| Boiling Point No. 13 - August 1987 |
Summary of a report by Tata Energy Research Institute (TERI), New Delhi, Feb 87, pp 132 (assisted by Dr. K.R. Smith)
This is a detailed technical report containing a large amount of data in tabular and graphic form and so is not easy reading. However, it is important for all stove workers for being one of the first major studies of the problem of smoke pollution from biomass fuelled, domestic cooking stoves in the Third World.
The report summarizes the objectives of the project as follows
(i) Develop a methodology to estimate emission factors of particulates and carbon monoxide for different fuelstove combinations.
(ii) To study if design changes incorporated in stoves to improve thermal efficiency also end up increasing emission factors as well.
(iii) To develop a performance index for judging stove performance that incorporates measures of both fuel economy and emissions.
(iv) To study if changes in thermal and environmental performance are more stove related or fuel related.
(v) To simulate common construction and operation flaws in the laboratory to study the change in performance of heavy mud and brick stoves.
'vi) To correlate concentrations and exposures reported from laboratory settings to those reported from in-situ measurements and to estimate the adverse health effects.
The justification for the project and study are given in the preface by Dr. R.K. Pachauri, Director of TERI:
"There has lately been a welcome increase in interest in energy and environment problems of rural areas. Some evidence has been accumulating on the large exposures to rural women resulting from cooking done in poor and open combustion conditions. Fears have been expressed that the more efficient chulhas that reduce the pressure on biomass resources and the time and effort spent by many rural families on gathering combustible materials, also lead to increases in a cook's exposure to harmful air pollutants. Given the scale of usage of chulhas, the importance of the subject is undeniable. So, with the objective of studying whether we were aggravating one problem while alleviating another, TERI with the support of the Department of Environment undertook as a first phase a study to measure simultaneously thermal efficiency and emission factors from cookstoves."
"This study while breaking some new methodological ground, confirmed that emission factors for biomass fuels do increase with the efficiency, but in half the stove-fuel combinations tested, the increase in emissions were offset by the increase in efficiency, so that total emissions from these stoves can be expected to be less. In some stoves such compensation is not achieved. If these findings are confirmed by field data, the obvious implication for stove disseminating agencies would be to go slow on promoting these particular stoves until ways are found to reduce emissions from them."
"This study is an important first step and represents TERI's commitment to find ways of using energy that are both efficient and non-polluting."
There are special characteristics of such stoves that make emissions monitoring difficult:
- Because there is no venting, it is not possible to measure pollutant concentration in a flue along with flue air velocity to determine total emissions as is the strategy for most emissions measurement methods;
- Because biomass fuels have substantially different emissions at different times during the burn, it is not appropriate to measure short-term steady-state emissions as is possible with liquid and gaseous fuels.
- Because cooking is not a single continuous process, measurements need to be made over some sort of cooking cycle in a manner analogous to the driving cycle used for monitoring the fuel and emissions characteristics of motor vehicles.
There have been basically two approaches to handle these problems. The first is to place the stove under a hood into which all the flue gases are drawn mechanically. The hood method (sometimes called the 'direct' method) has been used in studies of unvented gas cookstoves and kerosine space heaters in developed countries. It can be used for stoves with flues, as in the proposed standard method for wood heating stoves. Preliminary tests have been done also with the unvented biomass burning stove types of interest here. By either directly measuring the air flow in the hood or by estimating the air flow through mass balance calculations using nitrogen and carbon, the dilution of outside air can be estimated and the pollutant emissions per unit fuel burned determined. A variant of this approach has been used in which ratios of pollutants are monitored without any determination of air flow for combinations, such as gas cookstoves, where steady state combustion conditions are achieved..
The project has initiated research into the relationships between stove thermal efficiency, effective fuel consumption and harmful pollution for different types of stoves and fuels. In fact, the attempt to correlate so many different factors is a little confusing and sometimes unconvincing when the number of samples is very small.
The following extracts indicate the methodology used.
"In addition to fuel economy, many improved stove programmes have claimed reduction in human smoke exposures. Indeed, before the 1970s, smoke was often the primary concern of such efforts in developing countries. But in the years immediately after the first energy crisis, reduction in smoke exposures seemed to become relatively less important in most programmes. Recently, a number of studies have documented that unvented biomass burning stoves emit large amounts of health-damaging air pollutants, often leading to extremely high human exposures. As a result, there has been increased interest in relative smokiness as an important attribute by which to judge stoves."
"We do not intend to propose modifications to the methods recently developed by others to measure fuel economy, but will propose a new method to enable quantitative comparisons of stoves using a measure that incorporates both thermal efficiency and emissions. In addition, as will become clear, the two attributes are not independent and trade-offs exist between efficiency and emissions."
"Unfortunately, there are as yet no standard methods for measuring the 'smokiness' of unvented biomass burning cookstoves in the laboratory or the field, although there have been some initial efforts. There are, however, methods available for unvented gas fired stoves and wood-fired heating stoves with flues, as well as many other less closely related combustion devices with flues, chimneys, or exhaust pipes. This gap may seem surprising in that the open biomass burning stove is still the most common combustion device in the world.
"The second approach, the chamber (or 'indirect') method, requires no ductwork 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 measurement. The stove is simply put through a cooking cycle in a room and the pollutant concentrations are monitored within the same room. Thus, although sharing with the hood method the same need for pollution monitoring equipment, this method requires little other expense. In addition, airflow conditions around the stove can be simulated much more closely with this method."
"Both thermal and emissions performance during biomass combustion are affected by a bewilderingly large constellation of factors. These include the fuel species and conditions, stove designs, pot sizes, shapes, and materials, the manner in which the stove is tended, the types and amount of food being cooked, and of course, the ventilation conditions. Thus, in order to obtain a reliable ranking of stoves, as many factors as possible need to be held constant during the tests."
"A method is developed to measure emissions of air pollutants from unvented cookstoves and to incorporate a measure of these emissions in the existing way of rating cookstoves by thermal efficiency. 5 metal, 2 heavy chimney stoves and 2 kerosine stoves were tested. The fuels used were Acacia Nilotica, mustard stalks, dungoakes and kerosine. The efficiencies were highest for kerosine stoves, followed by stoves burning wood, crop residues and dungcakes in that order. For metal stoves, the decline in thermal performance was not appreciable when fuels were switched. The environmental performance on all stoves, however, changed markedly when fuels were changed.
The emission factors (EFs in g kg) were the highest for kerosine stoves estimated by us to be between 33-151 for CO and between 1.2 - 8.5 for TSP. On wood these were 13-68 for CO and 1.1 - 3.8 for TSP. Emission factors for CO crop residues and dung cakes were intermediate between the values for kerosine and wood and more for crop residues than for dungcakes. Emission factors for TSP were the lowest for wood but were higher and comparable for the other three fuels.
We also found that for any given fuel, the more efficient a stove, higher were its emission factors. Hence, we were led to propose emissions per a well defined standard task (a task was defined as raising temperature of 3.5 litres of water through 600) as a composite measure that combined emissions and fuel economy in one index. In half the cases we studied, the increases in efficiency were greater than the increases in emission factors, so that emissions per task were lower.
On emissions per task basis (for both CO and TSP emissions) kerosine stoves had the lowest values closely followed by two of the metal stoves burning wood, other wood stoves and stoves burning crop residues were next, while stoves burning dungcakes performed the worst. We also used this method to study the performance of heavy stoves with chimneys." "Our approach has been to design the emissions monitoring system such that it can be operated simultaneously with determinations of thermal performance. In this way, trade-offs between thermal and emissions performance can be investigated.
Two pollutants are monitored carbon monoxide (CO) and total suspended particulates (TSP). Both these pollutants are known to be health-damaging and to occur at high levels in biomass smoke inside houses burning biomass in unvented appliances. CO is perhaps most important as a potential short-term (chronic) toxicity. Indeed, TSP, under the name of 'tar', has been used as the single best indicator of the hazard of the most well-studied biomass smoke, that from tobacco. Furthermore, having a physical particulate sample enables later laboratory determination of individual hazardous compounds such as benzo(a)pyrene. Both are indicators of poor combustion for, unlike solid fossil fuels, most biomass fuels contain few toxic contaminants that remain after complete combustion.
Emission Factors vs Power
Increasing power generally implies more complete combustion, therefore higher combustion efficiency, lower emissions, higher thermal losses and therefore lower overall efficiency. (See figure 23 ). In the previous section we showed that EFs increase with increasing efficiency. In Section 4.D.3. we showed that the stoves efficiency reduces with increasing power (though the trend within a stove may be different). Therefore across all stoves, we should expect EFs to reduce with increasing power. These data for different fuels and unvented metal stoves are shown in Figure 11 and 1 2 for E(CO) and E(TSP) respectively. For each fuel there seems to be a value of power below which the emission factors for that stove increase very sharply. From figures 11 and 12, it can be seen that these values seem to be close to 2 KW for wood, 4 KW for dung and about 5 5 KW for crop residues. There are important implications in this for stove designers when they design stoves on which different fuels are likely to be used.
The stoves are all metal, one-pot, chimneyless.
Emissions per task: Et (CO)
Though a higher efficiency stove might have higher emission factors, the total emissions for the performance of a task, however, could be less because a lesser quantity of wood would be required. Emissions per task can then be used as performance index to compare different stove-fuel combinations. We had defined a uniform task as one that would raise the temperature of 3.5 kg of water by 60ºC or one that delivers 210 kcal to the pot. It is desirable that this performance index for a stove be as small as possible as it is directly proportioned to emission factors and inversely proportional to the thermal efficiency. Table 4.D.11 shows the means and ranges for emissions that result from stoves during the performance of this task. To get at emissions per unit heat delivered to the pot/s ( 1 MJ) these numbers should be multiplied 1.14.
Similar tests were carried out with two different, two pot, mud and brick stoves with chimneys, under various different conditions and so give a complete pattern of results. They indicate that venting the chimney outside the room removes 80-88% of the CO and 62-75% of the particulates. Leaving the second pot hole uncovered has little effect on thermal performance but causes large increases in the emissions into the room. Correction of a construction error of 3 cm reduction in the smoke tunnel caused an increase in efficiency from 12% to 22% which indicates how much mud stoves are effected by quite common errors in construction or by damage.
There have been very few studies that have attempted to measure dose and effects in the same group of people. When doses are difficult to measure, exposures are used as surrogate variables. Where even exposures are difficult to measure, concentrations are used as surrogates. In Table 4.D.18 an attempt is made to relate reported concentrations, concentrations of concern, and safe concentrations to. the adverse healths associated with various pollutants. With the exception of CO2 levels, higher concentrations for all pollutants listed in Table 4.D.18 have been reported from indoor settings than are considered safe.
In what follows we present a sample of studies wherein the researchers have associated adverse health effects with smoke from biomass combustion. In some cases no indoor concentrations were measured, some provide conflicting evidence, and others need to be corroborated.
Over 35% of the rural households surveyed by the NCAER complained of smokey kitchens and eye problems related to smoke. A little over 11 % complained of breathing problems as well. Anecdotal evidence of increased incidence of cataract and other eye problems, especially among women comes from Natwara, a village in Madhya Pradesh. The pollutant most responsible for effects on the eye is likely to be formaldyhyde, whose acute effects will manifest at concentrations less than 1.25 mg/m³.
A study conducted by the National Institute of Occupational Health among women in Ahmedabad found a statistically significant association between cooking with smokey fuels and the incidence of cough, dyspoenea and lung abnormalities. The particulate concentrations they measured were ranged from 4.7 to 11.5 mg/m³ for TSP.
Pandey found a higher prevalence of chronic bronchitis among non-smokers who cook than amongst those who do not, implicating domestic stove smoke as a possible etiologic factor. The prevalence increased with increasing number of hours, as reported by the respondent, spent near a stove. This was used as a surrogate variable for exposure. Inspite of the higher smoking rate among men (80%) than amongst women (60%), the prevalence of chronic bronchitis was the same in both groups. An extremely high incidence of acute respiratory infections amongst infants was also reported in this study. It is certainly possible that exposure to biomass smoke is one of the predisposing factors along with malnutrition, measles and parental smoking.
In a study of chronic cor pulmonale in Delhi, Padmavati and Arora found similar incidences amongst men and women even though only 10% of the women and 7 5% of the men admitted to smoking. In countries where cooking fuels are clean, incidence of this condition closely follows smoking patterns. The disease also appeared 10-15 years earlier in women than in men. While all the women were used to cooking, only 7% of the men said they were. Dung, wood and coal, in that order, were principle fuels. The investigators concluded that domestic air pollution was the cause of cor pulmonale in women and for its earlier onset.
In spite of the large quantities of carcinogens in biomass smoke, its link with the cancer has not been established. Evidence for such a link comes from two studies in Kenya and one in Malaysia that have reported higher incidence of nasopharyugeal cancer among people exposed to higher concentrations of woodsmoke. Other studies of nasopharyngeal cancer have failed to find an association with biomass smoke.
While exposure to 1000 ppm of CO for 5 minutes can be lethal, exposure to 100 ppm for 50 minutes is safe for normal adult subjects. However there are special groups on whom the effects of CO poisoning can be more pronounced that in normal subjects. These are patients with heart disease, pregnant women, fetuses and people living at high altitudes. It is particularly to these groups that high CO exposures should be avoided.
Risk estimates for particulates alone have been very difficult to compute since most of our information about the acute ill effects comes from urban settings where particulates are invariably associated with sulphur oxides. Nonetheless, various attempts have been made to associate increases in mortality resulting from chronic TSP exposures. These estimates range from I to 50 deaths per I million KU. For a person cooking 3 hours a day and exposed to an average concentrations of 4 mg/m³ during cooking (and zero at other times) would receive 500 EU annually. The death rate in a population exposed to such levels would go up by 0.5 to 25 per 1000. The crude death rate is about 15 per 1000 in India. Even if one were to believe the estimates of risk at the lower end of the range. It would then be possible that at least about three percent of the women cooking with biomass fuels have their lives shortened because of the particulate exposures they receive.
Benzo(a)pyrene is most frequently used as an indicator for a group of compounds called polycyclic aromatic hydro-carbanos (PAM) which themselves are a sub-set of a group called polycyclic organic matter (POM). Many of these compounds are carcinogenic and there are a number of occupational studies that have linked PAH exposure to lung cancer. Once again the range of risk estimates is very wide - being 2,000 to 50,000 deaths per l million KU. A person cooking for 3 hours a day and exposed to an average concentration of l ug/m³ would receive 0.125 EU annually. A million people thus .exposed could show increased lung cancer deaths between 250 and 6250. The lung cancer death rates per million for Indian women reported in the literature range from 33 in Nata, S.A.; 37 in Bombay to 97 in Singapore. Again the lower estimates seem more believable and would imply that almost all lung cancer deaths amongst women not attributable to tobacco smoking should be attributed to biomass smoke from cooking fires.
The report makes the following general recommendations:
"Any stove that we decide to disseminate on a large scale in a region should, as compared to the ones extant, meet the following criteria
1 Be able to cook the common meals of a place;
2 Not interfere with the structural integrity of dwellings (e.g. hole in the roof, fire hazards);
3 Reduce the exposure to the cook and other residents from harmful emissions from biomass combustion;
4 Require less fuel;
5 Take lesser time to perform a task;
6 Result in lower total emissions per task (so that the ambient air quality is not degraded further);
7 Be easy to construct, install, operate and maintain (marginal decline in performance with lack of care); and finally
8 Require minimum behavioral modifications from the users, (e.g. simultaneous fine tuning of two dampers, frequent chimney cleaning, etc.).
In addition, in come regions multi-fuel or multi-pot capabilities will be desirable. Clearly there is no single solution in terms of cookstove designs incorporating choice of features construction materials and venting arrangements that does well on all these attributes. It would be presumptuous on our parts, therefore, to recommend at this stage which designs of cookstoves should or should not be promoted. Earlier in the report, we had expressed our preference in favour of two attributes - fuel economy and emission factors of CO and TSP. A stove that is most thermally efficient (on the water boiling test) would also save the most fuel. A stove that has minimum emissions per task would result in minimum total emissions. For both these assertions to hold, the cooking efficiency has to be directly proportional to the water boiling efficiency. Ideally, one would like a stove that combines high overall thermal efficiency with low total emissions. Unfortunately, we saw that emission per task do not necessarily decline with increasing efficiency. Our preference, therefore, is not for the most thermally efficient stove, but for the one that results in the least emissions. In other words, stove performance, as defined by us, is strongly dependent on the biomass fuel that will be used and for low quality fuels like agricultural residues, a traditional stove in spite of lower efficiency might perform better than the 'improved' models.
If, because of the problems that have plagued the chimney stove dissemination programs (i.e. fire hazards, structural interference, large 'penalties' for improper construction (dimensional inaccuracies) operation and maintenance, low reductions in fuel consumption (and even in exposures) observed in practice, it is possible that the emphasis could shift to the dissemination of unvented metal or mud or ceramic stoves. In this case, the methodology outlined in this study would assume even larger importance, in that some such evaluation of the stoves will be essential before mounting any large scale dissemination programs."