|Fuel Saving Cookstoves (GTZ, 1984)|
It is the intent of this section to suggest how the field worker can integrate testing with the design process. Testing is; important because it permits comparison of the effects of various design features on technical objectives such as reduced fuel consumption or shortened cooking time. The results of testing indicate to what extent design modifications improve (or fail to improve), stove performance.
While providing quantitative data, testing also helps to:
- Develop an understanding of how stoves and fires work. This is especially important to people for whom cooking with fire is not a lifelong experience.
- Stimulate design ideas. Observations of the characteristics of air flow, heat loss, etc. in testing may suggest improvements in design.
Testing methods must take into account three factors that affect cookstove performance:
1. Behavior of the cook: The way the cooking fire is built and
tended strongly influences stove performance.
2. The variety of cooking tasks: Frying, boiling, and roasting involve different heat requirements. Meals may be cooked either simultaneously or in sequence.
3. Variations in fuel: Fuelwood may be of various species and moisture contents, shapes and sizes. Alternative fuels such as dung and agricultural wastes are sometimes used seasonally, with fluctuating availability.
Testing should represent local conditions. The simultaneous influence of these three variables can be studied in trials with users. These consist of cooking observations, to determine how people use stoves (remember that people don't actually "use stoves", they cook), and fuel use measurements, to compare fuel consumption of traditional cooking arrangements with that of improved stoves.
For specific technical analysis of stove performance tests of isolated variables are more easily controlled as they are conducted independently of the users. They are of two kinds: in standard meal tests, measured amounts of food are cooked while fuel consumption is measured; cooking simulation tests use simulated cooking conditions to study more precisely the effects that stove design and operation have on performance. Both of these tests are used in the development process to answer technical questions.
So appropriate testing starts with cooking observations as a basis of design and to establish performance standards for isolated variable tests. Isolated variable tests are then used for feedback on design performance. When improved stoves are sufficiently developed for trial introduction, performance of stoves in normal use is monitored to further refine designs, and so on, as illustrated in Figure 6-1.
Trials in users' kitchens
Consider how difficult you yourself might find it to precisely
describe something as routine (but complicated) as daily cooking habits. You can
talk to people, and ask a lot of questions, but it would be best to obtain this
information by actually watching people cook at home. Try to do this in as many
situations as possible, so as to fairly represent the local range of variation
in family size, economic status, etc. You will need to find the answers to such
- What are typical meals? - type of food and quantity.
- How does the cook control the fire?
- techniques of heat regulation, cooking sequence .
- How long do various cooking processes take?
- How often is the stove use; in a day?
- Does this change seasonally?
- How much fuel is consumed during each stove use?
- What other uses do fires have?
- special foods, social uses, heating.
- What kinds of pots are used?
- size, shape, material.
- What kind of fuel is used? - size, shape, species of wood - green or dry wood?
Get specific information, useful for comparison with improved stoves. Measure pot sizes, weigh fuel, and record the time involved in various cooking processes.
These variables of operation affect not only how well a stove performs, but also how different stoves or traditional cooking fires compare in performance when used under differing cooking conditions. For example: The amount of fuel an enclosed stove saves as against an open fire depends on the size of fire that is customarily built. A large open fire under a single pot loses large amounts of heat by radiation and convection, particularly in windy conditions. A lesser proportion of heat is lost from a smaller fire carefully maintained under the cookpot; a smaller open fire is more efficient (Fig. 6-2)
On the other hand, in an enclosed stove, much of the heat that would be lost from a large open fire is retained; efficiency is greater with a large fire (fig. 6:3).
In an actual test comparing an open fire and a simple enclosed stove (the Louga stove), the proportion of heat transferred to the cookpot (efficiency) was measured for different fire sizes. As can be seen in the graph in figure 6-4, the Louga stove would be only a small improvement over an open fire where traditional cookfires are built conservatively, but a great improvement where fires are large. This test indicates that varying factors such as the customary fire size affect the relative advantage of a given stove. Thus observations of cooking behavior are necessary to later design representative tests.
Watching people cook may provide insights to how a stove might affect traditional cooking techniques. For example:
New stoves may radically change the customary methods of building fires. A "three rock" open fire is easily controlled by adding or removing firewood from three directions, and the progress of the fire is visible to the cook. The techniques for control of a fire in an enclosed stove are quite different. The fire is controlled through the use of dampers, and the fire can be seen and tended only through the entrance door.
By observing these differences from the cook's point of view, design can be modified to incorporate some of the advantages of the open fire, for instance, by adding a second firebox door to the enclosed stove, or making the fire visible by raising the stove.
An understanding of cooking behavior may also suggest how design might be modified to deliberately change traditional practices. For example, even if people traditionally build larger fires than seem necessary (perhaps simply for convenience), they may accept an introduced stove with a small firebox to limit the size of the fire.
Fuel use measurements
Fuel use measurements made on stoves in use by local people are an integral part of follow-up work; they are the real measure of success for improved stove designs.
Tests should be administered by workers who have a good rapport with the families chosen for the tests, and fluency in the local language. They should also understand the local culture - how people measure time, the size of standard wood measurements, times of year when other fuels are used, etc. Equivalent tests should be made both before and after the introduction of an improved stove, over a period of several weeks of normal use. Compare changes in fuel consumption for the same family, using the same kind of firewood.
Concurrently with fuel use measurements, observations of stoves in use might be made, to assess what changes may have occurred in daily cooking routines, and how on-going design might be made more responsive to local conditions.
Simple tests should be designed for users to monitor their own fuel savings. These might be based on how long it takes to use a given quantity of fuel, or how much fuel is consumed in a given time. These tests could then be facilitated by a local representative, so that users could trade experience in a formal way.
Tests of isolated variables
Cooking efficiency: Analyzing variables
Tests of isolated variables measure the specific effects on stove performance of any one variable of stove design or operating conditions. Because they eliminate the varying circumstances of weather, cooking habits, etc., isolated variable tests yield more precise quantitative information than field trials.
Efficiency is a term used to express the proportion of the heat potentially available in cooking fuel that is actually captured by the contents of a cooking vessel. Efficiency decreases with heat losses due to incomplete combustion, incomplete heat transfer to the pot, and heat losses from the cooking pots (see Chapter 5).
Isolated variable tests can be used to assess where heat loss can be minimized by watching the effect on efficiency when one variable at a time is changed.
A simple example: heat loss was measured for a cast iron pot containing 1 liter of water at 100 °C, both with and without a lid. The graph in Figure 6-7 illustrates quantitatively the extent of heat loss from the uncovered pot. Clearly, the use of a lid greatly improves efficiency. Further tests might look at pot design as a variable.
In a second experiment, a smaller covered pot (having less exposed surface area), also containing 1 lifer of water at 100 °C, lost heat much more slowly than the pot in the first experiment, as seen in Figure 6-8.
These tests could be used in demonstrations with local people to show the influence of pot design on efficiency and the importance of covering the cookpot.
To assure that differences in stove performance are a function of only the one variable in question, other variables should be standardized within each test. The following is a list of some variables that have major effects on stove performance. Also listed are some variations that might be studied in subsequent tests.
- start with typical local pots:
- different types locally used
- metal vs. terra cotta
- improved types
- lids vs. no lids
Fuel type - wood used in comparative tests should be of consistent
species, size, and moisture content:
- green vs. seasoned wood
- various sizes
- other fuels used seasonally
Ambient conditions - constant air temperature and windless
- inside vs. outside
- simulated wind from an electric fan
Combustion air control - use the same chimney, same damper
- experiment with air flow pattern (e.g. secondary air inlet)
- dampers vs. no dampers
- change draft by using damper control or varying chimney size
Fuel feed rate
- should represent typical pattern:
- examine how changes in fuel feed rate affect heat transfer; how they affect heat distribution on multi-pot stoves
Cooking time - should represent typical pattern:
- study fuel feed rate vs. cooking time vs. fuel consumption
- vary the relative positions of pots on multi-pot stoves to find the best use of heat
Cooking sequence - start with local pattern:
- different designs
- compare to traditional cooking arrangements
- modify baffles
- vary internal geometry, e.g. firebox depth and shape
Stove - use one stove design in study of non-design
- test different typical foods
- could use heat input to water to simulate cooking
Food - different foods have different heat and time requirements;
- on multi-pot stoves, study the effect on performance of uncovered pot holes
Test operator - comparative tests should be conducted by the same
- observe how other people control tests and influence stove performance
Standard meal tests
Standard meal tests determine how well stoves operate under standardized local conditions; whether they actually save fuel when cooking, and if they are well adapted to local foods.
In a comparative test, if identical foods are cooked in the same way, performance can be assessed simply by measuring differences in fuel consumption.
Use what you know about local cooking practices to determine stove performance objectives. You might start with a cooking timeline (Fig. 6-9).
Using this, cook a few meals in traditional fashion, assuming the perspective of a local cook. When you have been 'trained' in local cooking techniques' measure the fuel used in preparing a typical regional meal. In comparison of improved designs, you need only prepare the same meal and measure fuel consumption.
If water is substituted for food in the cookpots, heat input to the water can be used as a measure of stove performance. The primary advantage of these "boiling water" tests lies in the uniform thermal properties of water.
The transfer of heat energy to water is easily quantified by measuring both its rise in temperature, and heat loss by evaporation:
Temperature increase (°C) x weight of water (kg) + weight of water evaporated (kg) x latent heat constant* ($40 Kcal/kg) = heat input (Kcal)
A figure for theoretical efficiency could then be calculated if the heat input to the water is divided by the chemical energy content of the fuel: efficiency (%) = 100 x heat input (Kcal)/(weight of fuel (kg) x energy content (Kcal/kg))
However, the efficiency of a stove in heating and evaporating water may be unrelated to cooking efficiency for several important reasons:
1. Consider the relationship between heating rate and useful heat when, for instance, cooking a grain in boiling water. As heat is put into the bottom of the pot, it is simultaneously lost from the top and sides. Ideally, the pot would be heated to the boiling point instantaneously, to minimize this loss. This would require a high power input for a short time. Then, the power input would be lowered to equal the heat loss, and the pot would simmer at 100 °C. If the heating rate exceeded the loss from the pot, the water would not increase in temperature, but would merely boil faster, increasing the rate of evaporation. In the case of constant high heat input, the efficiency calculated from a cooking simulation test would not represent actual cooking efficiency because the high heat input rate does not cook the food any faster, but consumes more fuel.
2. If a pot has a lid, some of the water evaporating from it will be recovered as it condenses on the inside of the lid, and thus evaporative loss is not measurable. Therefore in a boiling water test conducted with lids on the pots, efficiency would appear artificially low.
3. The thermal properties of water and many foods are different. Some of the heat utilized in cooking food is consumed in chemical changes and is thus not easily accounted for in a simulation test using boiling water.
For these reasons, boiling water tests do not directly measure cooking efficiency; they should not be used to project fuel savings that will occur under normal use conditions.
Nevertheless, boiling water tests have certain specific uses. Because heat input to water is precisely quantified, simulation tests may be more easily replicated than standard meal tests; the effects of some variables on stove performance might be easier to compare.
For example, heat distribution was studied in the following test. Comparing two multi-pot stoves, the relative amount of heat available to the first cookpot was measured. Using two kilograms of firewood, the fire was regulated so as to heat one lifer of water in the first pot quickly to a boil, and then maintain a simmer for as long as possible. One lifer of water was also placed in each subsequent pot to absorb the potentially useful heat available downstream from the first pot. The temperature profiles of the first pot, and of subsequent pots, for both stoves are graphed in Figures 6-10a and 6-10b. From this it appears that the # 1 stove (the Singer, see Chapter 7) might be better suited to efficient cooking where only one dish is to be prepared. The # 2 stove (the Lorena, see Chapter 7) appears to distribute the heat of. the fire more uniformly to all pots.
Evaporative loss of water was measured in this test only as a means of determining that the heat delivery rate to the first pot was approximately the same in each test.
The relative performance of these stoves might be quite different under normal cooking conditions, e.g. if the thermal mass of the pot contents were different, if lids were on the pots, etc. Thus the results of this kind of test indicate the relative performance under conditions that were simplified for expediency. The conclusions should be regarded as tentative.
Hints on testing
During each test note anything that might influence stove performance, as it is hard to know what might be significant later.
You don't need expensive equipment to conduct these tests. Fuel can be weighed on a produce scale or an improvised balance could be built from a wooden rod and paint cans (see Fig. 6-11). A milk bottle, or any container calibrated against a known standard (such as a canning jar) could be used to measure water and food volumes.
It might be useful to devise your own system of measurement, for some important variables. For example, you could describe a pot of water as boiling at a "light", "medium", or "furious" rate; or as a means of judging combustion efficiency, the opacity of smoke from the chimney of a stove could be rated as a percentage, no smoke at all being 0%, totally opaque smoke, 100%.