Particulate and Trace Gas Emissions from Open Burning of Wheat

Environ. Sci. Technol. 2007, 41, 6052-6058

Particulate and Trace Gas Emissions from Open Burning of Wheat Straw and Corn Stover in China XINGHUA LI, SHUXIAO WANG, LEI DUAN, JIMING HAO,* CHAO LI, YAOSHENG CHEN, AND LIU YANG Department of Environmental Science and Engineering, Tsinghua University, Beijing 100084, P.R. China

Field measurements were conducted to determine particulate emissions and trace gas emissions, including CO2, CO, CH4, NMHCs, NOx, NH3, N2O, and SO2, from open burning of wheat straw and maize stover, two major agricultural residues in China. The headfire ignition technique was adopted, and sampling was performed downwind from the agricultural fire. Particulate matter (PM) and gas emission factors were determined using the carbon massbalance method. Particle mass size distributions show a prominent accumulation mode peak at 0.26-0.38 µm. Submicron particles dominate PM emissions. Most measured chemical species measured show a similar size distribution as PM. Chemical composition analysis indicates that PM2.5 is largely composed of carbon, K, and Cl. PM2.5 emission factors of wheat straw and maize stover are 7.6 ( 4.1 g/kg and 11.7 ( 1.0 g/kg, respectively. It also indicates that 12.1-24.2% of N in biomass is released as nitrogen-based trace gases and 11.0-24.9% of fuel S is emitted as SO2.

Introduction Biomass burning is an important source of gaseous and particulate pollutants in the atmosphere and has a significant impact on global atmospheric chemistry and global climate change (1-3). In China, field burning of agriculture residues is a common way to eliminate waste after harvesting and a significant type of biomass burning. Some studies estimated that 17-25.6% of the total agricultural residue production, or 110-157.5 Tg of crop wastes were burned in the field in China every year (4-6). During harvest seasons, open burning of agricultural residues releases a large amount of pollutants to the atmosphere, including particulate matter (PM), CO, hydrocarbons, and others, which causes serious local and regional environmental impact (7, 8). In June 2006, wheat straw field burning in several regions, such as the Hebei Province and Shandong Province, caused serious air pollution in Beijing (9). In an extreme case, smoke emitted from crop waste open burning reduced visibility drastically and led to the temporary closing of an airport and highway. Limited studies on emissions from open burning of agricultural residues have been conducted. Jenkins and Turn developed a wind tunnel to simulate agricultural burning and measured its particulate and gaseous emissions (10, 11). Hays et al. simulated agricultural fire in an enclosure and * Corresponding author phone: +86-10-62782195; fax: +86-1062773650; e-mail: [email protected]. 6052

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 17, 2007

presented the physical and chemical characterization of particulate emissions (12). Zarate et al. performed combustion chamber experiments and a field burning experiment to estimate emission factors from cereal waste burning in Spain (13). Nguyen et al. conducted field experiments and measured CO2, CO, and CH4 emissions from rice straw burning in Vietnam (14). Dhammapala et al. performed wheat and bluegrass stubble burn experiments at the US EPA open burning test chamber and measured PM2.5, CO2, CO, and total hydrocarbon emissions (15). Data on emissions from open burning of crop wastes in China are scarce. Current estimates of emissions from field burning of crop residues in China (4, 6) mainly adopted emission factors given by Andreae and Merlet (2). Andreae and Merlet synthesized a large body of information on emissions from various types of biomass burning and presented a set of emission factors for separate species emitted from different types of biomass burning (2). Because of the lack of available measurement data, the emission factors in their database for agricultural residues are mainly based on extrapolation, which may cause high uncertainty in emission estimates. In addition, air quality is significantly affected because of agricultural field burning during the harvest period. To determine its contribution to ambient air quality, PM source profiles from agricultural fire are needed. Therefore, field measurements were conducted to characterize the emissions of particulate and trace gases from open burning of typical agricultural residues in China. The measurements quantified PM2.5 and trace gas emission factors for emission inventory development as well as provided profiles of PM2.5 for source apportionment. PM mass and chemical size distributions were also investigated in this study, considering their importance in exploring PM formation mechanisms and potential application in the air quality models.

Methods Experimental Procedure. Experiments were carried out at a rural area in Shandong province. Maize stover and wheat straw, which totally contribute about 59% of the total yield of agricultural residues (16, 17), were chosen as representatives of agricultural residues. Maize stover and wheat straw were collected from the local households and spread in windrows in an open area of the field after being reaped manually and by combine harvester, respectively. Two windrows of wheat straw and one windrow of maize stover were arranged for the experiments. The width of each windrow of wheat straw and maize stove was about 0.5 m and 2 m, respectively. Space between of two windrows of wheat straw was about 0.5 m, which is smaller than common windrow spacing (about 2 m), however, the space is sufficient so that the fire from the two windrows did not interfere with each other. The length of the fuel spread on the ground was about 15-20 m. About 40-60 kg of the fuel was burned in each run. The area densities of the tested maize stover and wheat straw were about 1.5 kg/m2 and 3.0 kg/m2, respectively, similar to normal burning in the field. The headfire ignition technique was adopted, simulating the common open burning practice in the field. Two windrows of wheat straw were ignited at the same time, which produced a sufficiently wide plume for sampling. A small weather station was installed to measure temperature, humidity, wind speed, and wind direction. Measurements were conducted under sunlight, light wind (less than 2 m/s), and a relatively stable wind direction. During the sampling periods, ambient temperature and relative humidity were between 25 and 30 10.1021/es0705137 CCC: $37.00

 2007 American Chemical Society Published on Web 08/03/2007

°C and 53-71%, and the sampling was performed downwind. The sampling instruments were set up on an agricultural vehicle creating a sampling height of about 2.5 m. The vehicle was moved to collect smoke emissions efficiently, depending on the wind direction, wind speed, and fire location. The sampling site was 5-10 m away from the fire, far enough for smoke emissions to dilute and cool to ambient temperature before sampling. The sampling time for each test ranged from 35 to 45 min, including the flaming and smoldering stages. For each agricultural residue, three successful test burns were conducted. Prior to the planned crop waste burning, ambient background levels were measured. Before each measurement, the burned agriculture wastes were weighed and a small bundle was sealed and brought back to the lab for proximate and elemental analysis. After the burning, the ash was collected for weighing and some of it was sealed and transported to the lab to determine the carbon content. The experiment site was located far from any point emission sources as well as the highway. The sampling time was staggered around cooking hours of residents in nearby villages. Sampling and Analysis. PM2.5 was collected by two parallel middle-volume samplers at 78 L/min (Dizhi Instrument Factory, China). The first sampler consisted of a 90 mm Teflon-membrane filter which collected samples for mass using a gravimetric method and for elemental analysis. The second sampler consisted of a 90 mm quartz-fiber filter which collected samples for organic and elemental carbon (OC and EC) and ion analysis. Before and after sampling, the Teflonmembrane filters were conditioned for 24 h at about 40% RH and 25 °C in an air-conditioned room and weighed on a microbalance with a resolution of 10 µg. Elements were analyzed by ICP-OES (IRIS Intrepid II XSP, Thermo) and inductively coupled plasma-mass spectrometry (ICP-MS, X serial, Thermo). To convert the samples into solution for element analysis, samples were digested using a microwave accelerated reaction system (Mars, CEM Corporation) employing high-purity reagents (subboiled HNO3, HCl, and HF). Prior to the sampling, the quartz filters were baked at 450 °C for 4 h to reduce blank carbon levels. OC and EC were determined by a Thermal/Optical Carbon Analyzer (DRI, Model 2001) using the IMPROVE protocol (18, 19). Ionic species including chloride, nitrate, sulfate, ammonium, and potassium were quantified by an ion chromatograph (Dionex600) (20). A low-pressure impactor (LPI, Dekati Ltd) was used to determine the size distribution of the chemical species (21). Operating at a flow rate of 9.89 L/min, the LPI had 50% cutpoint aerodynamic diameters of 0.028, 0.056, 0.095, 0.16, 0.26, 0.38, 0.61, 0.95, 1.60, 2.39, 4.00, 6.68, and 9.92 µm from stages 1 to stage 13. Aluminum foils were used as collection substrates. Prior to the collection, all substrates were baked at 450 °C for 4 h. Before and after sample collection, all substrates were also conditioned for 24 h at about 40% RH and 25 °C in an air-conditioned room and weighed on a microbalance with a resolution of 1 µg. After sampling, the substrates were analyzed for OC and EC and ionic species. The analysis methods are same as previously mentioned. Because of the inhomogeneity of the sample deposits, the laser signals reflecting from the substrate punches did not work well enough to distinguish the OC and EC split point. In order to correct OC pyrolysis in the aluminum foils, the pyrolysis ratio of OC in quartz filters was adopted (22). SO2, NOx, and NH3 were sampled and analyzed according to the Chinese standard method for sampling ambient SO2, NOx, and NH3 (23-25). SO2 was collected by an impinger containing the buffer absorption solution of formaldehyde, potassium biphthalate, and Na2-CDTA. After collection, the

solution was transferred to a colorimetric cylinder with sodium sulfonic ammonia, sodium hydroxide, and pararosaniline added sequentially and then analyzed colorimetrically at a wavelength of 570 nm (23). NOx was absorbed using an impinger containing the absorption solution of N-(1-Naphthyl)-ethylenediamine dihydrochloride, sulfanilic acid, and acetic acid. Before entering the impinger, the air passed through a CrO3 tube, where NO was oxidized to NO2. After sampling, the solution was analyzed colorimetrically at a wavelength of 540 nm (24). NH3 was collected by an impinger containing dilute sulfuric acid. After collection, the solution was transferred to a colorimetric cylinder with sodium salicylate-sodium tartrate, sodium nitroferricyanide, and sodium hypochlorite solution added sequentially and then analyzed colorimetrically at a wavelength of 698 nm (25). Other gases were collected using a sampling configuration that included, from upstream to downstream of the sampling train, a filter holder, subminiature sampling pump, and a 20 L Tedlar bag. Teflon tubes were used to connect each section. The flow rate of the pump was adjusted to fill the Teldar bag throughout the entire sampling period. After sampling, aliquots of gas samples were taken out of the 20 L Tedlar bags and placed into 1 L Tedlar bags for various subsequent chemical analyses. These included the determination of gas concentration of CO2, CO, CH4, non-methane hydrocarbons (NMHCs), and N2O. A gas chromatograph with a flame ionization detector (GC-FID) was used to measure CO2, CO, and CH4 (26). In this system, different packed columns were used to separate CO2, CO, and CH4. The separated CO2 and CO were converted to CH4 by H2 at 375 °C in a nickel catalytic converter, which was quantified by the FID. NMHCs were determined by subtracting CH4 from total hydrocarbons (THCs) which were quantified as CH4 concentration using FID with a blank column. Prior to quantifying THCs concentration, the small air peak was subtracted from the THCs peak (27). A gas chromatograph with an electron capture detector (GC-ECD) was employed for analysis of N2O (26). C, H, N content of the tested agricultural waste was analyzed by a CHN elemental analyzer (Model CE-440, Exeter Analytical Inc.), and other elements were quantified by using ICP-OES and ICP-MS. The carbon content of the ash was determined by the CHN elemental analyzer. Determination of Emission Factors. The carbon massbalance method was used to determine the emission factors (28). Zhang et al. gave details about the emission factors that were derived from the carbon mass-balance equation (29). This method has been widely used to evaluate emission factors during field experiments (28, 30-32). Dhammapala et al. compared the differences between emission factors determined by the carbon mass-balance method and the direct method and found them to be in good agreement (15). In this method, all the burned carbon is assumed to be emitted into the atmosphere as carbonaceous particles and carbonaceous gases such as CO2, CO, CH4, and NMHCs, which is described in the following equation,

Cf - Ca ) CCO2 + CCO + CCH4 + CNMHCs + CPM

(1)

where Cf and Ca are the carbon mass in the fuel and ash, respectively. CCO2, CCO, CCH4, CNMHCs, and CPM are the carbon mass in CO2, CO, CH4, NMHCs, and particles, respectively. OC and EC in PM2.5 are regarded in the calculation as carbon mass in the particles. According to results from the experiment conducted in the combustion chamber, the sum of carbon in CO2, CO, THC(CH4 plus NMHCs), and PM2.5 in the plume accounted for 93.1-111.2% of biomass carbon (15), indicating that the uncertainty of this method is acceptable. VOL. 41, NO. 17, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6053

TABLE 1. Tested Agricultural Residue Compositions agricultural residues

wheat straw

maize stover

Proximate (as received, mass %) moisture 9.59 volatile matter 65.54 fixed carbon 18.83 ash 6.04 C H N Na Mg Al Si S K Ca Fe a

8.79 68.93 18.43 3.85

Ultimate Analysisa (dry basis, mass %) 44.8 7.01 0.56

41.09 6.85 1.62

Elemental Analysisb (dry basis, mass %) 0.055 0.13 0.15 3.01 0.17 1.75 0.27 0.09

0.075 0.50 0.22 2.49 0.20 1.85 0.50 0.17

Analyzed by CHN elemental analyzer.

b

Analyzed by ICP-OES.

In this study, the sampling site was close to the fire; therefore, the physicochemical processes of pollutants in the atmosphere are considered to be negligible. The fuel and ash were weighed, and the carbon contents in both of them were analyzed. These data are used in the carbon balance equations to calculate emission factors, which may reduce uncertainties associated with this approach.

Results and Discussion

FIGURE 1. Size and chemical distribution of particles emitted from open burning of (a) wheat straw and (b) maize stove.

Compositions of the Tested Agricultural Residues. Compositions of the tested agricultural residues are shown in Table 1. The carbon content of the fuels is between 40% and 45%. The moisture content of wheat and maize stover is less than 10%. Si and K are present in relatively high concentration in elemental compositions of the fuels. Si accounts for 3.01% and 2.49%, and K accounts for 1.75% and 1.85% in wheat straw and maize stover, respectively. Particle Size and Chemical Distribution. Particle size and chemical distributions from open burning of agricultural residues during the entire sampling period, measured by LPI, are given in Figure 1. Organic matter (OM) is often derived from OC by multiplying a factor to account for the hydrogen, oxygen, and other minor species in the organic matter. This factor is generally estimated as 1.2-1.4 (33, 34). Recently, Turpin et al. concluded that the factors would vary with source, season, and age of the aerosol and recommended some values for specific conditions (35). However, there is few data in the literature on OM/OC ratios from agricultural residue burning. By applying 1.3 as the OM/OC ratio to the filter sampling data, the reconstructed mass from measured chemical species accounts for 88% and 92% of the gravimetric mass for wheat straw and maize stover, respectively. Well “closed” material balances were observed. No further discussion on the determination of OM/OC ratio was included in this paper, and OM concentrations presented in Figure 1 are estimated by multiplying the OC concentrations by the factor of 1.3. Particle mass size distributions show a single mode with peaks at 0.26-0.38 µm. Hays et al. also observed a unimodal size distribution in the accumulation mode for the rice straw burning and the mode peaks between 0.32 and 0.56 µm (12). OM contributes significantly to the individual size bins, especially to the fine particulate matter, while inorganic ions dominate the coarse particulate matter. In order to display the size distribution of individual species more clearly, Figure 2 shows the distribution of EC,

NH4+, K+, Cl-, NO3-, and SO42+ emitted from open burning of wheat straw. Except for NO3-, the species show size distributions similar to mass and OC size distributions, with a prominent mode peak between 0.26 and 0.38 µm. The results of PM mass and chemical size distribution reveal that PM emitted from agricultural fires is mainly from incomplete combustion. Inorganic species, such as K+, Cl-, and NH4+, which are produced by volatilization, subsequent nucleation and condensation, and further growth by coagulation, also contribute much to PM. NO3- displays a multimodal size distribution. The possible explanations for this result may be its volatile nature or errors in the collection, sampling treatment, or analysis. Figure 3 gives the cumulative percentage of PM10 mass size distribution from open burning. From Figure 3, we can conclude that submicron particles (less than 1 µm) and fine particle (less than 2.5 µm) contribute to over 93% of the wheat straw and 98% of the maize stover mass of PM10 emitted from open burning agricultural residues, implying a significant threat to human health. Chemical Compositions of PM2.5 Emissions. Chemical analyses of PM2.5 emissions from the burning of two typical crop wastes are listed in Table 2. The results shown do not include ambient background levels which have been subtracted out. The carbonaceous composition is dominant in the PM2.5 mass, accounting for about 46% for wheat straw and 37% for maize stover. A high ratio of OC/EC was observed; the ratio for wheat straw was 5, as compared to 11.2 for maize residue. For water soluble species, Cl-, the principal ion, accounts for 13.8% for wheat straw and 23.0% for maize stover; K+, often used as a marker for biomass burning (36), contributed 9.8% for wheat straw and 8.5% for maize stover; NH4+ is another important ion in our study, with a fraction of 3.7% for wheat straw and 10.0% for maize stover. For the ICP-OES analytic results, potassium is the major element comprising 7.2% for wheat straw and 6.9% for maize stover,

6054

9


FIGURE 2. Size distributions of individual species emitted from open burning of wheat straw.

FIGURE 3. Cumulative percent of PM10 mass size distribution from open burning. while all other elements are below 1%. Other trace elements analyzed by ICP-MS were less than 0.01%. It can be inferred that other than the carbonaceous composition, K and Clalso contribute significantly to PM2.5 emission. Similar data were observed in other research (10, 12). Most of the results listed in Table 2 indicate high uncertainties, compared with the data derived from the laboratory measurements (10, 12). Andreae et al. presented that a high Kexcess/soot ratio is a tracer of biomass burning, with observed values between 0.7-0.9 in aerosols mainly from agricultural burning and brush fire (36). Turn et al. reported the value from herbaceous fuels was 0.95 (10). A high value (2.83) was also found in

wheat straw burning in Hays’s data (12). Our studies show the values for wheat straw and maize stover are 0.94 and 2.29, respectively. Duan et al. investigated the effect of biomass burning on ambient air quality in Beijing and concluded that the particularly high K+/OC value (∼0.2) could be representative of on-field wheat open fire (8). In this study, the K+/OC value (0.26) from wheat straw burning is consistent with the value mentioned above. Cl- and K+ are an important anion and cation in PM from open burning of agricultural wastes, respectively. In our study, their ratios for wheat straw and maize stover are 1.4 and 2.7, respectively. Zarate’s research shows their ratio is about 0.8 for cereal waste (13). In Turn’s data, the ratios range from 0.33 to 2.0 for herbaceous fuels (11). From Hays’s study, the ratios are 1.1 for wheat straw and 2.8 for rice straw, respectively (12). It can be concluded that Cl-/K+ ratios vary considerably, possibly because of varying content of K and Cl in the individual fuels. Emission Factors of PM and Gas Pollution. Emission factors of PM2.5 and gaseous pollutants are shown in Table 3 and Table 4, respectively. The average PM2.5 emission factor from multiple tests of wheat straw and maize stover are 7.6 ( 4.1 g/kg and 11.7 ( 1.0 g/kg of biomass (dry basis), respectively. In Jenkins’s studies, PM2.5 emission factors of herbaceous fuels range from 3.2 to 7.4 g/kg (11). PM2.5 emission factors given by Hays are 4.71 g/kg for wheat straw and 12.95 g/kg for rice straw, respectively (12). Zarate et al. estimate the total particulate matter emission factors to be 13 ( 7 g/kg for cereal waste (13). Dhammapala et al. indicate VOL. 41, NO. 17, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6055

TABLE 2. Chemical Compositions of PM2.5 Emissions from Open Burning (wt % of PM2.5 Mass)

TABLE 3. Emission Factors of PM2.5 Emissions from Open Burninga

chemical species

wheat straw (n ) 3)

maize stover (n ) 3)

analytical method

chemical species

wheat straw (n ) 3)


ClNO3SO42NH4+ K+ OC EC Na Mg Al Si S K Ca Fe Sc Ti V Cr Mn Co Ni Cu Zn Ga As Se Sr Zr Mo Ag Cd In Sn Sb Ba Hg Tl Pb

13.8 ( 14.6a 0.24 ( 0.21 1.54 ( 1.25 3.69 ( 3.33 9.94 ( 11.8 38.5 ( 16.0 7.65 ( 3.97 0.17 ( 0.01 0.03 ( 0.01 0.14 ( 0.15 0.06 ( 0.07 0.62 ( 0.13 7.26 ( 4.28 0.13 ( 0.06 0.02 ( 0.01 0.0002 ( 0.0002 0.0033 ( 0.0023 0.0004 ( 0.0004 0.0041 ( 0.0071 0.0008 ( 0.0002 0.0001 ( 0.0001 0.0046 ( 0.0045 ND ND 0.0002 ( 0.0003 0.0007 ( 0.0002 0.0002 ( 0.0001 0.0014 ( 0.0007 0.0009 ( 0.0001 0.0002 ( 0.0000 0.0000 ( 0.0001 0.0004 ( 0.0002 0.0000 ( 0.0000 ND 0.0003 ( 0.0003 0.0026 ( 0.0032 0.0001 ( 0.0003 0.0000 ( 0.0000 0.0088 ( 0.0093

23.0 ( 7.05 0.60 ( 0.23 1.86 ( 1.08 9.97 ( 2.30 8.51 ( 4.77 33.6 ( 13.8 2.98 ( 0.68 0.17 ( 0.14 0.05 ( 0.01 0.04 ( 0.02 0.05 ( 0.03 0.67 ( 0.34 6.88 ( 3.49 0.06 ( 0.03 0.01 ( 0.01 0.0001 ( 0.0000 0.0012 ( 0.0014 0.0009 ( 0.0007 NDb 0.0011 ( 0.0002 0.0000 ( 0.0000 0.0003 ( 0.0003 ND 0.0003 ( 0.0005 0.0000 ( 0.0001 0.0006 ( 0.0001 0.0005 ( 0.0001 0.0007 ( 0.0001 0.0004 ( 0.0003 0.0001 ( 0.0001 0.0001 ( 0.0003 0.0006 ( 0.0002 ND ND 0.0001 ( 0.0000 0.0002 ( 0.0004 ND 0.0001 ( 0.0000 0.0095 ( 0.0037

IC IC IC IC IC TOR TOR ICP-OES ICP-OES ICP-OES ICP-OES ICP-OES ICP-OES ICP-OES ICP-OES ICP-MS ICP-MS ICP-MS ICP-MS ICP-MS ICP-MS ICP-MS ICP-MS ICP-MS ICP-MS ICP-MS ICP-MS ICP-MS ICP-MS ICP-MS ICP-MS ICP-MS ICP-MS ICP-MS ICP-MS ICP-MS ICP-MS ICP-MS ICP-MS

PM2.5 (g/kg) OC (g/kg) EC (g/kg) Cl- (g/kg) NO3- (g/kg) SO42- (g/kg) NH4+ (g/kg) K+ (g/kg) Al (mg/kg) Ca (mg/kg) Fe (mg/kg) Si (mg/kg) Mg (mg/kg) K (mg/kg) Na (mg/kg) S (mg/kg) Sc (mg/kg) Ti (mg/kg) V (mg/kg) Cr (mg/kg) Mn (mg/kg) Co (mg/kg) Ni (mg/kg) Cu (mg/kg) Zn (mg/kg) Ga (mg/kg) As (mg/kg) Se (mg/kg) Sr (mg/kg) Zr (mg/kg) Mo (mg/kg) Ag (mg/kg) Cd (mg/kg) In (mg/kg) Sn (mg/kg) Sb (mg/kg) Ba (mg/kg) Hg (mg/kg) Tl (mg/kg) Pb (mg/kg)

7.6 ( 4.1 2.7 ( 1.0 0.49 ( 0.12 0.83 ( 0.69 0.01 ( 0.01 0.09 ( 0.05 0.23 ( 0.15 0.58 ( 0.58 8.2 ( 7.0 10 ( 7.2 1.5 ( 0.44 6.1 ( 8.5 2.3 ( 1.3 480 ( 200 13 ( 5.9 45 ( 17 0.016 ( 0.012 0.28 ( 0.27 0.023 ( 0.017 0.22 ( 0.38 0.061 ( 0.026 0.005 ( 0.005 0.32 ( 0.23 ND ND 0.014 ( 0.013 0.046 ( 0.013 0.013 ( 0.005 0.11 ( 0.084 0.068 ( 0.037 0.017 ( 0.009 0.001 ( 0.003 0.027 ( 0.014 0.000 ( 0.000 ND 0.033 ( 0.044 0.14 ( 0.16 0.008 ( 0.013 0.003 ( 0.002 0.63 ( 0.55

11.7 ( 1.0 3.9 ( 1.7 0.35 ( 0.10 2.7 ( 1.1 0.07 ( 0.03 0.22 ( 0.15 1.2 ( 0.37 1.0 ( 0.65 4.2 ( 2.0 6.9 ( 2.3 1.4 ( 1.4 5.4 ( 4.2 5.4 ( 1.8 810 ( 440 20 ( 18 79 ( 47 0.007 ( 0.003 0.14 ( 0.15 0.10 ( 0.08 ND 0.13 ( 0.029 0.004 ( 0.003 0.034 ( 0.032 ND 0.028 ( 0.048 0.003 ( 0.006 0.069 ( 0.012 0.059 ( 0.014 0.085 ( 0.01 0.040 ( 0.029 0.007 ( 0.007 0.018 ( 0.030 0.070 ( 0.023 ND ND 0.005 ( 0.003 0.026 ( 0.045 ND 0.009 ( 0.002 1.1 ( 0.44

a The results are given as average weigh percent of PM 2.5 mass and standard deviation. b ND ) not detected or less than background level.

that PM2.5 emission factors for wheat stubble burned with low combustion efficiency are 4.7 ( 0.4 g/kg (15). Our data are comparable with these previous studies. Andreae and Merlet (2) concluded PM2.5 emission factors (3.9 g/kg) from agricultural residues are much lower than our study. For carbon emissions, OC emission factors are 2.7-3.9 g/kg and EC emission factors are 0.35-0.49 g/kg. These values fall within the ranges (0.9-8.94 g/kg for OC and 0.17-1.22 g/kg for EC) found in other similar sources (10, 12). Our data are consistent with those from Andreae and Merlet (2). Cl- has the second largest emission factor among all the species. Previous studies revealed that Cl in the fuel had a higher release ratio at low temperatures than other elements (37-39). Turn et al. proposed the notion of fuel element recovery in PM, which means the ratio of mass in the PM to that in the fuel for a selected element. In their studies, Cl has the highest recovery for herbaceous fuels with values between 10% and 35% (10). In this study, average fuel element recovery for PM2.5 is listed in Table 5. Unfortunately, ICP is not sensitive to chlorine, and the chlorine of the fuel is not measured in our study, so the Cl recovery in PM is not given. N as NO3and NH4+, K, S, and Na have high recovery. Most of the data agreed well with Turn’s data (10). Carbon emissions are dominated by CO2 which account for 89.2% and 89.9% of carbon in wheat straw and maize stover, respectively, with CO contributing 5.7% and 5.5%, 6056

9


a Dry fuel mass basis. The results are given as average emission factor and standard deviation. ND ) not detected or less than background level.

TABLE 4. Emission Factors of Gas Pollutions Emissions from Open Burning (g/kg)a chemical species

wheat straw (n ) 3)


CO2 CO CH4 NMHCs N2O NH3 SO2 NOx

1470 ( 46 60 ( 23 3.4 ( 0.85 7.5 ( 1.9 0.07 ( 0.02 0.37 ( 0.14 0.85 ( 0.57 3.3 ( 1.7

1350 ( 16 53 ( 4.0 4.4 ( 0.97 10 ( 5.3 0.14 ( 0.03 0.68 ( 0.52 0.44 ( 0.20 4.3 ( 1.8

a Dry fuel mass basis. The results are given as average emission factor and standard deviation.

CH4 contributing 0.6% and 0.8%, and NMHCs contributing 1.3% and 1.9%. ∆CO/∆CO2 ratios are 6.4 ( 2.7% for wheat straw and 6.2 ( 0.5% for maize stover, respectively. Both values are much less than 10%, indicating that the flaming phase is dominating for simulated agricultural fire (40, 41). Our ∆CO/∆CO2 and ∆CH4/ ∆CO2 ratios (0.6 ( 0.2% for wheat straw and 0.9 ( 0.2% for maize stover) are higher than those from rice straw burning (∆CO/∆CO2 is 13 ( 10% and ∆CH4/

TABLE 5. Average Fuel Element Recovery in PM2.5(%)a C N as NO3- and NH4+ Na Mg Al Si S K Ca Fe a

wheat straw

maize stover

0.71 3.8 2.3 0.18 0.56 0.02 2.6 2.8 0.37 0.16

1.0 5.3 2.7 0.11 0.19 0.02 4.0 4.4 0.14 0.09

Calculated from data in Table 1 and Table 3.

∆CO2 is 0.41%) in the dry season in Southeast Asia (14, 42). CH4 and NMHCs emission factors in the current work are larger than those from the wind tunnel experiment, which may be due to different ignition techniques (headfire versus backfire) (11). Emission factors for CO2, CH4, and NMHCs in our measurements are comparable with the values in Andreae and Merlet’s paper (2). The latter reported high CO emissions with high uncertainty (92 ( 84 g/kg) (2). Nitrogen-based trace gases and SO2 emissions are related to nitrogen and sulfur content of the biomass (31, 43). Considering spatial variation of N and S content, we adopted a better way to quantify their emissions other than emission factors (g/ kg) in Table 4, which is the fraction of fuel N or S emitted as each species. In this study, 24.9% and 11.0% of the S in wheat straw and maize stover was emitted as SO2. 18.0%, 0.7%, and 5.5% of N in wheat straw was released as NOx, N2O, and NH3, respectively; in maize stover, they are released at 8.1%, 0.5% and 3.5%, respectively. NOx is the dominant species among the nitrogen-based trace gas emissions measured. Jenkins et al. reported 7-19% and 1718% of fuel S released as SO2, and 10-12% and 10% of N emitted as NOx for wheat straw and maize stover, respectively (11). There is little data on N2O and NH3 emissions from agricultural residues. Considering the similarity between agricultural fires and savanna fires (44), we compared our results with previous studies on savanna fires. Hurst et al. presented 15 ( 8%, 0.76 ( 0.38%, and 8.6 ( 5.5% of N of savanna vegetation emitted as NO2, N2O, and NH3 in Australian savanna fire, respectively (45). Lacaux et al. showed that 17.2% of N is released as NOx, 3.5% as N2O, and 0.6% as NH3 in tropical African savannas fires (46). Andreae and Merlet (2) gave similar emission factors of N2O, NOx and SO2 as in our study, although they used a much higher NH3 emission factor (1.3 g/kg). Combustion efficiency is defined as the ratio of carbon released as CO2 to total carbon released during the combustion (32). In our study, the values are 91.6 ( 2.9% and 90.6 ( 1.1% for wheat straw and maize stover, respectively. Turn et al. report combustion efficiencies for herbaceous fuels with the exception of sugar cane ranges from 82% to 92% (10). Dhammapala et al. present high combustion efficiencies with an overall average value from 55 burns of 94.5 ( 0.8% (15). The orientation of the residue affects the combustion efficiency and the resulting emissions. Dhammapala et al. report that the average combustion efficiencies of the burns on randomly piled wheat stubble were 1% lower than those of the oriented burns, and PM2.5 emission factors of the former are greater than those of the latter (15). In our study, the orientations of residues were distributed according to the real condition under which they are normally burned in the field. Future Work. Rice straw, another major agricultural residue, is also often burned in the field after harvesting in

southern China; thus, more work is need to investigate its emissions.

Acknowledgments This work was funded by the National Key Basic Research and Development Program of China (Grant No. 2002CB211600), the National Natural Science Foundation of China (Grant No. 20521140077), and Toyota Motor Corporation. We wish to acknowledge the support of Toyota Motor Corporation, especially Mr. Satoru Chatani and Mr. Isawa Hiroyuki. We also thank Mr. Andrew Burnham of Argonne National Laboratory for his great help in revising this paper.

Literature Cited (1) Levine, J. S.; Cofer, W. R.; Cahoon, D. R.; Winstead, E. L. Biomass burning: a driver for global change. Environ. Sci. Technol. 1995, 29 (3), 120A-125A. (2) Andreae, M. O.; Merlet, P. Emission of trace gases and aerosols from biomass burning. Global Biogeochem. Cycles 2001, 15 (4), 955-966. (3) Andreae, M. O. Biomass burning: Its history, use, and distribution and its impact on environmental quality and global climate. In Global Biomass Burning, Atmospheric, Climatic, and Biospheric Implications; Levine, J. S., Ed.; MIT Press: Cambridge, MA, 1991; pp 3-21. (4) Street, D. G.; Yarber, K. F.; Woo, J.-H.; Carmichael, G. R. Biomass burning in Asia: annual and seasonal estimates and atmospheric emissions. Global Biogeochem. Cycles 2003, 17 (4), 1099, doi: 10.1029/ 2003GB002040. (5) Cao, G.; Zhang, X.; Zheng, F. Inventory of black carbon and organic carbon emissions from China. Atmos. Environ. 2006, 40, 6516-6527. (6) Yan, X.; Ohara, T.; Akomoto, H. Bottom-up estimate of biomass burning in mainland China. Atmos. Environ. 2006, 40, 52625273. (7) Guo, H.; Wang, T.; Simpson, I. J.; Blake, D. R.; Yu, X. M.; Kwok, Y. H.; Li, Y. S. Source contributions to ambient VOCs and CO at a rural site in eastern China. Atmos. Environ. 2004, 38, 45514560. (8) Duan, F.; Liu, X.; Yu, T.; Cachier, H. Identification and estimate of biomass burning contribution to the urban aerosol organic carbon concentrations in Beijing. Atmos. Environ. 2004, 38, 1275-1282. (9) http://www.bjepb.gov.cn/bjhb/tabid/68/InfoID/5897/Default.aspx. (10) Turn, S. Q.; Jenkins, B. M.; Chow, J. C.; Pritchett, L. C.; Campbell, D.; Cahill, T.; Whalen, S. A. Elemental characterization of particulate matter emitted from biomass burning: wind tunnel derived source profiles for herbaceous and wood fuels. J. Geophys. Res. 1997, 102 (D3), 3683-3699. (11) Jenkins, B. M.; Turn, S. Q.; Williams, R. B.; Goronea, M. et al. Atmospheric pollution emission factors from open burning of agricultural and forest biomass by wind tunnel simulations, Final report. CARB Project No. A932-126; California Air Resources Board: Sacramento, California, 1996. (12) Hays, M. D.; Fine, P. M.; Geron, C. D.; Kleeman, M. J.; Gullett, B. K. Open burning of agricultural biomass: physical and chemical properties of particle-phase emissions. Atmos. Environ. 2005, 39, 6747-6764. (13) Zarate, I. O.; Ezcurra, A.; Lacaux, J. P.; Dinh, P. V. Emission factor estimates of cereal waste burning in Spain. Atmos. Environ. 2000, 34, 3183-3193. (14) Nguyen, B. C.; Mihalopoulos, N.; Bonsang, B. CH4 and CO emissions from rice straw burning in South East Asia. Environ. Monit. Assess. 1994, 31, 131-137. (15) Dhammapala, R.; Claiborn, C.; Corkill, J.; Gullett, B. particulate emissions from wheat and Kentucky bluegrass stubble burning in eastern Washington and northern Idaho. Atmos. Environ. 2006, 40, 1007-1015. (16) Ministry of Agriculture, P. R.C. China agriculture statistical report 2000; China Agriculture Press: Beijing, China, 2001. (17) Department of Industry and Transport Statistics, National Bureau of Statistics, P.R.C., and Energy Bureau, National Development and Reform Commission, P.R.C. China energy statistical yearbook 2004; China Statistics Press: Beijing, China, 2005. VOL. 41, NO. 17, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6057

(18) Chow, J. C.; Watson, J. G.; Pritchett, L. C.; Pierson, W. R.; Frazier, C. A.; Purcell, R. G. The DRI thermal/optical reflectance carbon analysis system: description, evaluation and application in U.S. air quality studies. Atmos. Environ. 1993, 27A, 1185-1201. (19) Chow, J. C.; Watson, J. G.; Pritchett, L. C.; Crow, D.; Lowenthal, D. H.; Merrifield, T. Comparison of IMPROVE and NIOSH carbon measurements. Aerosol Sci. Technol. 2001, 34, 23-34. (20) Duan, F. K.; He, K. B.; Ma, Y. L.; Yang, F. M.; Yu, X. C.; Cadle, S. H.; Chan, T.; Mulawa, P. A. Concentration and chemical characteristics of PM2.5 in Beijing, China: 2001-2002. Sci. Total Environ. 2006, 355, 264-275. (21) http://www.dekati.com/dlpi5.shtml. (22) Kleeman, M. J.; Schauer, J. J.; Cass, G. R. Size and composition distribution of fine particulate matter emitted from wood burning, meat charbroiling, and cigarettes. Environ. Sci. Technol. 1999, 33, 3516-3523. (23) Standardization Administration of the People’s Republic of China (SAC). Determination of sulfur dioxide-formaldehyde absorbing-pararosaniline spectrophotometry; GB/T15262-94; SAC: Beijing, China, 1994. (24) Standardization Administration of the People’s Republic of China (SAC). Determination of nitrogen oxides-Saltzman method; GB/T 15436-95; SAC: Beijing, China, 1995. (25) Standardization Administration of the People’s Republic of China (SAC). Determination of ammonia-sodium salicylatesodium hypochlorite spectrophotometric method; GB/T 1467993; SAC: Beijing, China, 1993. (26) Wang, Y. H.; Wang, Y. S.; Sun, Y.; Xu, Z. J.; Liu, G. R. An improved gas chromatography for rapid measurement of CO2, CH4 and N2O. J. Environ. Sci. 2006, 18, 162-169. (27) State Environmental Protection Administration of China. Stationary source emission-Determination of nonmethane hydrocarbons-Gas chromatography; HJ/T 38-1999; SEPA: Beijing, China, 2000. (28) Radke, L. F.; Hegg, D. A.; Lyons, J. H.; Brock, C. A.; Hobbs, P. V. Airborne measurements on smokes from biomass burning, In Aerosols and Climate; Hobbs, P. V.; McCormick, M. P., Eds.; A. Deepak Publishing: Hampton, VA, 1988; pp 411-422. (29) Zhang, J.; Smith, K. R.; Ma, Y. et al. Greenhouse gases and other airborne pollutants from household stoves in China: a database for emission factors. Atmos. Environ. 2000, 34, 4537-4549. (30) Hurst, D. F.; Griffith, D. W. T.; Cook, G. D. Trace gas emissions from biomass burning in tropical Australian savannas. J. Geophys. Res. 1994, 99 (D8), 16441-16454. (31) Lee, S.; Baumann, K.; Schauer, J. J.; Sheesley, R. J.; Naeher, L. P.; Meimardi, S.; Blake, D. R.; Edgerton, E. S.; Russell, A. G.; Clements, M. Gaseous and particulate emissions from prescribed burning in Georgia. Environ. Sci. Technol. 2005, 39, 9049-9056. (32) Ward, D. E.; Susott, R. A.; Kauffman, J. B.; et al. Smoke and fire characteristics for Cerrado and deforestation burns in Brazil: BASE-B experiment. J. Geophys. Res. 1992, 97 (D13), 1460114619. (33) White, W. H.; Roberts, P. T. On the nature and origins of visibilityreducing aerosols in the Los Angeles air basin. Atmos. Environ. 1977, 11, 803-812.

6058

9


(34) Countess, R. J.; Wolff, G. T.; Cadle, S. H. The Denver winter aerosol: a comprehensive chemical characterization. J. Air Pollut. Control Assoc. 1980, 30, 1194-1200. (35) Turpin, B. J.; Lim, H. J. Species contributions to PM2.5 mass concentrations: revisiting common assumptions for estimating organic mass. Aerosol Sci. Technol. 2001, 35, 602-610. (36) Andreae, M. O. Soot carbon and excess fine potassium: Longrange transport of combustion-derived aerosols. Science 1983, 220, 1148-1151. (37) Bjorkman, E.; Stromberg, B. Release of chlorine from biomass at pyrolysis and gasification conditions. Energy Fuels 1997, 11, 1026-1032. (38) Jensen, P. A.; Frandsen, F. J.; Dam-Johansen, K.; Sander, B. Experimental investigation of the transformation and release to gas phase of potassium and chlorine during straw pyrolysis. Energy Fuels 2000, 14, 1280-1285. (39) Knudsen, J. N.; Jensen, P. A.; Dam-Johansen, K. Transformation and release to the gas phase of Cl, K, and S during combustion of annual biomass. Energy Fuels 2004, 18, 1385-1399. (40) Bonsang, B.; Boissard, C.; Lecloarec, M. F.; Rudolph, J.; Lacaux, J. P. Methane, carbon monoxide and light non-methane hydrocarbon emissions from African savanna burning during the FOS/DECAFE experiment. J. Atmos. Chem. 1995, 22, 149162. (41) Hurst, D. F.; Griffith, D. W. T.; Cook, G. D. Trace gas emissions from biomass burning in tropical Australian savannas. J. Geophys. Res. 1994, 99 (D8), 16441-16454. (42) Nguyen, B. C.; Mihalopoulos, N.; Putaud, J. P. Rice straw burning in South East Asia as a source of CO and COS to the atmosphere. J. Geophys. Res. 1994, 99 (D8), 16435-16439. (43) Lobert, J. M.; Scharffe, D. H.; Hao, W. M.; Kuhlbusch, T. A.; Seuwen, R.; Warneck, P.; Crutzen, P. J. Experimental evalution of biomass burning emissions: nitrogen and carbon containing compounds. In Global Biomass Burning, Atmospheric, Climatic, and Biospheric Implications; Levine, J. S., Ed.; MIT Press: Cambridge, MA, 1991; pp 289-304. (44) Hurst, D. F.; Griffith, D. W. T.; Cook, G. D. Trace-gas emissions from biomass burning in Australian. In Biomass Burning and Global Change; Levine, J. S., Ed.; MIT Press: Cambridge, MA, 1996; pp 787-792. (45) Lacaux, L. P.; Brustet, J. M.; Delmas, R.; Menaut, J. C.; Abbadie, L.; Bonsang, B.; Cachier, H.; Baudet, J.; Andreae, M. O.; Helas, G. Biomass buring in the tropical savannas of Ivory Coast: an overview of the field experiment fire of savannas (FOS/DECAFE 91). J. Atmos. Chem. 1995, 22, 195-216. (46) Hurst, D. F.; Griffith, D. W. T.; Carras, J. N.; Williams, D. J.; Fraser, P. J. Measurements of trace gases emitted by Australian savanna fires during the 1990 dry season. J. Atmos. Chem. 1994, 18, 33-56.

Received for review March 1, 2007. Revised manuscript received June 22, 2007. Accepted June 28, 2007. ES0705137