Environ. Sci. Technol. 2006, 40, 702-708
Emission of Polycyclic Aromatic Hydrocarbons in China SHANSHAN XU, WENXIN LIU, AND SHU TAO* Laboratory for Earth Surface Processes, College of Environmental Sciences, Peking University, Beijing 100871, China
Emission of 16 polycyclic aromatic hydrocarbons (PAHs) listed as U.S. Environmental Protection Agency (U.S. EPA) priority pollutants from major sources in China were compiled. Geographical distribution and temporal change of the PAH emission, as well as emission profiles, are discussed. It was estimated that the total PAH emission in China was 25 300 tons in 2003. The emission profile featured a relatively higher portion of high molecular weight (HMW) species with carcinogenic potential due to large contributions of domestic coal and coking industry. Among various sources, biomass burning, domestic coal combustion, and coking industry contributed 60%, 20%, and 16% of the total emission, respectively. Total emission, emission density, emission intensity, and emission per capita showed geographical variations. In general, the southeastern provinces were characterized by higher emission density, while those in western and northern China featured higher emission intensity and population-normalized emission. Although energy consumption in China went up continuously during the past two decades, annual emission of PAHs fluctuated depending on the amount of domestic coal consumption, coke production, and the efficiency of energy utilization.
Introduction For years human exposure to polycyclic aromatic hydrocarbons (PAHs) has been a concern as a result of the widespread occurrence of these compounds and their adverse impacts on ecosystem and human health (1, 2). Efforts have been made to reduce PAH emission in developed countries. Pacyna et al. (3) estimated that total benzo[a]pyrene emissions of 33 European countries were reduced from 1300 tons in 1970 to 590 tons in 1995. The emission of seven carcinogenic PAHs in the United States decreased from 2000 tons in 1990 to 1400 tons in 1996 (4, 5). However, evidence based on PAH deposition on the ice sheet of Greenland suggested that the global trend of PAH emission was constant from the beginning of the industrial period up to the early 1990s (6). It is expected that PAH emissions in developing countries including China have been continuously increasing owing to rapid growth in the countries’ economies and in energy consumption. High concentrations of PAHs have been measured in China, which has caused concerns (7). According to the results of an assessment on toxic persistent substances in China and neighboring countries, PAHs are recognized to be of particular importance both regionally and globally (8). PAHs are chiefly byproducts of incomplete combustion of fossil fuels and biomass and pyrosynthesis of organic * Corresponding author phone and fax: 0086-10-62751938, email:
[email protected]. 702
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 3, 2006
materials (9). Population growth and rapid economic growth in recent decades led to a rapid increase in energy consumption in China. As a result, China became the largest coal consumer and is among the largest petroleum users in the world. Annual consumption of coal and petroleum in China reached 1660 million and 287 million tons in 2003, respectively (10, 11). In addition, biomass burning is a traditional practice in rural China, and it was estimated that approximately 537 million tons of biomass was burned either as fuel or in the field in 2003 (11). It is necessary to estimate the emission of a pollutant for a better understanding of its fate and effects and for better practices in risk management and policy making. Emissions of PAHs in several countries or regions have been estimated. For instance, Tsibulsky et al. (12) estimated the annual emissions of six Borneff PAHs in 10 countries of the former USSR during the period 1990-1997. Berdowski et al. (13) reported that the emissions of six Borneff PAHs in 36 European countries in 1990 amounted to 15 800 tons. It was estimated that 1400 tons of the seven PAHs identified as potential biological carcinogens were emitted in the United States in 1996 (5), and the annual emission of 16 PAHs in the Great Lakes region in 1996 was 14 403 tons (14). When possible emission of a large amount of PAHs and potential for their long-range transportation is taken into consideration, PAH emission in China is of both regional and global significance. To this point, however, such an inventory has not been established, leaving a major information gap. For emission estimation of a pollutant, an emission inventory of a particular activity can be compiled by multiplying the activity rate by the corresponding emission factor (EF). Such an approach was used for various pollutants including PAHs (3, 12, 15). In a previous study, annual emission of PAHs from fossil fuel in China was estimated to be 9800 tons in 1999 (16). However, those from biomass burning and aluminum production were not included, leading to an incomplete estimation of the total anthropogenic sources. The objective of this study was to establish an emission inventory of PAHs in China, based on the rates of major PAHs emission activities and their EF values collected from the literature. Sixteen PAH species listed as U.S. Environmental Protection Agency (U.S. EPA) priority pollutants were included in the present study. The results are also presented at the provincial level to illustrate geographical distribution of the emissions. Temporal variation of the emissions from 1980 to 2003 was addressed as well.
Methodology PAH Species. Sixteen parent PAHs included in this study were naphthalene (NAP), acenaphthylene (ACY), acenaphthene (ACE), fluorene (FLO), phenanthrene (PHE), anthracene (ANT), fluoranthene (FLA), pyrene (PYR), benz[a]anthracene (BaA), chrysene (CHR), benzo[b]fluoranthene (BbF), benzo[k]fluoranthene (BkF), benzo[a]pyrene (BaP), dibenz[a,h]anthracene (DahA), benzo[g,h,i]perylene (BghiP), and indeno[1,2,3-cd]pyrene (IcdP). Emissions of individual PAH species as well as the total amount of the 16 PAHs (PAH16) were estimated. Source Identification. A preliminary investigation on potential emission sources in China was conducted. Each of the potential candidates was evaluated on the basis of data availability and a rough estimation on PAH emission. A final list generated for this study included (1) industrial coal combustion, including thermoelectric power station and heat plants; (2) domestic coal combustion, including residential and commercial sectors; (3) coke production, (4) nontrans10.1021/es0517062 CCC: $33.50
2006 American Chemical Society Published on Web 12/22/2005
port oil combustion, (5) transport oil combustion, (6) straw burning, (7) firewood burning, and (8) primary aluminum production. Coking industry was not included in the industrial coal item due to its exceptional PAH emission. Natural gas was removed from the original list simply because of its extremely small EF values. Source Quantification. Source-specific production (aluminum) and consumption (other sources) data for the period 1980-2003 in China were collected from officially published sources (10, 11, 17-20). Detailed information for individual provinces, autonomous districts, and municipalities on an annual basis was gathered. These data were used not only for total emission estimation in the country but also for spatially and temporally resolved emission presentation. For the provincial level estimation, data in 2003 were used because of the completeness of the data set for that particular year. Data on biomass burning from 1980 to 1990 were not available and were alternatively calculated by employing the data from 1991 to 2003. Linear regression models for prediction of firewood and straw consumptions were developed by use of rural population, per capita domestic coal consumption, and corn and soybean production as the independent variables (10, 11). The coefficients of determination of the regressions were 0.77 and 0.61 for straw and firewood, respectively. It should be indicated that, at this stage, these are the only data available for the source quantification. The accuracy of the statistics as well as the uncertainty associated with them should be further evaluated in the future. Emission Factors. Due to variability in the origin of the tested material, operation facilities and conditions, environmental settings, and measuring procedures, EFs were often reported as different values (21, 22). Therefore, an effort was made to collect a large data set from the literature (12, 21-33). Unfortunately, very limited data have been collected in China so far. Therefore, the emission factors, and subsequently the estimated emissions, are to be further evaluated in the future when more information becomes available. After outlier deletion by use of Grubb’s test (34), frequency distribution of EF values could be derived for each source. For the species with relatively larger sample size, the results of skewness-kurtosis tests (R ) 0.05) revealed that their EF values are log-normally distributed with similar coefficients of variation after log-transformation (34). Still, medians instead of geometric means were applied in the emission estimation, simply because the data numbers for several species in various source categories were too small to generate geometric statistics without great and unknown bias. The EFs were used throughout the study including temporal analysis. The EFs are provided in the Supporting Information. Uncertainty Analysis. High variabilities were expected for the EF values, which would result in high uncertainty in the emission estimation. Such uncertainty was evaluated by generating probability distributions of the emissions based on the distributions of both activity rates (sources) and EFs. The actual distribution statistics (log-normal) of all the 16 EFs were derived from the data set, while hypothesized lognormal distributions with a 5% coefficient of variation of log-transformed values were assigned to the source rates. Monte Carlo Simulation was repeatedly implemented by MATLAB (35, 36) with new input values randomly selected from within their respective probability distributions. Simulations were run 10 000 times for each source category, and the results manifested the effects of uncertainty in the input parameter statistics on the estimation.
Results and Discussion Annual PAH Emission in China. The annual consumption rates of industrial coal, domestic coal, coking coal, petroleum
FIGURE 1. Means and 10th, 25th, 75th, and 90th percentiles of PAH16 emissions from the major sources as the results of Monte Carlo simulation. for transport, petroleum for nontransport, straw, and firewood in 2003 in China were 1289, 132, 242, 85, 202, 333, and 204 million tons, respectively (10, 11). Meantime, 5.5 million tons of primary aluminum were produced (20). It was estimated that the total emission of PAH16 and total emission of seven carcinogenic PAH components among them (PAH7: BaA, CHR, BbF, BkF, BaP, IcdP, and DahA) were 25 300 and 3460 tons, respectively, in China in 2003. The emissions of individual PAH species from major anthropogenic sources in China in 2003 are given in the Supporting Information. It was expected that the estimates were associated with great uncertainties due to the large variation in the EF values applied. Such uncertainty is illustrated in Figure 1 as the result of the Monte Carlo simulation. Medians as well as several percentiles of the calculated results for four major sources are shown. If semi-interquartile ranges (the difference between the 75th and 25th percentiles) were calculated, they varied from 25% (firewood) to 42% (coke production) of the median values, indicating quite large uncertainties of the estimation. It was expected that there were slight differences between the directly calculated values and the Monte Carlo medians. It appears that the emission of PAH16 in 2003 was very close to the amount emitted in the United States in 1990 (26 500 tons), which has been reduced substantially since then (4, 5). If PAH7 was compared, however, emission from China was much higher than the 2000 or 1400 tons emitted from the United States in 1990 or 1996, respectively (4, 5). The relatively higher fraction of carcinogenic species in the PAH emission profile in China was closely related to the emission source pattern, which will be discussed later. Although the total emissions of PAH16 were similar between China (2003) and the United States (1990), which are analogous in area, there were significant differences in emission densities normalized to population (ca. 1.92 × 10-5 ton for China vs ca. 1.07 × 10-4 ton for the Unites States) and in emission intensities normalized to gross domestic production (GDP, in constant 1990 price) (1.47 × 10-5 ton/103 U.S. $ for China vs 4.6 × 10-6 ton/103 U.S. $ for the Unites States). These differences mostly resulted from low efficiency of energy utilization, widespread biomass burning, and large quantity of domestic coal combustion in China. In China, energy consumption for each dollar of GDP was about 2.5 times higher than that in the United States, and a total of 537 million tons of straw and firewood were burned in China in 2003 (11, 37). Contributions of Various Emission Sources. The major sources of PAH emission in China were identified as industrial and domestic coal combustion, coke production, primary aluminum production, transport and nontransport oil combustion, and biomass burning including straw and firewood. VOL. 40, NO. 3, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
703
FIGURE 2. Share of major PAH emission sources in China (2003), United Kingdom (2003), and United States (1990). Other industrial (Other ind.) in United Kingdom and in United States refers to industrial processes other than coking and aluminum production (Al production).
FIGURE 3. PAH emission profile in China (2003) compared with that in the Great Lakes region (GLR) (1996). Contributions of the major sources to the PAHs with various rings in China are also portrayed. The relative contributions of these sources to PAH16 and PAH7 emissions in China (2003) are illustrated in Figure 2. The results are compared with those in the United Kingdom (38) and the United States (4). PAH emission profiles were greatly different between China and the other countries compared (Figure 2). In terms of PAH16, biomass burning contributed 60% of the total in China, which is much higher than in the United Kingdom and the United States. if open fire is not included. In rural China, straws, stalks, and deadwood are most important fuels for cooking and heating, and they are often burned in stoves without forced blast. Moreover, open fire burning of straws in the fields is also a very common practice of getting rid of the material and of fertilizing agricultural soil in countryside (39). As the largest coal-producing country in the world, another important PAH source in China was domestic coal combustion, which contributed to 20% of the total and was a unique feature in the emission profile compared with the United States and United Kingdom. Most cities and towns in China rely on coal as the major fuel and the combustion facilities are often old-styled and not well maintained. In rural area, where coal is available, coal instead of biomass is also used in a careless manner (37, 40). Like biomass, 704
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 3, 2006
domestic coal combustion has very high EF values compared with industrial utilization. Consequentially, although the total amount of coal used domestically was around a tenth of that used in industry (132 vs 1289 million tons) (10, 11), PAH emission from the former was much higher than that from the latter (20% vs 0.3% for PAH16). This was particularly true for PAH7 (40% vs
