Saturday, June 11, 2011


One of my many life-time interests have been primitive and survival technologies, especially related to firemaking, cooking etc. Recently while researching another subject I came across some information related to new and improved types of woodstoves. Many of these could be improvised from cans, scraps of metal etc. Some incredible claims were being made for these stoves and I decided to investigate. After all, if one lives on a fault line one might be called to make one of these for real at some point.

There are many sources of information on these woodstoves on the net. There seem basically two kinds: The Downdraft Gasifiers and the Rocket stoves. The common principle is complete combustion of the fuel (often referred to as "biomass") which increases the calories available form a given quantity of fuel while at the same time reducing polutants such as smoke and carbon monoxide.

The best article on Downdraft Gasifiers is here:

It explains the principle in some detail which I am not going to repeat. Suffice to say that burning fuel from top to bottom creates a pyrolysis zone below the combustion zone, wood gas is generated there which then rises and is mixed with secondary air and burns to provide the actual heat for cooking etc. There are many versions of these described on the net and many videos on You Tube. Not all of these are actually what they claim to be and in many of them the authors show only tenuous grasp of the principles in play.

The definitive source for the Rocket stoves and its many variations is here:

The paper on design principles by Larry Winiarski is a must read. The Rocket stove is a more "linear" affair in terms of combustion: The design simply allows for a more complete combustion of fuel without the intermediate gasifier step. Again, there are many of these on You Tube etc. of varying quality.

I decided to have a go at building some of these starting with a small model to see what properties of these stoves are essential for correct functioning and how the various designs behave under different conditions.

Fig. 1
Fig. 2
I thought I would start with a Downdraft Gasifier stove. The simplicity of construction appealed to me as well as the theoretical ability to use a wide variety of fuel. I made one from two bean cans, 75 mm in diameter and attached their open ends to each other by cutting one in four places and sliding it into the other. I cut out one of the bottoms and deburred the sharp ends to make this the cooking end of the stove. I drilled holes in half of the bottom of the second can as well as one of the detached lids which I then riveted to the bottom thus providing myself with an air-flow regulator (Fig.1). More holes near the bottom and a row of holes about half way up the assembly to provide the secondary air intake, three legs riveted to the sides, and voila (Fig.2)

I tested the stove on top of my BBQ. The first fuel was the waste from my workshop - sawdust, chips and shavings. The result was disappointing. If enough of this "biomass" was loaded to fill the stove to the level below the secondary air intake the fuel burned very quickly and was used up in a couple of minutes. Loading the stove to the rim allowed the fuel to burn until the level of secondary air intake was reached and then the fire went out whatever one did . I interpreted this as lack of air flow through the primary holes. I increased the number of holes in the bottom (the air inlet regulator plate fell off later anyway - the aluminum rivet simply melted in half) and added few more holes around the priphery of the bottom. This made absolutely no difference to performance (Fig. 3 and 4)
Fig. 3
Fig. 4

 I figured then that the problem lies with the fuel which is too dense to allow the primary airflow (the author of the first article refers to difficulty he had gasifying pellets of 0.5 cm for that reason.)

I decided to try the stove with sticks cut from a "SPF" 2x4, guaranteed dry. The sticks were cut so the length which  corresponded to the depth of the stove. I crammed as amny as I could into the stove cavity and ignited them with a butane torch. Once the top part of the sticks burned through and the combustion zone reached the secondary air holes ther was a definite evidence of gasification in that a) there was no more smoke and b) the flames were clearly related  to the secondary air inlets (Vid. 1 and 2).


It took approximately 2 minutes for the stove to reach gasification stage. Disappointingly, the gasification stage lasted only about 7 minutes for a full load. The stove was re-loaded with cubes of the same wood appx. 1.5 cm per side. This revived the gasification but it was again short lived.

In the second video you can see an improvised shield made from aluminum foil. It did not seem to affect the pattern of the flame, it kept the wind out rather poorly. Many of the You Tube designs make much of the necessity of having such shield to pre-heat the secondary air. Whether my shield increased the temperature of the flames is impossible to determine. A boiling water test would be appropriate to determine its effect but this would require building an additional pot-holding fixture. That the nature of such fixture is critical to the stove's performance is explained in the Winiarski paper.

I did not use the "wick" as described in the first paper. It is not clear what effect, if any, adding the wick has on the stove performance. I have not seen the wick in use elsewhere and I simply did not have a suitable can to make it from.

The stove described in the first paper was, as far as can be determined, 12.5 cm internal diameter and 30 cm tall. My stove proportions were similar to that, consistent with the need to provide adequate space for the secondary air to mix with the wood gases and allow them to burn completely. Many of the You Tube variations have hardly any space above the secondary intake at all. I believe this shows inadequate understanding of the whole concept. The 3:1 ratio is also recommended by Larry Winiarski to achieve complete combustion in the Rocket Stove.

The authors had their stove burn for a total 37 minutes on a single load of fuel. It should be noted however, that at 20 minutes they reduced the primary air intake to only 20%. It is interesting to extrapolate from their graph to see what the duration of a full burn would be - I estimate it around 25 minutes. Given that the volume of fuel of my stove is about 1/4 of theirs the durations of full burn are consistent (assuming that the rates of burn is comparable which I suspect is not necessarily correct.)

The concept of restricting the primary air intake is problematic. Such restriction leads to incomplete burning of the fuel and therefore inefficiency. It also leads to increase of undesirable products such as carbon monoxide. The authors measured carbon monoxide emissions however they do not state at which point of the burn cycle. They do warn about "extingusishing the stove completely" because of this danger.

At the end of their burn cycle their stove produced a signinficant amount of charcoal. It is not clear to me what the direct benefit of making charcoal is. Howeve, presence of significant amount of charcoal confirms that at some stage the burning of fuel was inefficient. In my stove at the end there were only 4 small lumps of charcoal and I terminated the smoldering with water for safety's sake. There may have been even less if I let the stove go out by itself.

Larry Winiarski is quite clear that any air flow restricition is undesirable. The Rocket Stove regulates heat production by inserting or withdrawing fuel from the combustion chamber which, due to its design, is relatively easy to do. Such procedure in the case of the Donwndraft Gasifier is impossible, in fact refuelling the stove can be a problem while cooking.


1) The Downdraft gasifier technology is feasible and easily reproduced from readily available material.
2) The design is fuel sensitive in that dense and packed fuel will not allow the stove to function.
3) The size of the stove is of some importance due to limited burn time. It took 8 minutes to boil a litre of water with the original design thus the minimum useful size of a stove for such purpose (as defined by the volume of fuel) would be at least double that of mine.
4) Refuelling the stove is relatively awkward especially if there is pot on the stove. Additional features to allow this would add to complexity of construction.
5) Regulation of heat output by varying the amount of fuel burned other than by restricting the air flow is impossible. I find the concept of air flow restriction problematic and undesirable for reasons discussed above.
6) The concept of pre-heating the secondary air is interesting and probably should be investigated if one considered this design to be of further use. It is certainly considered essential feature of the Rocket stove.
7) The effect of the "wick" is intriguing. Same applies.
8) Overall my impression is that although this design clearly works it has many drawbacks which render it of limited use in all but emergency situation where such design is easily constructed using little but a can opener.
9) The Winiarski concept of the Rocket Stove has solved many of the above issues. Its design is clearly documented and the design principles firmly established. Hence I will experiment with it next.

No comments:

Post a Comment