| Boiling Point No. 21 - April 1990 |
by KKrishna Preread, E Sckutte and P Verhaart
Woodburning Stove Croup, Faculty of Physics, Eindhoven University of Technology, Eindhoven, The Netherlands
This contribution is based upon the work of many colleagues and students (Hasan Khan, Roger van Hock, Annemarie Dirks, Stein Kuiper and Etienne Moerman to be specific). It is perforce a short summary, but detailed reports of the work can be made available on request. The Woodburning Stove Group at Eindhoven has been formally associated with a technology geared to the conservation of wood for over 10 years now. There has always been a concern over the efficiency of combustion of wood, as can be gleaned from our earlier reports that invariably include results of gas analysis. In contrast, the work to be presented in this short report is almost exclusively concerned with the question of clean combustion of wood. The operative word here is "concerned" and thus at the outset we should like to disabuse the reader of any vision of a ready-to-use design for a cookstove that saves 50C/c of the wood compared to a conventional stove and yet produces no harmful pollutants.
The rest of this report is divided into basically three parts. The next section is a short primer on combustion of wood leading to a new concept for the design of a wood burning device. The second part provides some laboratory results comparing the performance of the proposed conceptual design with that of conventional wood burners. The concluding part discusses the prospects of incorporating some of these into practical designs.
A Wood Combustion Primer
Our concept of clean combustion of wood starts with the hypothesis that wood consists of only the harmless elements carbon, hydrogen and oxygen and thus the only products of combustion should be water and carbon dioxide with unused oxygen and nitrogen. While this might sound naive, we intend to demonstrate in the course of the article that such an ambition is not beyond the reach of a reasonable design without recourse to esoteric technical fixes like catalytic convertors and pressure supply of air etc.. Needless to say we do not consider the production of carbon dioxide by but Ding wood in a cookstove as contributing to the much-talked about greenhouse effect since we consider this to be caused by them guys who are energy guzzlers and forest burners.
Conceptually the combustion process could be represented in the form of a fire triangle first proposed bv Emmons. shown in figure 1.
This representation expresses 3 fundamental ideas behind combustion. Firstly, we need to bring the fuel and air together and this in principle means that they have to get thoroughly mixed. Secondly this mixture has to be introduced in a region of high enough temperature to sustain the chemical reactions that we call combustion. Finally there is a time element involved in the business since the processes of mixing and heating do not happen instantaneously but do take time (one might measure this in seconds and unfortunately many traditional systems do rot' the combustion process of these precious seconds).
What we have described above is applicable to all combustion systems. Now let us see how these ideas apply to the combustion of wood. Wood is a special type of solid fuel which on heating breaks up into a gas-like substance and a solid residue. The former is usually referred to as volatiles and burns with the familiar yellow flame. The latter is referred to as the char and burns with a glow on the fuelbed. These concepts are schematically illustrated in figure 2.
The volatiles burn according to the scheme indicated in the previous paragraph. The burning of char occurs at its surface and is more complex to describe in simple terms.
The way in which the above phenomena take place in a piece of wood could be visualized by considering a tall piece of wood standing vertically and burning steadily at its top (see figure 3).
The lowest portion is the virgin wood and the top port on is the char. On top of the char is the flame where the volatiles burn. What is good about this picture is that the volatiles have to pass through the hot char before reaching the flames and they can be expected to be at a reasonably high temperature. The bad thing is that air comes from the surroundings which are cold. The air should penetrate into the interior of the flame for complete mixing to occur. This rarely happens and the result invariably is poor combustion.
We do not burn wood in this manner anyway and thus will not push this idealized picture any further. As a matter of fact the cookstove provides a worse environment for clean combustion, as we shall presently see. Figure 4 shows schematically a cookstove. Two things make this system less effective for clean combustion. Firstly the volatiles have to pass through the "cold" unburnt wood before coming into the region of the flames. In addition the whole combustion space (unless designed properly) is in contact with the "cold" pan. Thus one of the principal requirements of the fire triangle is rather drastically violated by a conventional cookstove design - it does not provide an environment of sufficiently high temperature for good combustion. It is often assumed that secondary air assists the process of clean combustion.
If the admission ports are designed as a set of small orifices around the combustion chamber, mixing is greatly facilitated; but still the air is cold and unless special arrangements are made to preheat it, the already poor temperature conditions are made worse.
We now discuss a possible method of overcoming some of these drawbacks. We call the method down draft burning. The idea is schematically illustrated in figure 5. In this mode of burning the volatiles as usual are generated in the higher parts of the fuelbed and have to pass through the hot char before coming to the region where it can burn. Thus the temperature conditions are nearly perfect for complete combustion. Moreover since the gas stream has to take a right angled turn at the bottom of the fuel bowl the mixing between the volatiles and unused air is also improved. It has to be noted that it is essential to have a chimney for this mode of burning. In contrast to the conventional mode of burning where a chimney is considered as a device to remove obnoxious combustion products out of the kitchen environment, the chimney is essential to maintain down draft combustion. The question is whether this is the answer to our prayer. We shall present some test results in the next section to demonstrate its effectiveness is achieving clean combustion.
Before presenting results we need to discuss the methodology one needs to use to monitor the quality of combustion. In principle one has to measure a variety of components in the products of combustion. This is far too complex and does not necessarily lead to the promised land of a clean wood burner. Luckily for us Roger van Hoek established by detailed gas chromatographic measurements on a variety of stove designs that carbon monoxide could be used as a marker for clean combustion. In addition most of the experiments were supplemented by monitoring the chimney exhaust for signs of smoke and smell. It turned out whenever the carbon monoxide meter showed high values there was substantial amount of smelly, dense smoke at the chimney exhaust.
A fairly extensive set of measurements of the downdraft stove was carried out by Hasan Khan covering design, fuel and operational conditions. It is not possible to go into all the details in this paper. We will restrict ourselves to results that demonstrate the capability of the design compared to a conventional mode of burning. The stove with conventional mode of burning was the so-called experimental stove. This is a metal stove with double walls, a completely sunk-in pan and a 1 m chimney (see Sielcken and Vermeer, 1985; thus the stove is roughly comparable to the downdraft stove which was completely covered with a commercially available insulating blanket). Annemarie Dirks carried out fresh tests on this store under more or less comparable conditions to those used by Hassan Khan.
Figure 6a presents a global comparison of the performance of the two stoves in a plot of æ co against power output without reference to the output under which the stove was operated. Two facts emerge from the plot. Only three experiments on the downdraft combustion show CO content of more than 0.1~G while only three experiments on the conventional burning show CO content of less than ().1~o. The second point is more clearly seen from figure fib. Depending on the conditions of operation there is nearly a factor of 30 differences between the lowest and the highest value of CO registered in the experiments on downdraft combustion. The implication is that the actual amou nt of pollutants emitted by any stove is not just a function of power output but of many other factors characterizing the design and operation of a stove. What can be said at the present state of development of the two stoves is that the downdraft mode of burning seems to be more tolerant to abuse than the conventionally fired stove.
We illustrate the point made above in figure 7 which shows CO/CO2 percentages as a function of power output for three chimney heights. An important fact here is that if one were to require a stove to operate at 41<W it is much better to use a 43cm chimney than a 100cm one. Secondly, in general there is a steep increase in CO/CO2 as the power output is reduced for all chimney heights. A more disturbing factor in these results is that low power output and high CO/CO2 ratios are accompanied by high excess air factors. This suggests that bad combustion is not caused by low availability of oxygen but due to low temperatures caused by the dilution by too much air.
The results in Fig. 7 were obtained in the absence of a pan. As pointed out earlier, presence of a pan has a cooling effect with adverse consequences on combustion. While no extensive measurements on efficiency were carried out results of a few experiments are shown in table 1. In the design there were two possible positions for the pan. By removing the insulation between the fuel bowl and the chimney, a pan can be accommodated. It is also possible to place a pan on top of the chimney since one can get power outputs of the order of SkW with a chimney height of no more than 50cm. The table shows the drastic increase in the CO content with a pan. It seems that with two pans it is possible to realize total efficiencies of nearly 40%. It is also seen that there is a reduction in power output with the introduction of pan(s). We however repeat that the stove at its present state of development is not ready for use in a kitchen.
The next point is to compare the actual levels of pollution in a kitchen due to the two modes of combustion. Note that this is important since we do not work with chimneys that are tall enough to lead the offensive product of combustion out of the kitchen environment. A detailed estimate of this is dependent on the stove characteristics, the fuel used, its mode of operation and the architecture of the kitchen with emphasis on ventilation levels. This has not so far been dealt with in the literature and much more research is required.
In the absence of detailed work we provide a rough and ready estimate on the basis of a so-called well stirred reactor model. In plain English this means that the emissions of the chimney are instantaneously mixed with the air in the kitchen. To take into account the reality of in homogeneous mixing we introduce a correction factor of 0.4.
Figure 8 shows the results of the calculation for a kitchen with a volume of 25m3 and an air exchange rate of 4/fur (this means that the air in the kitchen is completely renewed once in 1> minutes.
The line marked MAC in the figure needs a bit of explanation It stands for the maximum allowable concentration and it represents the time one can stay at a given level of CO% in the environment (there is more to it than we are able to state here; see for example Sulilatu, 1987).
The symbol CON stands for the conventional stove. It is seen that this curve crosses the MAC curve after about 50 minutes of stove operation. This does not mean that the environment becomes dangerous after this period. One could stay in the environment for a few more minutes without any serious consequences. The curve DD represents the downdraft mode of burning and the cube never crosses the MAC line and it can be concluded to be safe for indefinite periods of operation. A note of caution is essential. The figure is no more than indicative of the merits of the concept of downdraft principle of combustion and should not be treated as received wisdom.
The last point we should like to address in this short summary on clean combustion is to present attempts to incorporate downdraft combustion in an existing laboratory model of a woodfired bakery oven (a mathematical description of the operation of this oven is available in Schutte et al 1989 and experimental results on it will be available shortly). Soot deposition in the flow passages of such systems has several undesirable consequences. Flow passages will be obstructed and heat transfer will be reduced. As a result maintenance requirements will be increased and if this is not done fire hazards and drastic life-time reductions can be expected. Thus poor combustion has considerable direct adverse financial consequences. Figure 9 shows the type of modifications that will be required to change a conventionally fired oven to a downdraft one.
Comparative tests on the two bring out three merits of downdraft burning. Firstly after 30 hours of operation (roughly 4 to 5 days of one-shift operation) with conventional mode of firing, thick layers of soot were deposited and cleaning was essential to maintain acceptable performance levels. On the contrary with downdraft burning for the same period, the passages were clean as a whistle and we expect periods between maintenance activities could be at a minimum 4 times as long. Secondly with downdraft burning the powers at which the oven could be fired with the same grate area and chimney height were at least a factor of two higher resulting in a preheating time reduction from 160 to 100 minutes. Finally, because of reduction in the preheating time, even at a higher rate of burning the fuel consumption was reduced by about Two.
In the foregoing we have indicated a set of preliminary results that hold sufficient promise for the application of the downdraft burning principle to a diversity of devices. Starting with the cookstove, figure 10 shows a design that can be put together with mud, bricks etc. The question of placing of pans and the effect on the performance will be studied by Hasan Khan in the coming year in the laboratories of the Bangladesh Institute of Fuel Research and Development as well as in kitchens. As far as the burning qualities of this design in our laboratory suggest, it is a sound candidate for further development.
One of the greatest drawbacks of the design sketched in figure 5 is that it needs to be charged with small amounts of wood at frequent intervals which will not be really appreciated by a user. In order to handle this, a design of the type shown in figure 11 is proposed. The design incorporates features for both primary and secondary air admissions. Preliminary tests suggest that the fuel bowl can be filled with about a kilogram of wood and the burning period will be about one hour.
The next point concerns a couple of tests with briquettes carried out by Moerman with the design of figure 5. The performance in terms of carbon monoxide emission compares favourabl with that of wood. What is more these fuels seem to tolerate much larger charges than will be feasible with wood.
A point that requires much more attention is the fact that the design of figure 5 does not tolerate turn-down ratios much larger than 2. For long cooking seasons with considerable periods of simmering, this is of paramount importance and will be one of the principal research tasks in the coming future.
The most gratifying feature of the downdraft mode of burning is that it can be incorporated into many existing small scale industrial furnaces with fairly simple modifications and achieve significant economic benefits.