Biochar Ovens

A number of people have requested information on low tech biochar production. This communication concerns some biochar oven designs that UBI/MoBI have begun working on. I have included a few pictures and brief comments on them to give the reader some idea of the designs and their use.

I use the term BIOCHAR OVEN (or simply ‘oven’ in context) to designate a simple, low tech pyrolysis apparatus dedicated primarily to the production of biochar, usually with a feedstock capacity of 50 – 5,000 l. Also, there is no intentional admission of oxidizers during pyrolysis into the feedstock chamber in a biochar oven.

We focus on ovens because they are more efficient and environmentally friendly than adaptation of more traditional methods of low tech charcoal production. The reasons for our focusing on low tech units in this size range will be readily apparent on reading a brief summery of the UBI concept that can be found at

‘Biochar stove’ or ‘kiln’ have also been used for these types of units. However, biochar stoves (and stoves in general) primary are for cooking and heating. They are usually of smaller capacity as well. Kilns can be of very large size and often have internal burning, particularly when used in charcoal making. ‘Retort’ is another word that is sometimes used, but they can be of very large size and the word is not familiar to the general public. We often deal with the general public and one advantage I see is that ‘oven’ is readily associated by them with what the apparatus is designed for. For all of these reasons I would encourage the use of the term ‘biochar oven’ for these sorts of units.


1. Hollow Core Fired (HCF) drum with lateral core arms design (KJF ‘08).

The next step in developing the Hollow Core design was to replace the simple stovepipe core with a heavy pipe of slightly larger diameter.


2. Holes were cut in the pipe large enough to loosely hold horizontal, open ended hollow pipes.


3. These pipes were arranged in a squed ‘X’ pattern with a hole cut in them at mid length that was faced up the core when the oven was placed in the firing position. The loose fit of the cross pipes in the core pipe holes allowed for p-gas (which is generated at a positive pressure) to enter the core and burn there. Although the primary design function of the cross pipes was to conduct heat outwards, the open ends and hole at the middle also allowed for generated p-gas to be carried into the core for burning. The upward rushing air in the central core may have caused a slight pressure decrease as well as it moved around the horizontal pipe and its central, upward facing hole. This might draw hotter gas generated nearer the core out towards the periphery. We saw no evidence that this drew O2 into the biomass chamber from the holes in the core pipe.

Firing was carried out in the same manner as illustrated as for the HC ExF design (see part 2).


4. As can be seen these modifications were still insufficient to cause good charring throughout a load of sawdust.


5. An improved heat retention cowling was made of cemented bricks surrounded by a layer of insulating sawdust for the next trial.


6. For that trial the drum was partitioned into vertical thirds. One each was filled with sawdust, ‘paddock’ (winter camp, packed) manure and field dried manure. As before, the sawdust was not charred completely through, nor was the paddock manure. However, the field dried manure was well charred throughout.

The paddock manure had been broken up into small pieces (though not as small as sawdust) to facilitate drying but still was not well dried. Both factors (dampness & packing) would impede heat penetration. However, the field dried manure was well dried and of relatively large size making for large pore size between pieces which facilitated heat penetration.


7. A final trial was run with a drum load of commercially available larch stove wood with a few pieces of thick larch bark. The wood and bark were well charred. A great deal of clean burning flame exited the top of the cowling during the burn indicating that there was insufficient O2 entering the core at the height of pyrolysis to allow for complete burning of the p-gas there.


The HCF design: It seems adequate for making biochar of dry feedstock that does not pose the heat penetration problems of sawdust. However, a number of design innovations may make it more efficient or practical.

Oil’ drums: ‘Oil’ drums (and the similar type with removable clamp lids) are made of fairly thin gauge metal and cannot be expected to stand up to long repeated usage. However, where available they can be inexpensive and readily adaptable and may still be more economical or practical than similar ovens constructed from scratch using heavier gauge metal. It would not be necessary to have a tightly fitting clamp lid if the lid is situated at the bottom of the drum, as any escaping p-gas or oil will be burned by the heating fire. This will be true even if the fire is confined to the hollow core if the lid is recessed within the rim of the barrel – the updraft in the core causes air along the bottom side of the drum to be sucked into the core fire. This means that standard bung type drums for containing liquids could be used by simply cutting free the lid and using the resulting disk as the bottom, held in place during firing by removable and/or fixed pegs.

Cylindrical polygons: Where metal ‘oil’ drums are either not economical or unavailable, oven drums can be fabricated from desired gauge metal. If equipment is not economically available to fabricate round cylindrical drums and hollow cores, then cylindrical polygons of 4 or more sides could be made by either break bending metal sheets or welding together rectangular panels. A removable lower end need not be epically tight fitting if it is slightly recessed for reasons described above.

Cowling: The cowling or oven housing is used primarily to improve heating efficiency. However the structure can be relatively expensive and make production operations cumbersome. With fire primarily concentrated in the hollow core, the cowling could be reduced to insulation iatrical to the outer wall of the oven. This should decrease initial expense while increasing mobility and ease of production operations.

Increasing heating efficiency: In the dry larch wood trial especially, a great deal of smokeless flame was seen immediately above the hot gas vents of the cowling at the height of the pyrolysis process. This indicates that the oxygen in the air being drawn into the hollow core was insufficient to fully burn the p-gas being vented there. It would seem that a series of channels could be designed immediately external to the outer drum wall that could draw down this hot p-gas, passively charge it with additional oxygen rich air and have it burn along the outer sides of the drum. Keeping it low tech in this way would be preferable to using some sort of forced draft which would add to operational expense and be difficult in rural areas where electricity or fuel for auxiliary engines might be hard to come by.

Gas tight horizontal pipes: Drawing the hot p-gas being generated nearer the core into the outward lying biomass may enhance heat penetration efficiency. This might be facilitated by sealing the connection between the horizontal pipe and the core. This could be done by either mating them with a screw fitting or welding them together. If the latter were done with the originally designed through-pipe with upward facing exit port, advantage could be taken of the greater pressure differential. A more even distribution of heat through the biomass load might be achieved by connecting each horizontal cross pipe to a feeding ring around the inside of the wall of the drum with inlet ports evenly distributed around the circumference.

Up-sizing: Fabrication of the drums would free us to increase drum volume through our size range of interest (50 – 5,000 l). Merely making the drum longer would probably have little effect on the other details of design except for accommodating the oxidation of the greater volume of p-gas generated. This might require increasing the diameter of the core and capacity of the outer drawdown channels if they are used. However, increasing the diameter of the drum would also require adjusting the diameter of the hollow core for adequate heat penetration of the biomass. Perhaps it would become necessary to switch to a pattern of multiple hollow cores within the drum as the diameter increases.

Horizontal burning: We have begun testing this design with the drum stood vertically to accommodate the external fire of the initial design phase and for the expediency of simplifying the makeshift cowling design. With fire confined primarily to the hollow core (and ‘draw down’ channels if they are incorporated) and a properly designed cowling, it should be able to operate the oven in a horizontal position which might facilitate biochar production operations, especially with longer drums.

Adaptation for sawdust: Although the current design was inadequate to completely char a load of sawdust, the design should be adaptable to sawdust by adjusting the diameter of the hollow core to the degree of heat penetration achievable by the burning of p-gas generated and any auxiliary fuel that is economically available, as indicated by the ‘20 cm Box’ design. (Details of the 20cmB design to follow.) However, this would decrease the volume of the load, though perhaps a second or third concentric drum could be added with the firing occurring between them.

Analysis of performance: In addition to testing some or all or the modifications alluded to above we plan to characterize the performance of these designs, including temperature profiles in various regions of the drums during operation using thermisters and data loggers as well as characterizing biochars produced from various feedstocks.

Open source designs: To the extent that these designs are not covered by priority or patents unbeknownst to me (I consider the Hollow Core designs original work), I consider the designs “open source” and invite their use and improvement asking only that I be kept in the loop. I hope to continue to work on improving the design and to make available the results, including detailed data, on the UBI and other web sites. I would like to invite all interested to collaborate in creating a data pool on the Hollow Core design, its construction, operation, maintenance and the characteristics of the biochars produced. You can contact me at

For more information on the UB International program please see our web page at:

I would like to thank the MoBI NGOs Mongolian Women Farmer’s Association for making their facilities and labor available for this work and People Centered Conservation of Mongolia for helping with funding.

-frog 9 Dec ‘09 

Karl J. Frogner, PhD, OZP
President & Project Development Head; UB International (UBI)
Advisory Committee, International Biochar Initiative (IBI)
Project Development Head; Mongolian Biochar Initiative (MoBI)
Project Development Consultant; Thai Biochar Initiative (ThBI)