SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Measurements
  5. 3. Results and Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

[1] The emission factors (EFs) of particles and their carbonaceous fractions, including black carbon (BC) and organic carbon (OC), are measured for residential burning of coal-chunks. Nine types of coals with wide-ranged thermal maturities were used. Particulate emissions from coal-stove are collected on quartz fiber filters through a dilution sampling system and analyzed for BC and OC by thermal-optical method. The EFs of particulate matter, OC, and BC from bituminous coal burning are 16.77, 8.29, and 3.32 g/kg, respectively, on the basis of burned dry and ash-free (daf) coal mass. They were much higher than those of anthracites, which are 0.78, 0.04, and 0.004 g/kg, respectively. Annual emission inventories of particles, OC, and BC from household coal burning are also estimated based on the EFs and coal consumption. The results of the calculations are 917.8, 477.7, and 128.4 gigagrams (Gg) for total particles, OC, and BC emitted in China during the year 2000.

1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Measurements
  5. 3. Results and Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

[2] Primary carbonaceous particles are mainly released from incomplete combustion of carbon-containing fuels, and are usually divided into two fractions, black carbon (BC) and organic carbon (OC). BC has a strong absorptivity of solar radiation, while OC mainly scatters solar radiation. Although there are debates about the net radiative forcing of carbonaceous aerosols [Penner et al., 2003], their importance on global and regional climate change has been well recognized [Andreae, 2001; Chameides and Bergin, 2002; Hansen et al., 2000; Hansen and Sato, 2001; Jacobson, 2001, 2002; Menon et al., 2002; Penner et al., 2001]. BC has been proposed as possibly the second most important global-warming species after CO2 [Hansen et al., 2000; Jacobson, 2001, 2002] and may be included in the post-Kyoto climate treaties [Streets and Aunan, 2005]. Additionally, carbonaceous particles have adverse effects on environment and human health. BC reduces atmospheric visibility, damages the appearance of buildings, provides adsorption sites for toxic pollutants, and can penetrate deeply into the respiratory system because of its fine size [Qiu and Yang, 2000; Hamilton and Mansfield, 1991]. OC contains hundreds of organic compounds, many of which are toxic and carcinogenic.

[3] China is thought to be the most important contributor to the global burden of carbonaceous aerosols according to the emission estimates that have been made [Bond et al., 2004; Cooke et al., 1999; Cooke and Wilson, 1996; Liousse et al., 1996; Penner et al., 1993; Streets et al., 2001, 2003]. But these estimates contain large uncertainties and differ greatly from each other, mainly due to the different emission factors of BC and OC assumed. Unfortunately, up to now, there are few accurate measurements on emission factors of carbonaceous aerosols based on experiments, especially on residential coal combustion [Bond et al., 2002; Chen et al., 2005], which is a major contributor of carbonaceous emissions in China [Bond et al., 2004; Streets et al., 2003].

[4] The primary object of our study is to develop carbonaceous emission factors from domestic coal combustion in China based on simulation experiments. However, the accurate measurements of emission factors have been hampered by the diverse burning conditions of coals adopted in Chinese kitchens, coupled with the wide range of coal maturities used. In our previous work, emission factors of BC and OC had been measured for five honeycomb-briquettes made of coals with different maturities [Chen et al., 2005]. In this study we focus on raw-coal-chunks since they represent a larger part of household coal consumption nowadays in China, and they emit a much greater fraction of carbonaceous particles. Nine coal samples of different maturities are selected due to the fact that not only anthracite but also various bituminous coals are burned for domestic purpose in China, especially in coal-producing remote countryside. This household coal utilization has also caused severe indoor air pollution [e.g., Mumford et al., 1987]. Particulate emissions from residential coal burning in our work are collected and analyzed for BC and OC concentrations to calculate their emission factors; finally the annual emissions of BC and OC are estimated for residential sector of coal consumption in China.

2. Measurements

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Measurements
  5. 3. Results and Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

[5] The nine coal samples were collected from different state-owned coal mines and their characteristics are shown in Table 1. These coals cover a wide range of maturity with volatile matter content on a dry and ash-free basis (Vdaf) varying from 8.1% to 38.4%, and can be categorized into high-volatile bituminous (HVB), medium-volatile bituminous (MVB), low-volatile bituminous (LVB) and semi-anthracite coals (SA) following the American Society for Testing and Material [2004] standard classification of coals. Compared to the honeycomb-briquettes of our previous study, these coals cover a wider range of thermal maturity, and two coals remain the same (XW and XA) in order to investigate the differences between two burning styles [Chen et al., 2005].

Table 1. Characteristics of the Nine Coal Samples Tested in This Study
Coal IDVdaf,a %Ad,b %Cdaf,c %RO,d %RankeProducing Area of Coal
  • a

    Volatile matter on dry and ash-free basis.

  • b

    Ash on dry basis.

  • c

    Carbon of elemental composition on dry and ash-free basis.

  • d

    Mean reflectance of vitrinite in coal.

  • e

    Rank by ASTM standard classification of coal [American Society for Testing and Material, 2004], HVB is for high-volatile bituminous coal, MVB for medium-volatile bituminous coal, LVB for low-volatile bituminous coal, SA for semi-anthracite.

ZG38.424.4172.780.58HVBZhungeer, Inner Mongolia
YL37.348.3581.930.72HVBYuling, Shaanxi Province
XW30.8323.3278.641.12MVBXuanwei, Yunnan Province
DT30.364.2484.530.95MVBDatong, Shanxi Province
CX30.089.6789.341.00MVBCixian, Hebei Province
XA20.7410.2885.011.70LVBXing'an, Henan Province
CZ16.007.6091.401.90LVBChangzi, Shanxi Province
YQ12.1911.0490.471.98SAYangquan, Shanxi Province
AY8.0910.3193.172.47SAAnyang, Henan Province

[6] The coal-stove and sampling system have been described in detail elsewhere [Chen et al., 2004]. As an improvement, an additional fractionating system is integrated to enhance the sampling capacity. Briefly, the stainless steel sampling system is composed of several parts: a hood which gathers the emissions from the coal-stove, two large-size pipes with a long curved pipe which cools down the hot flue gas to ambient temperature, a pump at the end of the system which draws the flue gas through the pipes at a flow rate of about 1 m3/min, and a branched pipe upstream the pump which ducts a portion of exhaust to a sampler. The sampler can collect particles and gaseous organics simultaneously using pre-baked quartz fiber filters (QFF, Whatman) and polyurethane foam plugs (PUF) at a flow rate of 20–60 L/min. Two flow meters are installed to monitor the flow rates of the pump and the sampler respectively, recording every 10 minutes during sampling, to calculate the actual fraction sampled of the total emissions. The sampling systems are mounted in a laboratory with an air-cleaning system to provide particle-free air circulation inside.

[7] The sampling procedure is similar to what is described in our previous study [Chen et al., 2005], as only summarized briefly here. Large chunks of raw coal were broken into pieces of 3–5 cm in diameter at first. Some charcoal (pre-weighed) was burned in the stove until smoking faded out and the temperature of the inner stove was high enough, and then coal chunks were put into the stove for ignition from the bottom and left burning without any disturbance. The sampling procedure started when the coal chunks were put into the stove and lasted for 2–4 h until combustion ended completely. The weights of coal chunks before and after combustion were recorded to obtain the actual weight of coal burned. Duplicate samples were collected for each coal to check the reproducibility. Procedure blanks and charcoal emission samples were also collected to determine background contamination.

[8] The weights of particulate matter collected on QFFs were acquired by weighing the filters before and after sampling under the same conditions of temperature and humidity. A filter punch (1.5 cm2) was cut for organic and elemental carbon (OC/EC) analyses using a thermal-optical carbon analyzer (Sunset Laboratory Inc.). The protocol was similar to UTS-3 [Yu et al., 2002], i.e., the stepwise temperatures are similar to those in NIOSH method 5040 [Birch and Cary, 1996]. However, the residence time of each step is much longer, because it has been shown that a sufficiently long residence time reduced charring [Yu et al., 2002; Chow et al., 2004] and thus improved the accuracy of the EC and OC analysis [Yang and Yu, 2002]. Three to five portions for each QFF sample were analyzed to eliminate the nonuniformity of filter samples. Although the difference between EC and BC had been stressed [Bond et al., 1998], we consider BC here as the same mass as EC due to the current lack of a standardized BC measurement method [Bond and Bergstrom, 2006].

3. Results and Discussion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Measurements
  5. 3. Results and Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

[9] The emission factor of particulate matter (EFPM) for residential combustion of each coal tested on dry and ash-free basis is calculated according to the particle weight collected on the filter and the ratio of sampled to total emissions, corrected for the blank and charcoal emissions. A similar method is used to calculate the emission factors of BC (EFBC) and OC (EFOC). All results are presented in Table 2, including the ratios among PM, OC and BC.

Table 2. Emission Factors and Their Ratios of PM, OC and BC for Residential Raw-Coal Combustion Based on Burned Dry and Ash-Free Coala
Coal IDEFPMEFOCEFBCOMb/PM, %BC/PM, %OC/BC
AverageSDAverageSDAverageSD
  • a

    Emission factors are given in g/kg coal.

  • b

    OM refers to organic matter (1.3 × OC).

ZG4.051.632.661.060.200.1585.194.8113.61
YL23.244.2711.522.935.340.9164.4122.992.16
XW24.923.8212.551.7210.100.5765.4540.521.24
DT26.8510.1611.587.0810.120.7356.0737.691.14
CX37.8119.9717.0110.4812.675.1558.4733.501.34
XA22.960.099.293.036.970.6352.5930.361.33
CZ6.801.253.850.980.480.0873.537.018.07
Geomean of bituminous16.772.528.292.753.320.5564.3219.832.50
YQ0.970.180.0510.0130.0070.0016.790.677.78
AY0.620.110.0300.0050.0020.0006.180.3414.08
Geomean of anthracite0.780.140.0390.0080.0040.0006.480.4810.46

[10] It can be seen in Table 2 that there are notable differences in EFs for coals with different maturity. For example, CX (MVB) has the highest EFs of PM, OC, and BC, which are about 60, 600, and 6000 times higher than those of AY (SA), respectively; great differences also exist among different bituminous coals, for example, EFBC of CX is 65 times of that of ZG. Therefore, it seems that there is no easy way to estimate realistic carbonaceous emissions from the residential sector of coal burning in China, since coals with wide range of maturity are burned in Chinese countryside and their relative consumptions are difficult to acquire.

[11] The relationship between EFs and coal rank, as well as the good correlations among EFPM, EFOC and EFBC which were found previously [Chen et al., 2005], is confirmed by the present data (Figure 1). As can be seen, coals with Vdaf of about 30% (MVB, vitrinite reflectance (RO) close to 1.0%, Table 1) produce the highest EFs. It has been shown that bituminous coals with such ranks can yield the most abundance of tar [Radke et al., 1980]. Under residential burning conditions, coal tar will be released into the flue and form significant particles [Chen et al., 2005; Bond et al., 2002]. Coal tar is composed mainly of organic matter (1.3 × OC). The main part of the PM mass is organic matter except for PM from anthracite (Table 2). Anthracite coals emit much more mineral matter due to the higher burning temperatures resulting in more complete combustion. Other research showed that tar is the precursor of BC [Richter and Howard, 2000]. Therefore, good correlations of EFOC versus EFPM (R2 = 0.98) and EFBC versus EFPM (R2 = 0.94) are expected. The ratios of OC to BC (Table 2) also show significant variability within different coal ranks. These primary ratio values are useful when using the BC tracer method to estimate the contribution of secondary organic aerosol [e.g., Yu et al., 2004].

image

Figure 1. Relationship of emission factors of PM, OC, and BC versus volatile matter (Vdaf) of coals on dry and ash-free basis.

Download figure to PowerPoint

[12] By comparing the two burning styles, both EFPM and EFOC show no obvious differences, while the average EFBC of bituminous coal chunks is about 15 times higher than that of briquettes (3.32 versus 0.22 g/kg). The XA coal shows the largest discrepancy of EFBC with a factor of 109 between the two burning styles (6.97 versus 0.064 g/kg). However, the differences caused by burning styles are just a little higher than the divergence among various bituminous coals (Table 2) [Chen et al., 2005]. It seems that under residential burning conditions, coal maturity has more significant effects on carbonaceous EFs than burning style.

[13] There are some EFPM data measured previously for residential coal burning, and they are comparable with our results [e.g., Bond et al., 2004, and references therein]. These data were compiled and the values calculated as 7.7 ± 6.5 to 12 ± 8 g/kg for bituminous coal combustion [Bond et al., 2004], and 0.5 g/kg for briquette or anthracite [Streets et al., 2001]. But so far, there is only limited data for EFOC and EFBC acquired from experiments simulating the residential sector of coal burning. The latest data of EFBC derived from the calculation by Streets et al. [2001] based on EFPM and heterogenous ratios of BC to PM, which are 0.12 g/kg for briquette and 3.7 g/kg for raw coal, are unexpectedly similar to the mean values of our study (Table 2). However, the average EFs calculated in our study contain some uncertainties because the actual ratio of different coals is not taken into account.

[14] According to the EFs of PM and carbonaceous fractions acquired in our studies and the coal consumption, primary estimates about the emissions of PM, OC and BC from residential coal burning in China can be made (Table 3). The total residential coal consumption in China was about 79 Tg (teragrams) in 2000 [National Bureau of Statistics of China, 2004], and the percentages of anthracite, bituminous and lignite coal in the total production of raw coal were about 17.6%, 78.1% and 4.3%, respectively. Furthermore, 40% of the coals were burned as briquettes while the remainder as consumed in raw-coal-chunks [Chen et al., 2005]. The yearly BC emission we calculated for the domestic sector is 128.4 Gg, several times lower than the estimate of 605.4 Gg by Streets et al. [2001]. The measurement results obtained during Transport and Chemical Evolution over the Pacific (TRACE-P) experiment suggested that the emission estimates of the domestic sector in East Asia are highly uncertain and had been underestimated [Carmichael et al., 2003]. Therefore, in addition to improving the measurement methods for the carbonaceous matter, more work should be carried out in the future to reduce the uncertainty in the estimates of carbonaceous emission from residential coal combustion in China, including the acquisition of detailed information on consumption of coals of different maturities, the fractions of briquettes and raw chunks consumed, and more tests with other fuel/stove combinations and burning conditions.

Table 3. Emission Estimates of PM, OC, and BC From Residential Coal Combustion During 2000 in China
 BituminousAnthraciteTotal
Raw CoalBriquetteRaw CoalBriquette
Consumption, Tg37.0424.698.345.5675.62
Emissions estimate, Gg 
 PM621.13282.856.497.39917.85
 OC307.04170.230.330.09477.69
 BC122.975.330.030.02128.36

Acknowledgments

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Measurements
  5. 3. Results and Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

[15] This work was funded by the Chinese National Natural Science Foundation (grants 40590392 and 40133010) and China Postdoctoral Science Foundation (grant 2005037597).

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Measurements
  5. 3. Results and Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Measurements
  5. 3. Results and Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information
FilenameFormatSizeDescription
grl21897-sup-0001-t01.txtplain text document1KTab-delimited Table 1.
grl21897-sup-0002-t02.txtplain text document1KTab-delimited Table 2.
grl21897-sup-0003-t03.txtplain text document0KTab-delimited Table 3.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.