Quest for Fire: A Rocket Mass Heater for the Greenhouse

Being off grid (without utility hookups to water, gas, and electricity) is challenging even just for a dwelling –  but even more so for a farm, especially in this climate. In the spring, we have to start our plants indoors, so they will be developed enough to produce before it gets too hot for the cool weather plants in summer, and before the killing frosts of fall take out the heat lovers. They all need to be kept warm day and night, especially when germinating from seeds – but they also need to be under good bright sunlight. The sunlight is great in the greenhouse, which also gets plenty warm most days if it’s not totally cloudy – but greenhouses lose heat quickly after sundown. We solved for this in previous years by nightly driving all the baby plants up to the trailer, where they slept toastily on floor-to-ceiling shelving next to the wood stove.

plus trays on the front passenger seat and foot well.
plus trays on the front passenger seat and foot well.

Then every morning just after dawn, we drove them all back down to the greenhouse for their daily sunshine. We worked out a good two-person method to get it done as efficiently as possible, but it was a time-consuming process – and we couldn’t have the lights on near the plants when they were inside (or they’d get leggy trying to get fed by the artificial lighting). Yeah, it worked … but we needed to come up with something better.

Heating the greenhouse at night was the obvious solution, but it seemed impossible – the plastic covering lost heat so quickly that it would cost us a fortune in propane … and we’d always be fearful that the tank was going to run out in the night, leaving all our crops to freeze to death. And of course our little solar power system was not capable of generating enough electric heat (which is inherently inefficient) to do the job. A wood stove like the one we use in the trailer would require repeated fueling throughout the night, and again, the heat would mostly be going to the roof and then out through the plastic film.

Research led us toward a possible solution – “rocket mass heaters,” which use small amounts of wood burned efficiently at high temperatures to heat up a thermal mass, which then slowly radiates the heat outward for many hours after the fire has gone out.

Rocket-Stove-Mass-Heater

It would still be a challenge to heat the whole greenhouse with this method unless we built a prohibitively large heater, however. Hmmm.

Well, we’d already been experimenting with a technique that Elliot Coleman promotes, using row cover fabric to create a greenhouse-within-a-greenhouse – just putting the plants on the ground and covering  them inside a low tunnel of fabric kept them warmer than the surrounding air, thanks to the day’s heat in the soil. That only got us a few degrees, but it was a crucial difference when temps might be just barely dipping below freezing.

What if we did the same thing atop a thermal mass? It seemed like a winning idea – so we decided to build a heated bench sized for seed trays, and put a cover over them at night to hold heat in.

For a year, we researched and gathered materials for the build – scavenged from garbage and the Free section of Craigslist; two trailer loads of clay from a retired sculptor, stove pipe of various diameters, a 55 gallon drum, river rocks from the neighbor’s pile they’d unearthed when building their house, cement board, and bricks – the best score of all was a huge free load of white “insulative firebricks,” which withstand high heat without heating up much themselves – ideal for our purposes.

IMG_0741

Last fall, after the killing frost had come and the 2015 CSA was over, we got to work, starting to test combustion chamber designs the day after our CSA potluck.

IMG_1120

There was a long period of trying out different brick layouts – the proportions of the combustion chamber are crucial to get right – for the rest of the fall, “CSA” ceased to mean “Community Supported Agriculture,” and became ‘Cross Sectional Area” – the entire system had to maintain the same CSA throughout the system, from fuel feed to exhaust, from rectangular tunnels to round pipes. I won’t bore you with all of the proportional rules that had to be followed, but know there was a lot of subtle and annoying math at this stage.

an early mockup, seeing how tall the system would be - and how close we could put it to the curving north all of the greenhouse. (The original plan was for a 8" pipe system, although we switch to a 6" design before building)
an early mockup, seeing how tall the system would be – and how close we could put it to the curving north all of the greenhouse. (The original plan was for a 8″ pipe system, although we switched to a 6″ design before building)

 

creating a lid for a "pocket rocket" - a way to create intense fire that would burn the paint off the 55-gallon drum
creating a lid for a “pocket rocket” – a way to create intense fire that would burn the paint off the 55-gallon drum
burning the paint off of the 55 gallon drum - so it wouldn't create toxic fumes in the greenhouse later
burning the paint off of the 55 gallon drum – so it wouldn’t create toxic fumes in the greenhouse later
early layout with the larger pipes - determining how many rocks we'd want, how wide it should be for the seed trays, and how deep we'd want to make it,
early layout with the larger pipes – determining how many rocks we’d want, how wide it should be for the seed trays, and how deep we’d want to make it,
The WWOOFers joked about me playing with my blocks - I had bricks in the trailer that I'd use to try out different configurations for the heat riser ... on edge? flat? how much space inside? It was a lot more complex than I'd thought it would be ...
The WWOOFers joked about me playing with my blocks – I had bricks in the trailer that I’d use to try out different configurations for the heat riser … on edge? flat? how much space inside? It was a lot more complex than I’d thought it would be …
even in bed it was rocket heater worktime
even in bed it was rocket heater worktime

 

Many variations were tried out and discarded in the design process. The red bricks are used in the fuel feed for their durability, while the softer insulative bricks were used throughout the rest of the combustion system.

IMG_1152 IMG_1153 IMG_1155 IMG_1154

determining bench dimensions. (for awhile we thought we'd use foam insulation but wound up not doing so)
determining bench dimensions. (for awhile we thought we’d use foam insulation but wound up not doing so)

 

creating the subsurface foundation for the combustion chamber - the air spaces in the bricks add insulation to keep heat where it's wanted - not absorbing into the ground
creating the subsurface foundation for the combustion chamber – the air spaces in the bricks add insulation to keep heat where it’s wanted – not absorbing into the ground

 

IMG_1178

laying the base layer, with pits for ash cleanouts. White bricks insulate against heat loss, keeping temperatures high for the heat riser, where even the smoke will burn up
laying the base layer, with pits for ash cleanouts. White bricks insulate against heat loss, keeping temperatures high for the heat riser, where even the smoke will burn up

 

IMG_2424

constant vigilance was required to ensure that proper dimensions were maintained in all directions
constant vigilance was required to ensure that proper dimensions were maintained in all directions

IMG_1218 IMG_2456

had to learn how to score and split bricks with a chisel to make some of the puzzle fit together
had to learn how to score and split bricks with a chisel to make some of the puzzle fit together

IMG_1230 IMG_2460 IMG_2462

heat riser is Widget-approved
heat riser is Widget-approved

IMG_1239

fitting the barrel
fitting the barrel

 

exhaust manifold - where falling gases inside the barrel are channeled into the bench exhaust pipe run
exhaust manifold – where falling gases inside the barrel are channeled into the bench pipe run
although the heat riser was made of insulative fire brick, I realized it would be even better with more insulation - so I cut apart an old water heater and added it - later filling the space between it and the riser with perlite insulation mixed with clay slip.
although the heat riser was made of insulative fire brick, I realized it would be even better with more insulation – so I cut apart an old water heater and added it – later filling the space between it and the riser with perlite insulation mixed with clay slip.

 

heat riser insulation finished and capped off with clay
heat riser insulation finished and capped off with clay

IMG_1591

trench for thermal mass foundation/insulation layer
trench for thermal mass foundation/insulation layer
insulative firebricks and perlite to separate the thermal mass from the infinite heat sink of the earth
insulative firebricks and perlite to separate the thermal mass from the infinite heat sink of the earth
Forms for the thermal mass (they wound up bowing outward some between the internal supports, but not too badly and I like the organic wavy lines that resulted)
Forms for the thermal mass (they wound up bowing outward some between the internal supports, but not too badly and I like the organic wavy lines that resulted)

IMG_1755

And then winter came, and we hit the road, leaving the wet clay of the combustion chamber to slowly dry out over the wintertime.

When we returned in March, we got to work on the thermal mass bench … which turned out to be a lot more work than we’d bargained for.

laying out the exhaust ducting, and taping all the joints.
laying out the exhaust ducting, and taping all the joints.

 

Kristin did all the mixing. ALL of it. We mixed 2 parts sand for every 1 part of clay. We went through at least a full yard of sand (two trailer loads) and almost all of the free clay we'd scored. It took us several solid days to get it all mixed and added to the bench ... Kristin's feet and my hands looked like they'd been to war by the end.
Kristin did all the mixing. ALL of it. We mixed 2 parts sand for every 1 part of clay. We went through at least a full yard of sand (two trailer loads) and almost all of the free clay we’d scored. It took us several solid days to get it all mixed and added to the bench … Kristin’s feet and my hands looked like they’d been to war by the end.

 

Exhaust duct was coated in clay slip for maximum heat transference. We tried to use as much river rock as we could to save on clay and sand, but it still took an incredible amount to fill the bench (23' long, 14" high, and 24" wide - wider where the forms bowed out)
Exhaust duct was coated in clay slip for maximum heat transference. We tried to use as much river rock as we could to save on clay and sand, but it still took an incredible amount to fill the bench (23′ long, 14″ high, and 24″ wide – wider where the forms bowed out)
first test fire! It went .. OK. We decided to wait until we had more completed before really testing her out.
first test fire! It went .. OK. We decided to wait until we had more completed before really testing her out.
feeling the heat at the top of the riser
feeling the heat at the top of the riser

IMG_4014

slowly filling the forms with clay/sand mix and rocks. Kristin down at her mixing station.
slowly filling the forms with clay/sand mix and rocks. Kristin down at her mixing station.
ha ... so it got too hot for pants in the sunny greenhouse
ha … so it got too hot for pants in the sunny greenhouse

IMG_4024

Music was a required tool to help maintain spirits during the tedious mixing process.
Music was a required tool to help maintain spirits during the tedious mixing process.

IMG_4029

Kristin developed a technique of slicing off slabs of clay and spreading them out on the sand, then topping that with more sand, before starting to stomp and twist.
Kristin developed a technique of slicing off slabs of clay and spreading them out on the sand, then topping that with more sand, before starting to stomp and twist.

IMG_4628 IMG_4042

Finally getting to the top layer! The last foot or so seemed to take forever to fill in. (This is still sand/clay mix - but some of the clay was a white instead of red.)
Finally getting to the top layer! The last foot or so seemed to take forever to fill in. (This is still sand/clay mix – but some of the clay was a white instead of red.)
end of the bench before the exit to the chimney - was insulated with perlite/clay mix end cap
end of the bench before the exit to the chimney – was insulated with perlite/clay mix end cap

IMG_4671

Bench completed!! Now to start drying out several tons of wet clay ...
Bench almost completed .. just had to let it set up before removing thos internal braces, and filling in the gaps they left behind. Then it was time to start drying out several tons of wet clay …

IMG_4722

Chimney V1 - uninsulated and rather short. Initial burns couldnt even drive smoke up the pipe - it just pooped out onto the ground through the open cleanout hole at the bottom, because all the heat had ben leached out by the wet, cold clay bench.
Chimney V1 – uninsulated and rather short. Initial burns couldnt even drive smoke up the pipe – it just pooped out onto the ground through the open cleanout hole at the bottom, because all the heat had ben leached out by the wet, cold clay bench.
added some thermal mass to the barrel to limit amount of heat lost to radiation off the metal. Using the heat at the top for a wood drying rack. Artichokes enjoying the radiant heat.
added some thermal mass to the barrel to limit amount of heat lost to radiation off the metal. Using the heat at the top for a wood drying rack. Artichokes enjoying the radiant heat.
Chimney V2 - taller. (I don.t
Chimney V2 – taller. (There doesn’t seem to be a photo of Chimney V3 aka “Paul Baxter” – with insulation and a wind cap courtesy of a nameless donor.)

 

IMG_4820

Booster fire! This proved crucial during the bench-drying phase for a fast drafting system with hot, oxygen rich flame - the little fire would be pulled up the chimnet, creating suction on the rest of the system, pulling air through the fire in the combustion chamber. (Now that it is dried out, we only use this to start the system up, to avoid cold air plugging up the system)
Booster fire! This proved crucial during the bench-drying phase for a fast drafting system with hot, oxygen rich flame – the little fire would be pulled up the chimney, creating suction on the rest of the system, pulling air through the fire in the combustion chamber. (Now that it is dried out, we only use this to start the system up,  avoiding a cold air plug)

 

the wickets that support the row cover fabric are visible here (yard sign posts we scavenged)
the wickets that support the row cover fabric are visible here (yard sign posts we scavenged)

 

Before the chimney was upgraded further and before the thermal mass was fully dry – it was already working! Temperatures on the plant try bench stayed over 20 degrees warmer than outside, night after night – with no need to tend the fire after bedtime, and zero risk of the heat going out in the night – you can’t stop a warm giant rock from radiating! Plus it was warming the soil itself, and not just the air around the plants – great for happy, healthy root development.

IMG_4941 IMG_4828

IMG_5170
toasting bread for an experiment in making kvass

IMG_5166

reheating and crisping leftover pizza while firing the mass for the night
reheating and crisping leftover pizza while firing the mass for the night
added a blast furnace window to the aluminum pot I use as a lid (saved from a now demolished Ford assembly plant)
added a blast furnace window to the aluminum pot I use as a lid (saved from a now demolished Ford assembly plant)

IMG_4932 IMG_4936

It’s been a success. In recent nights just below 40 degrees, it kept the hot weather plants happily over 60 all night – maintaining the same 20+ degree heat increase seen at lower temps. It feels great to have it finally done, and really working wonderfully for our needs. It was a lot of work, but it was absolutely worth it. Plus, now we have experience in mass heater construction, so a mass heater in a future home is definitely a possibility …

 

8 thoughts on “Quest for Fire: A Rocket Mass Heater for the Greenhouse”

  1. I’m not scientific, and not living off the grid, but this is an amazing story and process. Impressive! The two of you worked so hard and seem to be a great team…thanks for sharing . Will you patent this?

    1. thanks! I don’t think we can get a patent on it – it’s not a common implementation, but I’m sure there are many others who have come to similar conclusions – in fact, someone from Japan posted a similar system when I shared the video with a Rocket Mass Heaters group on Facebook. It sure was a lot of firsts for us, though!

  2. Hooray. Thanks for sharing. Our rockets help us heat our house for $75 per year. Building one right now for my daughter’s live-in treehouse. Also terrakote.com works great for adding water resistance to the cob bench –Uncle Mud

  3. I am living in Ontario Canada and am looking for someone who will help me build a rocket mass furnace for my greenhouse. If you know of anyone I would be very grateful.

    Thank you

  4. Very impressive indeed!

    What type of maintenance does it require?

    Are there some things you would have done differently now that you’ve gone through the process?

    Thanks,
    Stephen

    1. For maintenance I just have to scoop ash out of the burn chamber between uses. If I did it over I would probably use stronger forms for building the bench; mine bowed out quite a bit, requiring considerable extra clay to fill (and creating a larger mass). I’d also have bett r insulat d beneath the combustion tunnel, and likely under the bench as well, rather than trying to save money on perlite.

  5. This adventure was documented really well and sharing it is really appreciated. We will be building at least one rocket mass heater this year and will be using either hempcrete or kenafcrete for insulation. I will share our adventure as well on Kenaf Partners USA our Facebook page. Keep up the good work and thanks again for sharing.

Leave a Reply

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