Cover Image
close this bookBoiling Point No. 36 - November 1995 (ITDG, 1995, 35 p.)
close this folderR & D News
View the documentCoal briquetting and clays for Zambian stoves
View the documentMalawi Rural Stoves, Socio-economic Research
View the documentImproving the three-stone fire
View the documentComparative tests of solar box cookers
View the documentParabolic Solar Reflector and Heat Storage Cooker
View the documentAn Affordable Parabolic Solar Cooker

Coal briquetting and clays for Zambian stoves

Two projects of the Zambian National Council for Scientific Research (NCSR), Director W C Mushipi, reported by N R Hill.

The twin projects of ceramic stoves for domestic cooking (A) and briquetting of coal (B) resulted from official concern over the rate of deforestation caused by the demand for fuelwood as well as for the production of charcoal. Both projects, which began over five years ago, are housed in a purpose-built pilot plant about 0.5 km from the main NCSR site in Lusaka.

Project A - Coal briquetting

The main raw material for the briquettes is a washed, sub-bituminous coal having a carbon content of 58 per cent and ash content of 24 per cent.

The first stage is carbonization of the coal which reduces the volatiles to 6 per cent - but also reduces the calorific value by 25 per cent. To ensure the briquettes will ignite readily, some bagasse is separately carbonized and mixed in with the coal. The mixture is formed into solid, part-rounded, smooth surfaced briquettes, less than 100mm across (see Figure 1), which are then passed through a de-smoking machine (Figure 2).

The briquettes can burn for 6 to 8 hours, and when 10 are loaded in one of the NCSR ceramic stoves the fire keeps burning for 9 to 10 hours. A disadvantage of the process is that as much as a quarter of the heat value of the coal is lost, and further energy is consumed in running the de-smoking machine. It is understood that the initial carbonizing of the coal enables it to be briquetted without the need for a binder.

No details of costs of the new items of plant or for making the finished products, were obtained.

(Editorial note. Unfortunately NR Hill's summary does not give any information about the quality and quantity of the fumes emitted when the briquettes are made or when burned in the NCSR stove. For domestic situations this is now probably more important than their calorific value or fuel economy.)

Project B - Clays for Zambian stoves

The Geological Survey of Zambia has published a comprehensive document on the clay deposits of Zambia entitled 'Chikankata clay deposit: a candidate material for heat-resistant, energy-saving, cook stoves' by N C H Lubaba, J S Banda, S Mambwe & Y Minamikawa

So far, following six or seven ears of evaluation work, the clay-most suitable for stoves, of the three samples studied in detail, is that from Chikankata, which contains smectite clay mineral, talc and quartz. It is considered that the content of MgO, 14-15 per cent, from the talc as well as the 10 per cent of quartz are important in its performance. Sufficient plasticity is given by the 30 to 40 per cent of the particles which are less than 2 microns.

The Chikankata clay is the most economic as the firing can be done at 850°C, and it then attains a higher density than the other clays. It also has low thermal conductivity and the lowest thermal expansion. Its performance is claimed to be better than that of clay used for the University of Zambia's improved 'mbaula' stove.

Cook stoves made from Chikankata clay lasted more than 12 months and, using charcoal as a fuel, can maintain for one hour a fire temperature 100°C higher than a traditional brazier. The durability of the stoves was influenced by design or geometrical factors, heat-treatment, and the size, shape and toughness of the grog. More durable stove compositions were obtained at lower firing temperatures and by incorporating tough and large grog particles with a high aspect ratio.

In 1993 2,000 NCSR stoves were sold and it is expected about 4,000 will be sold in 1994

NCSR, PO Box 310158, Chelston, Lusaka, Zambia. Tel. +260 1 281081

Cross section of the NCSR, Lusaka, domestic clay stove

Malawi Rural Stoves, Socio-economic Research

Socio-economic constraints and incentives relating to the adoption of improved cooking technology for rural households in Malawi.

by Lewis Mhango & Mike Thomson

The Malawi Ministry of Energy and Mining and the UK-based Natural Resources Institute are currently planning a joint research project based on the above theme.

As in many parts of the world, and despite many potential benefits to householders (such as fuelwood and time saving and positive nutrition and health effects), rural domestic energy programmes in Malawi have met with very limited success. This study will analyse rural attitudes towards improved stoves and cooking techniques so that future technology transfer efforts are better directed and have more chance of success.

Early in 1996, a selection of villages will be chosen with different energy problems, cooking techniques, diets, access to markets, agro-climatic conditions, cultures, traditions etc. Within each chosen village, the research team will analyse how different people perceive energy problems, what they do to try to overcome these problems, and what constraints they face.

The success of improved stove, biogas, and solar technologies in resource-poor rural areas hinges on a proper understanding of the socio-economic forces operating within villages and within households.

Anyone requiring more details or wanting to comment on the study should contact Lewis Mhango at the Ministry of Energy and Mining, Private Bag 309, Liongwe 3, Malawi. Fax: + 265 784 236, or Mike Thomson, Natural Resources Institute, Central Avenue, Chatham Maritime, ME4 4TB, UK Fax: +44 1634 883706.

Improving the three-stone fire

by C Ballard-Tremeer and H H Jawurek.


The open, wood-burning fire built on the ground with a pot supported above it - the 'three-stone fire' - is still an extremely common rural cooking device. The raising of such a fire on to a grate could lead to reduced heat losses to the ground and to increased completeness of combustion. This would favour both increased efficiency and reduced emissions. If these gains were significant the 'improved' open fire (IOF) might well constitute a good, essentially no-cost, stove.

The aims of the present study therefore are:

· To compare the efficiencies and the emissions of an open fire with those of an improved open fire built on a grate

· To place these findings in perspective by comparing them with similar measurements on a commercial low-cost stove.

See Figure 1. The open fire was built on a 50mm thick sand base with the pot supported on a metal tripod 10mm above the base. The side lengths of the triangular frame were 250mm as was the diameter of the floor of the pot.

The improved open fire was of the same dimensions as the open fire, but was fitted with a wire grate of 10mm square pitch positioned 10mm above the ground. The grate area (maximum fuel bed area) was approximately half the pot floor area.

The one-pot chimneyless metal stove was a typical example of current trends in stove design. It had light-gauge steel inner and outer walls, ceramic insulation and an adjustable fuel door to control the inlet air. The firebox diameter was 200mm. The stove was top-fed (the pot had to be removed to add fuel).


The percentage heat utilized (PHU) was determined by carrying out the standard water boiling test (VITA 1985).

Emission measurements

The stoves being tested were placed beneath an extraction hood fitted with a fan, a butterfly damper and an orifice plate flow meter. The extraction rate was set slightly above the minimum required to capture all smoke. The resulting air velocities at the stove were lower than typical air currents in a closed room.

The pollutants measured were particulates (smoke), carbon monoxide (CO) and sulphur dioxide
(SO2). Since the particulates emitted by wood fires are all in the respirable size range it is appropriate to measure total suspended particulates (TSP). TSP was measured by means of a light obscuration meter mounted in the exit duct of the extraction hood.

Figure 1

Fuel and fuelling

The exotic hardwood Eucalyptus grandis was used to simulate the hardwood types preferred in rural areas. Its moisture content was 11(±1) per cent. The open fires were fed radially and semi-continuously, following traditional practice. The stove was fed intermittently, once during the heating up phase and one to three times during the simmering.

Results and discussion

It was found that smoke, CO, and SO2 follow similar emission patterns, see Table 1.

Hydrocarbon emissions were not separately measured in our study. However, the most hazardous group of these, the polyaromatic hydrocarbons (PAHC) are emitted as particulates (Calle & Zeighami 1984) and are thus included in TSP.

In summary: A wood-burning stove or fire having high emissions of CO has high emissions of smoke (which includes PAHC), SO2 and non-methane hydrocarbons. The emission of CO is thus used in the present study as an approximate health hazard index of wood-burning cooking devices.

Figure 2 shows efficiency (averaged over the heating-up and the simmering phases) versus fire power for six tests on each of the cooking devices. Clearly there is a general trend to reduced efficiencies with increasing fire power. This would seem to be attributable largely to 'heat transfer efficiency' effects.

The averages of the efficiencies shown in Figure 2 are given in Table 1. The fire built on a grid with an efficiency of 21 per cent is a substantial improvement on the open fire at 14 per cent (the corresponding reduction in fuel consumption is 32 per cent), and competes successfully with the metal stove at 20 per cent. The metal stove, however, reaches boiling in 16 minutes, as opposed to 22 minutes for the fires, a feature that is likely to be favoured by users.

In presenting results on emissions we prefer to use the total mass of pollutant (CO) produced per test (cooking task). This mass is directly related to human exposure. Emission factors, because of their denominator term (per mass of fuel burned) include the effects of efficiency. They are tabulated for comparison.

Figure 3 shows efficiency on the x axis and total CO emissions on the y axis for the 18 tests under discussion. The figure is not intended as a 'plot' in the functional relationship sense, but as a 'performance map' on which the top left-hand corner is the region of good performance and the bottom right-hand corner the region of poor performance.

Figure 2: Efficiency and fire power

Figure 3. Efficiency and tote/ CO emissions

Table 1: Performance characteristics of the cooking devices

Open fire

Improved open fire

Metal stove

Efficiency, %




Total CO per test, g




CO emission factor, g/g




Total SO2 per test, g




Time to reach boiling, minutes




Summary and conclusions

Efficiencies and emissions of an open fire, an improved open fire built on a grate, and a commercial, low-cost, one-pot, chimneyless metal stove were compared. Preliminary tests together with published data showed that CO emissions can be used as an approximate health hazard index of wood-burning cooking devices.

On the basis of total CO emitted per test (cooking task) the improved open fire was the lowest-emitting device, followed closely but without statistically significant difference by the open fire. The stove was higher-emitting than the fires by a factor of three. With respect to SO2 emissions this factor was 10.

The average efficiency was 21 per cent for the improved open fire which is substantially better than the open fire at 14 per cent (these figures correspond to a 32 per cent reduction in fuel consumption). The mean efficiency of the stove was 20 per cent, which is no improvement on the fire built on a grate.

Clearly the improved open fire is a good overall performer, with efficiencies equal to that of the metal stove and emissions marginally (but not statistically significantly) lower than those of the open fire.

Editorial note

The laboratory tests carried out at Witwatersrand University show results which are in line with our field experience of three stone fires and unimproved metal stoves. We would be interested to see similar, comparative tests using a more popular and efficient metal stove, such as the improved KCJ (or a wood burning version of a similar stove) or the 'UMIME' shielded fire or a modern ceramic stove, such as the Sri Lankan Anagi.

As the authors suggest, fire power (and consequent speed of cooking) along with smoke reductions are often the two most important considerations for stove users. If the research continues with field tests, the authors may find their claims for 32 per cent reduction in fuel consumption is not very meaningful at household level.

Report of Research Project of the School of Mechanical Engineering, University of the Witwatersrand, Johannesburg, Private Bag 3, WITS 2050, South Africa
Tel: +2711 7162558
Fax: +2711 3397997

Correspondence should be addressed: E-mail: hhjaw @

Comparative tests of solar box cookers

by the European Committee for Solar Cooking Research (ECSCR) Almeria, Spain. 1994. Editorial extracts.

Solar cookers on the market are very different in size, type, performance and user friendliness and until now comparative data has not been available. To give full information, tests have to be conducted within specific user contexts. However, before going into such field tests there are some important features of solar cookers which can be tested independently of a specific user context:

· thermal performance - how quickly does a cooker heat up, what temperature does it reach in given conditions?

· user friendliness - is the cooker safe, how easy is it to use, to assemble, to transport?

These and other features were studied in the tests conducted at the Plataforma Solar de Almeria by the ECSCR in Spain in 1994. The tests were performed by the Zentrum fur Sonnenenergie and Wasserstoff-Forschung of Germany, and funded by the German Research Ministry. A total of 13 solar box cookers from eight different countries, (France, Germany, Ghana, India, South
Africa, Spain, Switzerland and USA) were tested.

They were tested around noon during a clear day, for speed of heat-up (heat-up time) to 80°C and to boiling minus 3°C. Where possible, cookers were realigned to the sun every half hour, although they can be used all day without tracking and just one realignment.

Table 1 shows, as an example, the test results for two of the stoves described in this edition of
Boiling Point and Table 2, the results for two others. Results of all the tests on the box and concentrator stoves (25) tested are given in the Summary of Results, June 1994 available from ECSCR.

The European Committee for Solar Cooking Research (ECSCR) was founded in 1991. It has set out to improve the situation in solar cooking before the end of the century.

Its philosophy is to bring together users on one side, and individuals and institutions active in solar cooking research and development on the other side; to put solar cooking efforts on a more objective basis; to tackle the problem by a parallel procedure of development of better cookers and a better understanding of the user situation.

Contact: European Committee for Solar Cooking Research (ECSCR), Office of the Co-ordinator, c/o Synopsis, Route d'Olmet, F-34700 Lodeve., France Tel. +33 67440410, Fax: +33 67440601

Table 1

Table 2

Parabolic Solar Reflector and Heat Storage Cooker

by Dr Gerhard L Jobst

Effective solar cooking

Gaining and retaining the sun's power for cooking - these are the two principles that can generally be employed by solar cookers.

While box-type solar cookers try to use both principles at the same time, the use of one or more haybox cookers together with a powerful parabolic type cooker makes a brilliant combination.

The SK 12 parabolic solar reflector

This is a high-capacity solar cooker. The size and design of its cooking pots are especially adapted for haybox cooking. Once in the haybox the food is cooked further by the retained heat. It can be left unattended and keeps warm for later consumption while the cooker can immediately be used for other cooking or sterilizing water.


Manufacture of the SK 12 is possible by local craft workers using simple cutting, bending and punching tools and utilizing easily available materials throughout, such as strip steel, wire and ordinary nuts and bolts.

With its 1.4 metre diameter mirror the SK 12 is a powerful solar collector, capable of bringing three litres of water to the boil within half an hour, as well as allowing frying. So it is a high-heat cooker, and together with cooking pots of 28 cm in diameter, ie pots of up to 10 litres, it is also a high-capacity cooker which can meet the cooking needs of up to 20 people.

Traditional cooking

'Sitting by the fireside, waiting for the meal to be ready after a hard day's work...' Everybody knows that the family's traditional fireside has more significance than just cooking. Unfortunately only a few are aware of the fact that fuel-saving solar cooking can blend perfectly with fireside traditions if only a simple and comfortable system of ancient origin is used.

Back to our roots

Daily cooking frequently includes a long simmering period which is required for many beans, grains, stews and soups. The amount of fuel needed to complete these cooking processes can be greatly reduced by cooking with retained heat in a heat storage cooker such as a hay box.

The haybox

The haybox is simply a well insulated box or basket lined with a reflective material into which a pot of food previously brought to the boil is placed. The insulation greatly slows the loss of conductive heat, convective heat in the surrounding air is trapped inside the box, and the shiny lining reflects the radiant heat back into the pot.

The food is cooked in one to two hours by the heat retained in the insulated box. This works best when the pot fits snugly into the insulation with no air in between.

Be inventive

Such a box or a basket can easily be made of inexpensive, locally available materials. It can be wooden, or a can-in-a-can, or card board, or any combination. Hay, straw, rushes, feathers, sawdust, rags, wool, shredded paper, etc are all good insulating materials.

Comparison of some normal and parabolic/haybox cooking times


Normal cooking times

SK 12 Cooking time + time in the haybox


60 minutes

12 + 55 = 67 minutes


90 minutes

25 + 70 = 95 minutes

Stewed meat

70 minutes

25 + 50 = 75 minutes


45 minutes

20 + 40 = 60 minutes

Principles to be kept in mind are:

the insulation should cover all six sides of the box;
the box should be airtight;
the inner surfaces of the box should be of a heat reflective material such as aluminium foil.

After the food is cooked it keeps hot for many hours.

Haybox cooking is different

There are some adjustments involved in cooking with haybox cookers:

· less water should be used since it is not boiled away;

· less spicing is needed since the aroma is not boiled away;

· cooking must be started earlier to give the food enough time to cook at a lower temperature than on the solar cooker or over the fire;

· the food should boil for several minutes before being placed in the box. This ensures that all the food is at boiling temperature, not just the water,

· haybox cookers work best for large quantities (over four litres) as small amounts of food have less thermal mass and cool faster than larger quantities.

E C Solar Development Group, State Technical College Altotting, Germany. Tel: +49 86718028. Fax +49 8671 84689

An Affordable Parabolic Solar Cooker

by Ari Lampinen and Rajesh Sharma

Most of the solar cookers used in the developing world are box type. The higher performance parabolic cookers are extensively used by households only in Tibet (mirror area about 2m²). Elsewhere, they are most often used in large configurations in schools and other community kitchens where big mirrors can be used to cook tens of litres of food at once (eg ULOG solar hybrid community kitchen with mirror area of about 7m²).

The price tags of the Chinese/Tibetan cookers exceed $100 and the ULOG cookers exceed $1000 and so are not affordable by ordinary people. This paper describes a design that can be fabricated by rural people with their existing skills. Its performance is not so good as the Chinese and ULOG models but is considerably better than box cookers which cost more because of the high cost of glass.

A working model has been made by rural, illiterate Nepalise women belonging to the untouchable caste. Participants for this project and a larger project including other types of solar cookers were chosen, based on their skills, by local women's committees set up with support from a Plants for Life (a Nepalese NGO) integrated rural development programme.

A local, skilled bamboo weaver wove two parabolic baskets about 1 m diameter each, ie 0.8m² using a metal former. A mixture of clay, cow dung, mustard oil cake and paddy husk was pasted on to the parabolic baskets and left to dry for a day. Next day, this was smoothed with fresh mustard oil cake and left for another day. On the third day the surface was polished with sandpaper, and aluminium, reflecting paper was glued on. (See figure 1)

A supporting frame was made with the help of a local carpenter using bamboo to hold the baskets and also to move them vertically from about 30 to 90 degrees. Using a mirror. the two reflecting baskets were focused on an oven plate and the cooking pot.

During cooking, adjustments were made every 12 to 15 minutes. Vertical adjustment was done by using a moving bamboo mechanism.

The material cost of the 60kg cooker was US$3 and labour for the prototype was 18 hours. Cooking time was 35 minutes for 0.5kg of rice and 40 minutes for 0.5 kg of potatoes.

This work was financed from FINNIDA NGO funds.

Ari Lampinen works at the Department of Physics, University of Jyvaskyla, PO Box 35, FIN - 0351, Jyvaskyla, Finland. E-mail Rajesh Sharma works for Plants for Life, PO Box 21, Hetauda 5, Makwanpur, Nepal