not catch the bus on fire

This combination of requirements drove me towards a hybrid solution- a structural metal containment vessel supporting a masonry core which serves double duty as thermal mass and combustion chamber insulation.

Construction

First step was to build the containment vessel. For this I utilized old propane tanks salvaged from metal recycling at the dump, an old air compressor tank and four of the many bus seat legs which were lying around. Some cutting, grinding and welding later, I had a workable containment vessel.

A leaky air tank about to begin its next life

With the containment vessel complete, I took a trip up to an old clay pit on Blackwater River State Forest and collected several five gallon buckets of clay plus some coarse grained sand. The only item I had to purchase was a bag of perlite.

Form for the feed tube and combustion chamber assembled from 2.5 inch cardboard tube and masking tape

All ingredients ready to mix
50/50 clay and perlite mix
Only enough water was added to enable workability
Much packing required
Finished and ready for a long curing period

If I had it to do over again I’d probably introduce fiberglass (the same stuff used for reinforcing concrete) into the mix to improve the durability of the finished product, but as it is the mix of clay and perlite has been reasonably durable. Significant shrinkage was experienced during the multi-month curing period, to the point where the masonry almost entirely separated from the containment vessel in all dimensions. I used dry sand to fill all the gaps, then finished with a sand and clay plaster after curing and running the stove several times.

I completed the masonry element of the stove in early December, and did not get around to the first lighting until February, as I wanted to ensure the masonry was dry as possible before firing.

Low and slow first firing meant lots of smoke as the mildewed cardboard tubes burned off

Testing revealed that while the stove worked well without the drum in place, it did not draw sufficiently with the drum in place to prevent some smoke escaping from the feed tube. Most of the designs I’ve seen use a larger diameter feed tube and combustion chamber with significantly more volume in the drum, so I hypothesize that this rocket stove is too small to be naturally aspirated.

Once everything was tested I finished the clamps for the drum, prepped and painted all exposed metal surfaces, then left the stove to sit for several months while I worked on other things.

Installation

Only after virtually everything else was finished did I get around to installing the stove in the bus. By this point it was named Reggie and Hannah had come around from her initial skepticism to grudging acceptance of its unconventional styling.

The bulk of the installation work consisted of fabricating the exhaust and clamps (more recycled stock, much grinding and welding), though actually getting the stove inside was a bit of a challenge on its own since we’d long had the back of the bus closed off and had to enter through the front door. Reggie may be small, but he’s quite dense and probably weighs in the neighborhood of 150-200 lbs.

After some consideration I opted to route the exhaust under the bus instead of out the side. This simplified the mounting system, allowed for a secondary expansion and fresh air mixing chamber ahead of the exhaust fan, and eliminated any external visual indicators that there is a wood burning stove aboard. There is a bit of smoke while warming up, but once at operating temperature and provided dry fuel there is scarcely any smoke at all- surplus oxygen and high combustion chamber temperatures result in exhaust that is mostly water vapor and CO2.

Exhaust assist is provided by a 12v ventilation fan- the cheapest fan I could find with a fully enclosed motor and aluminum blades. The fan is mounted in a recycled propane tank under the floor which serves as the secondary expansion and mixing chamber. The fan is controlled via a PWM controller mounted beside the primary ventilation unit control. Time will tell how long the fan motor lasts, but it has been fine so far- exhaust temperature is low by the time it reaches the fan, so thermal failure seems unlikely.

Yet another hole in the floor...
The exhaust makes a great toe warmer
Somewhat counter-intuitively, no smoke escapes
Firewood and starter material lives behind the drum (we move it when Reggie gets lit)
Burning merrily and smelling of hot enamel pain

Conclusions

We’ve now used Reggie a number of times. Initial concerns regarding masonry durability were unfounded- Reggie has survived roughly 6000 miles of travel with only minor cosmetic damage. Fuel is easy to come by- sticks up to 1" in diameter and two feet long burn readily and self-feed into the combustion chamber.

The only issue we’ve had is the black paint offgassing from the drum once it heats up- the odor is foul and strong even after several hot fires, and cannot be healthy. We need to run several more hot burns with the windows open and the bus vacated before writing it off, but so far I’m not impressed with the grill and stove paint we used. Fortunately we haven’t encountered much cold weather, and are now heading south along the Pacific coast, so still have another month or more before we’ll need to start using Reggie in earnest. If the offgassing issue hasn’t been resolved by then I’ll take more drastic measures involving an angle grinder, a wire brush and an afternoon spent in PPE.