Cover Image
close this book Boiling Point No. 28- August 1992
View the document Biomass Combustion, Chimneys & Hoods
View the document WOOD FUEL :
View the document Chimneys & Hoods for Smoke Removal
View the document Biomass Combustion & the Environment
View the document Charcoal & the Environment - Pros & Cons
View the document Smoke Measurement
View the document Stove Emission Monitoring
View the document Successful Mud Brick Chimneys
View the document Alternative Approach to Wood Combustion
View the document Triple Cone Stove Burning Ricehulls & Woodsmoke
View the document "Energy Assistance Revisited - A Discussion Paper"
View the document Clays for Stoves
View the document ITDG & The Maendeleo Review
View the document NEWS
View the document HEDON
View the document PUBLICATIONS
View the document Research & Development News
View the document Letters To The Editor

Stove Emission Monitoring

by G F. Nangale, Department of Engineering, University of Reading. Whiteknights, Reading , RG6 2A Y. UK.


In a project to investigate methods of smoke measurement, three pollutants were measured, i.e., Carbon monoxide (CO), Total suspended particulates (TSP) and Hydrocarbon acidity (HCs). These are known to be health damaging and occur at high levels in biomass smoke and inside houses burning biomass in unvented appliances. CO is generally classified as an asphyxiant because it has a strong affinity to haemoglobin, which carries oxygen to body tissues. TSP is the best indicator of chronic toxicity. Acidity levels due to hydrocarbon emission from combustion processes are known to cause severe eye disorders.

There are two methods commonly used to monitor emissions from unvented biomass-burning cookstoves. The first is the hood method where the stove is placed under a hood into which all the flue gases are drawn mechanically. This method has been used in studies of unvented gas-cookstoves and kerosene space heaters. By directly measuring the air flow in the hood, collecting total particulates in the sampling gas-bottle and direct measuring of carbon monoxide by a flue gas analyzer, the dilution of outside air can be determined.

The second approach, the chamber method, (see also article above by P Young), requires no duct work and air flow calibrations. In principle, it can be done in any chamber, or even in a remote village house, where the ventilation conditions are relatively constant over the period of measurements. The stove is simply put through a cooking cycle in a room and the pollutant concentrations are monitored within the same room.

The choice of the hood method was based on the unavailability of an appropriate chamber (hut) and the conducive laboratory setting for the hood method within the Department of Engineering Workshop of the University of Reading.

Table 1- Stove Design Parameters


Weight (kg)

Wall thickness (cm)

Combustion volume rate to pot(litres)

Height from bottom (cm)






Three Stone






A ten-litre capacity aluminium pan with a flat base (weighing 1.45kg) was used.


The approach was designed in such a way that the emission monitoring system could be operated simultaneously with the determination of thermal performance. In this way, trade-offs between thermal and emission performance could be investigated. Using fuelwood, pollutant emission performance from a Zimbabwe improved stove was compared to that from a simulated traditional three stone fire.


Fuel characteristics: Wood, "Fraxinus excelsior" (Ash), was obtained locally. This was split into four quadrants (6-10 cm diameter and 8-12 cm in length) and air dried where used. Moisture content on dry basis of the wood was found to be 10.46% and using the bomb calorimeter method, the gross calorific value was found to be 18.01 MJ/kg.

Description of the stoves and cooking pan: A Zimbabwe improved metal stove and a traditional three stone stove simulated by three concrete blocks were used in the study. The stove design parameters are summarized in Table 1.



Placement of hood and pollutant monitors:

The hood (Fig 1) was constructed in such a way that it could be connected to piping already installed in the laboratory connected to an existing extractor fan. The air velocity in the hood was maintained at 0.2 m/s and adjusted by a butterfly damper fitted in the hood duct. The hood was fabricated using metal sheet welded or brazed together. The heights at which the TSP monitor was placed inside the hood were 0.45 m and 0.71 m above the centroid of the Zimbabwe stove and the three-stone stove combustion zones, respectively.

Particulates were collected on a 25 mm diameter 0.8 um millipore filter with a low-flow personal air sampling pump (model AFC 123, Cassella London). Oversized particulates were collected in the grit pot. The flow rate of the pump was adjusted to about 2 litres/min. and calibrated before use.

Carbon monoxide concentrations were measured with a flue gas analyzer (Automation Ltd model 402 solid state NDIR). The analyzer was calibrated before each experiment.

The sampling pump was fitted inside the hood duct so that a sample of the flue gas emitted from the combustion process was bubbled in a bottle partially filled with water, maintained at a temperature below 4ÂșC withice. The flow rate of the sampling pump was adjusted and calibrated at 25 litres/min. In this way, most organic acids from the flue gas dissolved in the water in the gas bottle. Solubility of most organic acids increases with decrease in temperature.

Experimental sequence:

The method proposed here attempts to minimise variations in emission characteristics by using a single charge of fuel and minimal tending of fire so that a single test could be used to estimate both thermal and environmental performance. An international standard method called the "water boiling test" was used for the evaluation of the thermal performance of the stoves.

Thermocouples to monitor temperatures were placed under the pan, in the water, on the lid and in the room space (for ambient). The thermocouples and a flue gas analyzer were connected to the data logger, which in turn was connected to the computer. Fig 2 shows the schematic diagram of the experimental set-up.

Both types of pollutant monitors were fumed on when the fire was satisfactorily stable. The water boiling test was terminated and the TSP pump was stopped when the water temperature started dropping steadily. The pan was reweighed to determine the mass of water evaporated. The remaining pieces of fuel were weighed to determine the mass of fuel used during the experiment.


Fig 2 - Schematic Diagram of the Experimental Set-Up

Results and Discussion

The Zimbabwe improved cookstove was compared to the traditional simulated three stone stove using the same wood and the same cooking vessel, that is, the aluminium pan. The mean burn rates did not differ greatly from one another. The emission factors and efficiencies for the Zimbabwe stove were, however, higher then those of the three-stone stove.

A plot of mean emission factors and acidity levels against mean efficiencies (Fig 3) showed that there is a large difference in emission factors and acidity levels and only a small difference in thermal efficiency between the IWO stoves. When emission values and efficiencies of the individual stoves were compared within experimental sets, no trend was observed. The large variability of emissions from individual stoves shows that replicating combustion conditions is very difficult.

The increase in mean efficiency factors and acidity levels with the Zimbabwe stove compared to the three stone stove shows that the small improvements in thermal efficiency come from improvements in heat transfer efficiency with the simultaneous deterioration in combustion efficiency. This was probably caused by the air flow resistance in the small annular gap between the cooking vessel and stove's shield. Bussmann (Ref 1) demonstrated that the ratio of carbon monoxide to carbon dioxide increased with decreasing gap widths and that the ratio increased sharply when the gap became less than 4 mm. The shield-pan gap in the experiment was 2.5 mm.


It was observed that the mean emission values associated with the Zimbabwe stoves were very high compared to emission values from the three-stone stove. However, the difference of thermal efficiencies observed from the two stoves was very small. The low mean efficiency (24.2%) from the Zimbabwe improved stove was probably due to the small pan-shield gap, which tends to resist air flow across the stove.

Some modifications are required on the experimental set-up to facilitate monitoring of total solids emitted from cookstoves. The solids deposited in the pipe work could be minimised by replacing the copper sampling pipe with a shortened glass piping. In this case deposited solids could easily be collected by washing the glass sampling pipe in a bowl containing distilled water before filtration and evaporation processes.

The thermal efficiency of the Zimbabwe stove could be improved by employing a smaller cooking vessel of about 25.9 cm such that the shield-pan gap is increased from 2.5 mm to 8 mm. Bussmann reported that a shield-pan gap of 8 mm can provide efficiencies in the range of 30 to 40% for pan sizes of 20 to 28 cm diameter with moderate combustion quality.

The hood method used cannot be carried out in the field in developing country conditions due to its set-up and supporting equipment. In this case the chamber method is recommended. It might be expected, however, that stove performance on both parameters, i.e., fuel and smoke, will be both poorer and more variable under field conditions.

Fig3 - A Plot Mean Emission Factors and Acidity Against Mean


Acknowledgement This work is part of my MSc research project at the University of Reading. I wish to express my sincere gratitude to Dr Anne Wheldon and Dr David Fulford for their great help and consultance.


Bussmann, P J.T. (1988). Wood Staves: Theory and Application in Developing Countries. Eindhoven.


Ed Note: It is a common feature in stove work that Percentage Heat Utilized is often increased at the expense of lower fuel combustion efficiency resulting in high emissions.