Building and Environment 44 (2009) 898–903

Contents lists available at ScienceDirect

Building and Environment journal homepage: www.elsevier.com/locate/buildenv

Indoor air pollution from particulate matter emissions in different households in rural areas of Bangladesh Bilkis A. Begum a, Samir K. Paul b, M. Dildar Hossain b, Swapan K. Biswas a, Philip K. Hopke c, * a

Chemistry Division, Atomic Energy Centre, P.O. Box-164, Dhaka, Bangladesh Department of Physics, Jahangirnagar University, Savar, Bangladesh c Center for Air Resources Engineering and Science, Clarkson University, 8 Clarkson Avenue, Box 5708, Potsdam, NY 13699-5708, USA b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 29 January 2008 Received in revised form 7 May 2008 Accepted 15 June 2008

Indoor air pollution from the combustion of traditional biomass fuels (wood, cow dung, and crop wastes) is a significant public health problem predominantly for poor populations in many developing countries. It is particularly problematic for the women who are normally responsible for food preparation and cooking, and for infants/young children who spend time around their mothers near the cooking area. Airborne particulate matter (PM) samples were collected from cooking and living areas in homes in a rural area of Bangladesh to investigate the impact of fuel use, kitchen configurations, and ventilation on indoor air quality and to apportion the source contributions of the measured trace metals and BC concentrations. Lower PM concentrations were observed when liquefied petroleum gas (LPG) was used for cooking. PM concentrations varied significantly depending on the position of kitchen, fuel use and ventilation rates. From reconstructed mass (RCM) calculations, it was found that the major constituent of the PM was carbonaceous matter. Soil and smoke were identified as components from elemental composition data. It was also found that some kitchen configurations have lower PM concentrations than others even with the use of low-grade biomass fuels. Adoption of these kitchen configurations would be a cost-effective approach in reducing exposures from cooking in these rural areas. Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Particulate matter Indoor air Biomass fuels Reconstructed mass Carbonaceous matter

1. Introduction Air pollution occurs because of anthropogenic activities such as fossil fuel combustion, i.e., natural gas, coal and oil, to power industrial processes and motor vehicles. Combustion puts harmful chemical constituents into the atmosphere such as carbon dioxide, carbon monoxide, nitrogen oxides, sulfur dioxide, and small solid particles and liquid droplets. In developing countries, exposure to indoor air pollution also comes from the combustion of traditional biomass fuels (wood, cow dung, and crop wastes). The byproducts of such combustion produce a significant public health hazard problem predominantly affecting poor rural and urban population in many developing countries [1]. Large numbers of people are exposed on a daily basis to harmful emissions and other health risks from biomass burning that typically takes place in low efficiency, traditional stoves burning low-grade fuel without adequate ventilation. The majority of those individuals exposed to enhanced concentrations of pollutants are women, who are normally responsible for food preparation and cooking, and the infants and/

* Corresponding author. Fax: þ1 315 268 4410. E-mail address: [email protected] (P.K. Hopke). 0360-1323/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.buildenv.2008.06.005

or young children who spend time around their mother near the cooking area. Recent studies [2–4] have shown that indoor air pollution levels from combustion of biofuels can be extremely high, often many times the ambient air quality standards in various developing countries. The measured concentrations depend on where and when the monitoring takes place, given that significant temporal and spatial variation may exist within a house including roomto-room differences. Indoor air pollution from wood burning, animal dung, and other biofuels is a major cause of acute respiratory infections that constitute the most important cause of death for young children in developing countries [5]. In general, airborne fine particles are considered to be of greater health significance [6] than any of the other air pollutants. The measurement of indoor particles is thus essential in order to assess the total particulate exposure of the general population. Recently, several studies have been conducted to assess mass concentrations and chemical characteristics of indoor PM2.5/PM10 and their relationships with the corresponding outdoor PM2.5/PM10 [7–10]. It was found that the correlations between indoor and outdoor PM2.5/PM10 mass concentrations varied over a wide range in different areas with different particles emissions and meteorological characteristics [6]. With regards to chemical compositions, the mass

B.A. Begum et al. / Building and Environment 44 (2009) 898–903

concentrations of the elements, organic/elemental carbon (EC or BC) and ionic substances were determined. It was observed that chemical profiles of indoor and outdoor PM2.5/PM10 particles vary from area-to-area because of different characteristics of indoor/outdoor (I/O) sources as well as infiltration of ambient particles into the interiors of the homes. Indoor air pollution depends on fuel type used, time spent cooking, structural characteristics of houses and household ventilation practices [11,12] (opening of windows and doors). All of these factors are important for households where there is diversity in cooking fuels, stove types, cooking locations and quality of ventilation. It has found that most of the traditionally used biofuel stoves have a thermal efficiency between 10 and 30% [13,14] and emit large quantities of pollutants, exposing the users to high concentrations of toxic and carcinogenic emissions [15]. In general, the PM emission rate increased with increasing temperature [16] and it was found that elemental carbon formation increases with increase of combustion temperature [17]. In Bangladesh, middle-income and upper-income households in urban areas typically use electricity or relatively clean cooking fuels such as liquefied petroleum (LPG) or natural gas. However, lowerincome households in peri-urban and rural areas rely primarily on biomass fuels. These fuels include wood, twigs and leaves, animal dung, and agricultural residues such as straw, rice husks, bagasse (fiber derived from sugar production), jute sticks, etc. Seasonal and economic factors may also dictate the use of different biomass fuels over the course of the year. These biomass fuels are considered as low-grade fuel with higher emissions that are likely to have greater health impacts. The particulate concentrations in an air shed depend on the emission rate from the fuel use and the dispersion of the emitted particles. The extent and duration of smoke in the kitchen, and the amount of smoke leaking from the kitchen to the outdoors or to other living spaces depend on several structural factors: the location of the kitchen, the ventilation rate, and the porous nature of materials used to construct the roof and walls of the kitchen. The duration of the particle exposure may also depend on other characteristics of a house that affect ventilation, such as the number of rooms, the number, size, and placement of doors and windows, and materials used in the construction of walls and roofs. In Bangladesh, houses incorporate many combinations of these characteristics. Bangladesh is a tropical country, has warm weather (20–39  C) and high relative humidity (65–90%) [18] throughout the year. Because of direct transport to indoors through open doors and windows all year round (typically doors and windows are kept open for ventilation), the contributions of ambient particles to indoor particle characteristics would be different from those observed in colder areas. The aim of the present study is to investigate the impact of fuel type and use in kitchen, kitchen location, and ventilation rates on indoor air and apportion the possible source contribution from the measured trace metal concentrations and black carbon (BC) concentrations of indoor particulate matter collected during the investigation.

899

entrance is entirely open), with or without a roof. Some kitchens have four walls with gaps of a few inches between the walls and the roof. Fig. 1 provides descriptions of four typical kitchen configurations. These structural arrangements are expected to have a significant effect on the particle concentrations in the kitchen and adjacent living area. Sampling was conducted during the months of February and March 2006 in a rural area of Savar in the Dhaka district. Five houses were chosen depending on fuel use and kitchen configurations. Table 1 presents the characteristics of the kitchens that were selected as being representative of the variations in fuel use, cooking arrangements, and structural characteristics that affect ventilation. Before choosing these households, a survey was undertaken in which 33 individual households were surveyed to understand the lifestyle, household configurations, fuel use, etc. within the area of investigation. The information was then used to select these five different houses as representative of the households in the locality. Five PM10 samples (four from the kitchen and one from the living room) were collected from each house using MiniVol Portable Air Samplers (AirMetrics, Eugene, OR, USA) [19]. The PM10 samples were collected in the kitchen over a 4 h interval from 7 a.m. to 11 a.m. covering the cooking period and for 8 h from 7 a.m. to 3 p.m. in the living area of each selected household. One collocated PM2.5 sample was also collected for 4 h period along with PM10 from each kitchen in order to estimate the PM10/PM2.5 ratio for the different types of fuels typically used in these kitchens. PM10 samples were also collected outside the house during sampling period to assess the influence of ambient air pollution on indoor air quality. The MiniVols were programmed to sample at 5 l/min through PM10 and PM2.5 particle size separators (impactors) and then through 2 mm pore Teflon filters. The actual flow rate should be 5 l/ min (Lpm) at ambient conditions for proper size fractionation. To ensure a constant flow of 5 Lpm through the size separator at different air temperatures and atmospheric pressures, the sampler flow rates were adjusted for the ambient conditions at the sampling site. The sampler was placed in the room with the nozzle at approximately the height of the breathing zone. 2.1.1. PM mass determination PM mass was measured by the Chemistry Division of the Atomic Energy Centre, Dhaka (AECD). The PM10 and PM2.5 samples were determined by weighing [20] the filters before and after exposure using a microbalance (METTLER Model MT5). The filters were equilibrated for 24 h at constant humidity of 50% and temperature (22  C) in the balance room before every weighing. A Po-210 (alpha emitter) electrostatic charge eliminator (STATICMASTER) was used to eliminate the static charge accumulated on the filters before each weighing. The difference in weights for each filter was calculated and the mass concentrations for each PM2.5 or PM10 sample were then determined. The concentration of black carbon (BC) in the fine fraction of the samples were determined by reflectance measurement using an EEL type Smoke Stain Reflectometer [21].

2. Materials and method 2.1. Sampling site description and analysis In general, Bangladeshi rural and suburban households have a number of cooking arrangements. In many cases, kitchens are not enclosed by four walls and a ceiling. Some poor households do not have separate kitchens. Cooking during the rainy season takes place inside the single room with the adjacent space used by others in the household. During the dry season, cooking would be done outside the structure. In other houses, kitchens have three walls (i.e., the

2.1.2. Elemental analysis A radioisotope-induced energy-dispersive X-ray fluorescence (EDXRF) [22,23] spectrometer was used to analyze the elemental composition of all of the filter samples. Energy-dispersive X-ray Fluorescence (EDXRF) analysis directly measures the energy as well as the intensity of the fluoresced X-rays. EDXRF analysis was based on emissions from a Cd-109 source (emitting Ag–K X-rays). The EDXRF spectra were processed and quantified using the Qualitative X-ray Analysis System (QXAS) and the Analysis of X-ray spectra by Iterative Least square fitting (AXIL) [23]. Calibration was performed

900

B.A. Begum et al. / Building and Environment 44 (2009) 898–903

Household Type K1

Household Type K2

Living room Living room

Kitchen

Kitchen

Household Type K3

Household Type K4

Living room

Living room Stove Sampler Door

Open Kitchen Kitchen

Window Fig. 1. Cooking location and descriptions of four typical kitchen arrangements.

using MicroMatter thin elemental standards (MicroMatter Co., Eastsounds, WA, USA). Because of the limitations of the XRF system, only eight elements: K, Ca, Ti, Cr, Mn, Fe, Cu and Zn could be determined. 3. Data analysis 3.1. Reconstructed mass (RCM) The analysis of PM samples provided elemental concentrations. These values permit the development of fingerprints for a variety of particle sources [24]. It is also useful to combine selected elements to estimate the concentrations of composite variables to represent the most likely form of the measured element such as estimating the amount of ammonium sulfate from the measured sulfur concentration. It is also possible to derive other combinations of elements that represent interesting aerosol components. These combinations are called pseudo-elements such as ‘‘soil’’ [25]. These composite variables and pseudo-elements help provide a better

understanding of possible sources and their contributions to the average ambient aerosol. 3.1.1. Organic carbon The OC was estimated from suitable EC(BC)/TC ratios obtained from literature. Emissions from diesel engines as well as from oiland coal-fired stationary sources exhibit EC(BC)/TC ratios [26,27] in the range of 0.6–0.7. In contrast, for emissions from biomass combustion, EC(BC)/TC ratios are around 0.3 that have been reported in prior studies [28,29]. 3.1.2. Smoke Fine potassium is an indicator for smoke from biomass burning/ brick kiln. Most biomass fuels are characterized by high-alkali contents [30] leading to high concentrations of fine aerosols in the flue gases [31]. In order to obtain a reliable smoke value from the fine potassium, it is necessary to subtract the fine potassium associated with the soil and sea salt components from total K [32]. The XRF analysis does not provide reliable measurements of Na and

Table 1 Description of different types of kitchen investigated. Kitchen group

Fuel type

Kitchen configuration

Ventilation

Construction material

K1 K2 K3A K3B K4

LPG Wood Wood, pollen Wood Wood, cow dung, pollen

Inside house Attached Separate enclosure, 4 ft from house Separate enclosure, 4 ft from house Open area, 12 ft from house

Window, 2 doors 2 Doors 1 Door 1 Door Open

Wall: Wall: Wall: Wall:

brick; roof: tin bamboo; roof: tin bamboo; roof: tin bamboo; roof: tin

B.A. Begum et al. / Building and Environment 44 (2009) 898–903 Table 2 PM2.5/PM10 ratios for biomass fuels.

901

Fuel type

Present study

LPG Wood Wood, pollen Wood, cow dung, pollen

0.40 0.64 0.62 0.59

Literature value [29] 0.51 0.56

Cl is generally below detection limits because of acid displacement by nitric and sulfuric acids. Hence, it is only possible to correct for soil K. Therefore, smoke was calculated as

Smoke ¼ ðKtot  0:6FeÞ

Predicted mass concentration (µg/m3)

1200

Kitchen 1000 800 600 400 200

0

3.1.3. Soil Soil is composed mainly of the oxides of Mg, Al, Si, Ca, Ti and Fe with many other trace elements. The average composition of sandstone and sedimentary rocks and the summation of the 5 major oxides of Al, Si, Ca, Ti and Fe accounts for more than 85% of the total composition. So the equation for soil is

Soil ¼ 2:20Al þ 2:49Si þ 1:63Ca þ 1:94Ti þ 2:42Fe This equation assumes that the two common oxides of iron Fe2O3 and FeO occur in equal proportions. The factor of 2.42 for iron also includes the estimate for K2O in soil through the (K/Fe) ¼ 0.6 ratio for sedimentary soils. The sum of all of the composite variables is expected to provide a reasonable estimate of PM10 and PM2.5 mass for comparison with the measured gravimetric mass collected on the filters. The definition of the reconstructed mass is

RCM ¼ ðNH4 Þ2 SO4 þ Salt þ Soil þ Smoke þ BC þ OC where (NH4)2SO4 and salt are the terms for sulfate and sea salt. Because of the analytical limitations, these terms have been omitted for lack of data needed to estimate their concentrations. 4. Results and discussion 4.1. Comparison of PM10 and PM2.5 for different fuels PM10 particles were the main focus in this study. However, PM2.5 was also monitored in each kitchen, to assess the emissions of PM10 and PM2.5 from different types of fuels. Table 2 provides the PM2.5– PM10 ratios measured for each fuel type. It was found that the ratio varied from 0.40 to 0.64. The low-grade fuels contain high molecular weight hydrocarbons that produce high concentrations of organic carbon during the combustion process. As a result, the amount of PM2.5 increased and the ratio of PM2.5/PM10 has increased. The ratios for most biomass fuel are comparable to the ratios obtained by Dasgupta et al. [33]. The PM2.5/PM10 ratio for LPG is lower than the other fuels suggesting relatively low PM2.5 mass emissions relative to those from biomass fuels. Biomass such as wood, twigs and leaves, and various agricultural byproducts, usually contain a substantial fraction of ash-forming inorganic compounds in addition to the organic byproducts produced by

0

200

400

600

800

1000

1200

1400

µg/m3) Measured mass concentration (µ Fig. 2. Plot of predicted mass by RCM method as a function of measured mass for the kitchen area measurements.

incomplete combustion. These emissions can cause operational problems such as corrosion and blockage [34,35]. Large-scale combustion of wood has been shown to produce fly ash mass size distribution with two distinct modes [36,37]. LPG contains lower molecular weight hydrocarbons, burns cleanly, and does not contain any ash-forming inorganic compounds. 4.2. Characteristics of PM from different fuel types Table 3 presents the PM mass, black carbon (BC), organic carbon (OC) and possible sources of PM (both PM10 and PM2.5 particulates) apportioned by RCM method based on the fuel types used. Table 3 also presents the estimates of major components of the PM10 and PM2.5 for different types of fuel used by the households in the study area. The reconstructed mass for the samples accounted for 65– 128% of the measured mass. The least squares fit to the data gave RCM ¼ 0.91(measured mass) with R2 ¼ 0.87 (Fig. 2). These results demonstrate a satisfactory estimate of the PM components. The lowest RCM value of 65% was observed in the kitchen for PM10 fraction where cow dung and pollen were used as fuel. This low value may be due to an underestimation of organic matter since it was not actually measured. It is also observed that the major portion of the PM (both PM10 and PM2.5) is carbonaceous matter, especially for biomass fuels. 4.3. Indoor air pollution 4.3.1. Kitchen area Table 4 presents the mean PM concentrations in the different types of kitchens during cooking period along with estimated RCM components. In general, it was found that kitchens using biomass fuels (K2–K4) have much higher PM10 concentrations than the kitchen using LPG (K1). In a separate study in Bangladesh, it was also reported that for most common biomass fuel (namely, firewood, dung and jute sticks), PM10 concentrations varied from 60 to

Table 3 PM apportionment based on type of fuel used for cooking. Fuel type

PM type

BC (%)

Smoke (%)

Soil (%)

OC (%)

RCM (%)

LPG

PM10 PM2.5 PM10 PM2.5 PM10 PM2.5

38.9 45.4 32.9 33.3 27.6 27.3

9.95 – 1.47 3.03 10.3 4.44

40.9 82.4 10.8 16.0 7.94 4.20

– – 75.4 76.7 63.6 62.7

89.9 128 121 129 109 98.7

Wood, cow dung, pollen Wood

902

B.A. Begum et al. / Building and Environment 44 (2009) 898–903

Table 4 PM10 apportionment (mg/m3) of samples collected in different kitchen settings. Kitchen group

PM10 concentration

BC

Smoke

Soil

OC

RCM (%)

K1 K2 K3A K3B K4

133  48 1086  86 846  318 785  189 647  286

27.2  8.3 220  68 244  51 207  42 197  71

42.0  23.9 59.7  47.3 63.0  27.0 50.6  12.5 15.0  10.3

68.7  28.6 64.4  30.4 65.2  13.9 63.9  4.2 43.6  18.9

– 506  156 560  116 477  95 453  162

104 78 110 101 110

Table 5 Comparison PM10 concentration (mg/m3) in living area with kitchen. Kitchen group

Living room

Kitchen

Kitchen/living ratio

K1 K2 K3A K3B K4

178 332 104 159 188

197 1177 867 592 990

1.11 3.54 8.36 3.72 5.27

1165 mg/m3 [4]. The present results suggest that indoor air pollution may be substantially higher especially in rural poor households. The study in urban Dhaka also showed insignificant variations between indoor and outdoor air pollution concentrations in homes where piped natural gas was burned for cooking. However, another study [38] suggested that natural gas does produce some indoor PM10. The average PM10 concentration in the kitchen using LPG (K1) was 132 mg/m3 and during that period, the mean ambient PM10 concentration was 63 mg/m3. The observed PM10 in K1 kitchen was higher than the ambient concentration and may be the result of the influence from adjacent kitchens where biomass fuels were used for cooking and other purposes. Source component analysis of the kitchen with LPG emissions showed that more than 40% of PM10 was soil, further suggesting the influence of external sources. The observed larger standard deviation in the PM10 concentration reflects the day-to-day variations in fuel use and cooking patterns. In the case of the K4-type open kitchen, the high standard deviation of the PM mass concentrations is mainly due to the effect of meteorological parameters such as wind direction and wind speed. The use of different types of fuels in different cooking sessions may also have some influence on the observed variation. In general, the PM mass concentrations were relatively lower than other kitchens using biomass fuel because of the dilution effect resulting from the increased ventilation because of the open air cooking location. In addition, the PM concentrations depended on the specific fuels employed during monitoring interval. 4.3.2. Living area Living areas of each household type were monitored to understand the impact of emissions from the kitchens. Table 5 presents PM concentrations of the living areas and a comparison with the concentrations in the kitchen areas. The data are for those days when PM10 sampling was done simultaneously in the kitchen and in the living room. The PM10 concentrations in the living areas are in general lower than the kitchen or the cooking locations. It is observed that the PM10 concentration levels mainly depended on the kitchen position from the living areas and/or ventilation from the kitchen to the living room. Although kitchen group K1 shows

lowest PM10 concentration in the living area as well as in the kitchen, the highest influence was observed to be from kitchen emissions because of the ventilation characteristics of this particular house. The kitchen and the living room of this group is made of bricks and apparently has a lower ventilation rate. In the similar study by Dasgupta et al. [4], the results also suggested that stove location, building materials, and opening doors and windows after cooking significantly affected the household PM10 concentrations. It is also reported that households with inadequate ventilation had relatively higher PM10 concentrations even for detached or open kitchens. Table 6 presents the estimates of apportioned major components for the PM10 in the living areas of different households. The reconstructed mass for the samples accounted for 66–110% of the measured mass. The least squares fit to the data gave RCM ¼ 0.73(measured mass) with R2 ¼ 0.56 that suggests that only 73% of the measured mass could be reconstructed using the present identified sources. The R2 value also shows lower correlation between predicted RCM and measured mass for the living areas. The source apportionment calculation shows major components of the PM10 in living areas are also carbonaceous matter supporting the importance of the kitchen emissions in driving the airborne concentrations. 5. Conclusions In order to determine the indoor air pollution levels in households in rural areas, PM samples were collected from the rural area at Savar, about 30 km north of Dhaka city. The collection of sample was designed to explore the variation in fuel type, kitchen type, and position of kitchen within the home. It is observed that PM concentrations are lower in household using LPG, a cleaner fuel than other biomass fuels. It is also observed that due to position of kitchen and ventilation practice, PM10 concentration in living area are influenced by emissions from kitchens and obviously found higher than the ambient PM10 concentrations. Although fuel choice may affect the indoor air pollution, its role is secondary to the ventilation factors for households. Open or well-ventilated kitchen lowers the PM10 concentration in the cooking and living areas. Carbonaceous material was found to be a major component of PM10 in both kitchen and living room. The present study shows that some kitchen settings can provide relatively clean conditions in terms of PM concentrations even when ‘‘dirty’’ biomass fuels are used. Since these arrangements are already within the means of poor families, adopting such kitchen settings rather than switching to more expensive clean fuels to enjoy significantly cleaner air would be cost-effective in rural areas.

Table 6 Sources apportioned by RCM method from PM10 collected from living room.

Acknowledgement

Kitchen group

Smoke (%)

Soil (%)

BC (%)

OC (%)

RCM (%)

K1 K2 K3A K3B K4

10.0 2.61 18.6 8.81 3.40

30.5 5.27 28.3 28.4 21.2

11.2 17.5 16.1 15.6 13.8

25.8 40.4 36.8 35.9 31.8

78 66 110 89 70

The authors wish to thank Director, Atomic Energy Centre, Dhaka (AECD) and Head, Chemistry Division, AECD for their continued support and cooperation throughout the work. The authors are thankful to the staff members of the Chemistry Division for their continuous help during the course of this work. The work is financially supported partly by the RCA/IAEA, and the Ministry of

B.A. Begum et al. / Building and Environment 44 (2009) 898–903

Science and Information & Communication Technology (MOSICT), Government of the People’s Republic of Bangladesh under a special grant. The authors thankfully acknowledge this assistance. References [1] Smith K, Mehta S, Macusezahl-Feuz M. In: Ezzati M, Rodgers AD, Lopez AD, editors. Comparative quantification of health risks; the global and regional burden of disease due to selected major risk factors. Indoor air pollution from household use of solid fuels, vol 2. Geneva: World Health Organization; 2004. p. 1435–93. [2] Smith KR, Samet JM, Romieu I, Bruce N. Indoor air pollution in developing countries and acute lower respiratory infections in children. Thorax 2000;55:518–32. [3] Parikh J, Balakrishnan K, Laxmi V, Haimani B. Exposures from cooking with biofuels: pollution monitoring and analysis for rural Tamil Nadu, India. Energy 2001;26:949–62. [4] Dasgupta S, Huq M, Khaliquzzaman M, Pandey K, Wheeler D. Indoor air quality for poor families: new evidence from Bangladesh. Indoor Air 2006;16:426–44. [5] Murray C, Lopez A. The global burden of disease. Cambridge MA: Harvard School of Public Health, WHO, World Bank; 1996. [6] Li CS, Lin CH. Carbon profile of residential indoor PM1 and PM2.5 in the subtropical region. Atmospheric Environment 2003;37:881–8. [7] Hidy GM, Lachenmyer C, Chow J, Watson J. Urban outdoor–indoor PM2.5 and personal exposure in the deep south. Part II. Inorganic chemistry. Aerosol Science and Technology 2000;33:357–75. [8] Chao CY, Wong KK. Residential indoor PM2.5 and PM10 in Hong Kong and the elemental composition. Atmospheric Environment 2002;36:265–77. [9] Fischer PH, Hoek G, van Reeuwijk H, Briggs DJ, Lebret E, van Wijnen JH, et al. Traffic-related differences in outdoor and indoor concentrations of particles and volatile organic compounds in Amsterdam. Atmospheric Environment 2000;34:3713–22. [10] Geller MD, Chang M, Sioutas C, Ostro BD, Lipsett MJ. Indoor/outdoor relationship and chemical composition of fine and coarse particles in the southern California deserts. Atmospheric Environment 2002;36:1099–110. [11] Freeman NCG, Tejada SS. Methods for collecting time/activity pattern information related to exposure to combustion products. Chemosphere 2002;49:979–92. [12] Brauer M, Saxena S. Accessible tools for classification of exposure to particles. Chemosphere 2002;49:1151–62. [13] Venkataraman C, Rao GUM. Emission factors of carbon monoxide and size-resolved aerosol from biofuel combustion. Environmental Science & Technology 2001;35:2100–7. [14] Zhang J, Smith KR, Uma R, Ma Y, Kishore VVN, Lata L, et al. Carbon monoxide from cookstoves in developing countries: 1. Emission factors. Chemosphere Global Change Science 1999;1:353–66. [15] Venkataraman C, Negi G, Sardar SB, Rastogi R. Size distributions of polycyclic aromatic hydrocarbons in aerosol emissions from biofuel combustion. Journal of Aerosol Science 2002;33:503–18. [16] Venkataraman C, Joshi P, Sethi V, Kohil S, Ravi MR. Aerosol and carbon monoxide emissions from low-temperature combustion in a sawdust packedbed stove. Aerosol Science and Technology 2004;38:50–61. [17] Farmer R, Edelmann R, Wong E. Modeling soot emissions in combustion systems. In particulate formation during combustion. In: Siegla D, Smith G, editors. General motors research laboratories. New York: Premium Press; 1981. p. 299–317. [18] Azad AK, Kitada T. Characteristics of the air pollution in the city of Dhaka, Bangladesh in winter. Atmospheric Environment 1998;32:1991–2005.

903

[19] Begum BA, Biswas SK. Comparison of PM collection efficiency of Gent and Airmatrics MiniVol portable air sampler. Nuclear Science and Application 2005;I14:79–83. [20] Begum BA, Akhter S, Sarker L, Biswas SK. Gravimetric analysis of air filters and quality assurance in weighing. Nuclear Science and Applications 2006;15:36–41. [21] Biswas SK, Tarafdar SA, Islam A, Khaliquzzaman M, Tervahattu H, Kupiainen K. Impact of unleaded gasoline introduction on the concentration of lead in the air of Dhaka, Bangladesh. Journal of Air and Waste Management Association 2003;53:1355–62. [22] Molnar P, Gustafson P, Johannesson S, Boman J, Barregard L, Sallsten G. Domestic wood burning and PM2.5 trace elements: personal exposure, indoor and outdoor levels. Atmospheric Environment 2005;39:2643–53. [23] Bernasconi G, Tajani A, Kregsamer P. Manual for QXAS/AXIL. Version 3.5. Vienna: IAEA; 2000. [24] Begum BA, Biswas SK, Hopke PK, Cohen DD. Multi-element analysis and characterization of atmospheric particulate pollution in Dhaka. Aerosol and Air Quality Research 2006;6:334–59. [25] Malm WC, Sisler JF, Huffman D, Eldred RA, Cahill TA. Spatial and seasonal trends in particle concentration and optical extinction in the United States. Journal of Geophysical Research 1994;99:1347–70. [26] Cass GR, Boone PM, Macias ES. Emission and air quality relationships for atmospheric carbon particles in Loss Angeles. In: Wolff GT, Klimisch RL, editors. Particulate carbon: atmospheric life cycle. New York: Plenum Press; 1982. [27] Rau JA. Composition and size distribution of residential wood smoke particles. Aerosol Science and Technology 1989;10:181–92. [28] Cachier H, Ducret J, Bremond MP, Youbue V, Lacaux JP, Gaudichet A, et al. In: Levine JS, editor. Biomass burning aerosols in a savanna region of the Ivory coast in global biomass burning. Cambridge, MA: MIT Press; 1991. p. 174–80. [29] Cachier H, Liousse C, Buat-Menard P, Gaudichet A. Particulate content of Savanna fire emissions. Journal of Atmospheric Chemistry 1995;22:123–48. [30] Jenkins BM, Baxter LL, Miles TRJ, Miles TR. Combustion properties of biomass. Fuel Processing Technology 1998;54:17–46. [31] Jimenez S, Ballester J. Formation and emission of submicron particles in pulverized olive residue combustion. Aerosol Science and Technology 2004;38:707–23. [32] Weast RC, Astle MJ. CRC Handbook of chemistry and physics. 63rd ed. Boca Raton, Florida: CRC Press Inc.; 1982. [33] Dasgupta S, Huq M, Khaliquzzaman M, Pandey K, Wheeler D. Indoor air quality for poor families. The World Bank, Development Research Group; 2004. Policy Research Working Paper 3393. [34] Baxter LL, Miles TR, Jenkins BM, Milne T, Dayton D, Bryers RW, et al. The behaviour of inorganic materials in biomass fired power boilers: field and laboratory experiences. Fuel Processing Technology 1998;54:47–78. [35] Miles TR, Baxter LL, Bryers RW, Jenkins BM, Oden LL. Boiler deposites from firing biomass fuels. Biomass Bioenergy 1996;10:25–138. [36] Pagels J, Strand M, Szpila J, Gudmundsson A, Bohgard M, Lillieblad L, et al. Characteristics of aerosol particles formed during grate combustion of moist forest residue. Journal of Aerosol Science 2003;34:1043–59. [37] Strand M, Pagels J, Szpila A, Gudmundsson A, Swietlicki E, Bohgard M, et al. Fly ash penetration through electrostatic precipitator and flue gas condenser in a 6 MW biomass fired boiler. Energy Fuels 2002;16:1499–506. [38] Beer T, Grant T, Morgoan G, Lapszewicz J, Anyon P, Edwards J, et al. Comparison of transport fuels: final report (EV45A/2/F3C)to the Australian greenhouse office on the stage 2 study of life-cycle emissions analysis of alternative fuels for heavy vehicles. CSIRO, University of Melbourne, Southern Cross Institute of Health Research; 2000.