Atmospheric mercury emissions from China's primary nonferrous metal (Zn, Pb and Cu) smelting during 1949–2010

Atmospheric mercury emissions from China's primary nonferrous metal (Zn, Pb and Cu) smelting during 1949–2010

Atmospheric Environment 103 (2015) 331e338 Contents lists available at ScienceDirect Atmospheric Environment journal homepage: www.elsevier.com/loca...

1MB Sizes 1 Downloads 8 Views

Atmospheric Environment 103 (2015) 331e338

Contents lists available at ScienceDirect

Atmospheric Environment journal homepage: www.elsevier.com/locate/atmosenv

Atmospheric mercury emissions from China's primary nonferrous metal (Zn, Pb and Cu) smelting during 1949e2010 Xuejie Ye a, Dan Hu a, Huanhuan Wang a, Long Chen a, Han Xie a, Wei Zhang b, *, Chunyan Deng a, Xuejun Wang a, * a b

Ministry of Education Laboratory of Earth Surface Processes, College of Urban and Environmental Sciences, Peking University, Beijing 100871, China School of Environment and Natural Resources, Renmin University of China, Beijing 100872, China

h i g h l i g h t s  High-resolution mercury emission from nonferrous metal smelting is calculated.  Mercury speciation information is involved in the inventory.  A dynamic historical inventory with mercury species is developed.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 20 October 2014 Received in revised form 24 December 2014 Accepted 24 December 2014 Available online 26 December 2014

Primary nonferrous metal smelting is one of the most significant anthropogenic mercury emission sources. A spatially resolved mercury emission inventory over a long time span is essential for assessment of mercury source attribution and mercury transport modeling. In this study, based on updated technology-based emission factors, the atmospheric mercury emissions originating from primary zinc, lead and copper smelting in China were calculated. The inventory indicated that the total mercury emission from nonferrous metal smelting in China was 14.65 Mg in 2010, lower than the estimations in previous studies. The contributions of point and non-point sources were 23.3% and 76.7%, respectively. In 2010, the mercury emission from primary zinc, lead and copper smelting was 7.49, 6.05 and 1.10 Mg, respectively, and the Hg2þ, Hg0 and HgP emissions were 8.10, 6.16 and 0.75 Mg, respectively. Spatially, the province with the largest emission was Sichuan, followed by Henan, Gansu, Shaanxi, Hunan and Yunnan provinces. The historical emissions were estimated based on dynamic emission factors that take the temporal technology changes into consideration. During 1949e2010, the cumulative mercury emission from China's nonferrous metal smelting was 323.0 Mg, of which the emission from lead smelting accounted for 44.6%, followed by zinc smelting (32.8%) and copper smelting (22.6%). From 1949 to 2010, the contribution of mercury emission from zinc smelting increased from 1.4% to 53.7%, while that from lead smelting showed a decreasing trend. For copper smelting, its contribution reached the maximum (40.1%) in 1987. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Mercury emission inventory Nonferrous metal smelting Zinc Lead Copper

1. Introduction Mercury is a persistent environmental pollutant with high toxicity, and it has aroused a global concern due to its long range transport and bioaccumulation in the environment (Hu et al., 2012). Studies on mercury emissions from anthropogenic sources have

* Corresponding authors. E-mail addresses: [email protected] (W. Zhang), [email protected] (X. Wang). http://dx.doi.org/10.1016/j.atmosenv.2014.12.062 1352-2310/© 2015 Elsevier Ltd. All rights reserved.

continued for many years. Previous studies showed that China's mercury emissions from anthropogenic sources have reached 600 Mg/y (Streets et al., 2005; Pirrone and Mason, 2009; Wu et al., 2006; Pacyna et al., 2010), which accounts for half of Asia's emissions (Pan et al., 2006) and approximately 28e40% of the global emissions (Pacyna et al., 2006; Pirrone and Mason, 2009). Nonferrous metal smelting is believed to be one of the most important anthropogenic mercury emission sources. Wu et al. (2006) estimated that China's annual mercury emissions from nonferrous metal smelting reached 230e320 Mg during 1995e2005, contributing approximately 46% of the total

332

X. Ye et al. / Atmospheric Environment 103 (2015) 331e338

anthropogenic emissions of China. Of that contribution, 86% was from primary smelting of zinc, lead and copper. Pirrone et al. (2010) reported that the global emissions from nonferrous metal smelting had reached 310 Mg in 2007, of which 230 Mg was from China. Hylander and Herbert (2008) considered mercury removal efficiencies in their study and reported that approximately 83 Mg of mercury was released into the atmosphere from zinc, lead and copper smelting in China in 2005. With the mercury mass balance in the smelting processes, Wu et al. (2012) estimated that the 2010 mercury emissions from China's primary zinc, lead and copper smelting were 39.4, 30.6 and 2.5 Mg, respectively. Large uncertainties exist in anthropogenic mercury emission estimations. In the estimation of nonferrous metal smelting, the main impact factors include the mercury concentration in the ore, smelting technology, type of air pollution control devices (APCDs) and application percentages of the smelting technology and APCDs. For a bottom-up atmospheric mercury emissions inventory, the emission factor (EFs) method is used widely, and the associated uncertainties originate mainly from the lack of measured EFs for the various smelting processes. Field measurements showed that EFs of nonferrous metal smelting significantly depended on the smelting technologies and the presence or absence of emission control measures (Wang et al., 2010; Zhang et al., 2012). Although technology-based EFs are essential to reduce the uncertainty in the mercury emissions inventory, they were seldom adopted in previous studies. The atmospheric mercury species are primarily gaseous elemental mercury (Hg0), gaseous oxidized mercury (Hg2þ), and particulate-bound mercury (Hgp). These species not only have different risks on the environment and human health but also behave differently in the atmosphere (Corbitt et al., 2011). With the improvement of abatement technologies, the mercury removal efficiency and species profile changed significantly with time. This change is especially significant with sulfuric acid plants, where approximately 95% of the gaseous mercury could be removed from the flue gas (Hylander and Herbert, 2008). In this paper, mercury emissions from primary zinc, lead and copper smelting in China were estimated using the updated technology-based EFs. Mercury emissions from both point and non-point sources were estimated in 2010. The emissions from large-scale smelters (with metal production capacities larger than 30,000 Mg/day) were considered as point sources, while the emissions from small-scale smelters were allocated spatially at 0.1  0.1 as non-point sources. In addition, a dynamic historical inventory with mercury species was developed. The mercury emissions from 1949 to 2010 were estimated using dynamic EFs considering the technology changes over time. The changes of mercury speciation profiles with and without the equipped acid plant were taken into consideration, and mercury emission inventories of Hg0, Hg2þ, and HgP were developed accordingly. 2. Data sources and methodology 2.1. Emission factors (EFs) Various smelting technologies are adopted in China's nonferrous metal smelters. For zinc smelting, the types of smelting processes, including the hydrometallurgical process (HP) and pyrometallurgical process (PP), were important in mercury emissions. The pyrometallurgical process can further be divided into four types: the retort zinc smelting process (RZSP), imperial smelting process (ISP), electric zinc furnace (EZF) and artisanal zinc smelting process (AZSP). Lead smelting processes consist of five major types: the rich-oxygen pool smelting process (RPSP), sinter machine process (SMP), sinter pan or pot process (SPP), imperial sinter process (ISP), and artisanal lead smelting process (ALSP). For

copper smelting, the flash furnace smelting process (FFSP), richoxygen pool smelting process (RPSP), imperial furnace smelting process (IFSP) and roasting-leaching-electrolyzing process (RLEP) are currently the dominant techniques, and mercury emissions from RLEP are negligible (MEP, 2010). The electric furnace smelting process (EF) and revelatory furnace smelting process (RF) are backward technologies and have been eliminated. The technologybased EFs and related percentages of metals produced in selected years are sourced from the Ministry of Environmental Protection of China (MEP, 2010) and Wu et al. (2012), and they are listed in Table 1. In general, the smelting processes with air emission control measures and low mercury EFs are mostly applied in large-scale plants. USGS reported that the numbers of China's large smelting plants (with metal production capacities larger than 30,000 Mg/day) of primary zinc, lead and copper were 12, 17 and 19 in 2007, respectively (USGS, 2010). These large plants are regarded as point sources in this study (Fig. 1). Previous surveys showed that there were few changes in the number of the large smelting plants of nonferrous metals during 2007e2010 (CNMIA, 2008e2011). Therefore, it is reasonable to believe that the locations of the point sources in 2010 were the same as those in 2007. Meanwhile, the numerous small-scale plants (with metal production capacities smaller than 30,000 Mg/day) are treated as non-point sources. On the basis of the proportions of smelting technologies adopted and the related EFs, we can calculate the weighted national equivalent (average) EFs of zinc, lead and copper smelting, and the results are presented in Table 1. In 2010, the equivalent EFs of point and non-point sources was obtained according to the corresponding smelting technologies used in point and non-point sources respectively (shown in Table 1), and listed in the last line in Table S1

Table 1 The technology-based EFs and related contribution percentages in selected years. Metal

Time Process c

Percentage (%)a EFs (g/Mg)b Equivalent EFs (g/Mg)

81.2 2010 HP ISPc 7.1 RZSP 7.9 EZF 1.3 Others 2.5 2004 HP 71.8 ISP 7.7 RZSP 13 EZF 5.9 AZ 1.6 1998 HP 67 ISP 10 RZSP 20 EZF/AZ 3 Lead 2010 RPSPc 47.3 ISPc 5.1 SMP 20.2 SPP 27.4 1984 SMP/others 100 c Copper 2010 FFSP 34.2 RPSPc 52.4 IFSP 9.8 RLEPc 0.2 EF/RF 3.4 2006 FFSP/RPSP 83.2 IFSP 9.7 EF/RF 7.1 2000 FFSP/RPSP 75 Others 25 1985 Others 100

Zinc

a

0.37 0.37 5.9 10 23.7 0.37 0.37 5.9 10 23.7 0.37 0.37 5.9 23.7 1.4 0.23 3 3 11.1 0.041 0.041 1.5 0 6.7 0.041 1.5 6.7 0.041 6.7 6.7

1.52

2.03

2.18

2.10

11.1 0.41

0.66

1.71 6.7

Percentages of smelting processes were sourced from Wu et al. (2012). Technology-based EFs were cited from MEP (2010). These smelting processes are generally used in large-scale plants (regarded as point sources), which have more advanced technologies and lower EFs. b

c

X. Ye et al. / Atmospheric Environment 103 (2015) 331e338

333

Fig. 1. Mercury emissions from point sources of nonferrous metal smelting in China in 2010.

in the supporting materials. Based on the mercury content in concentrates consumed in each province and the average mercury content in concentrates in China in 2010 (Wu et al., 2012), we adjusted the national equivalent EFs of point and non-point sources into the respective provincial equivalent EFs (Table S1). EFs for Hg0, Hg2þ and HgP were obtained by multiplying mercury EFs and the speciation fractions. In previous studies, the mercury speciation profile was assumed to be 80% Hg0, 15% Hg2þ and 5% Hgp for nonferrous metal smelting (Pacyna and Pacyna, 2002), but the mercury removal effect of APCDs was not considered. Because various smelting processes and APCDs were used in China's smelters, especially the equipping of an acid plant with a conversion and absorption tower (CAT), the mercury speciation profile has changed significantly (Kamata et al., 2008; Lee and Bae, 2009). The sulfuric acid can oxidize mercury in situ, and mercury can be oxidized by the vanadium pentoxide catalyst bed in the acid plant, which was utilized for conversion of SO2 to SO3 (Habashi, 1978; Straube et al., 2008; Li et al., 2010). Zhang et al. (2012) and Wang et al. (2010) confirmed a high percentage of Hg2þ in the flue gas after the acid plant, which has a shorter lifetime in the atmosphere. Therefore, in this study, based on whether CAT was applied, the species profile was divided into two groups. For the smelters adopting CAT, data measured by Wu et al. (2012) were used, i.e., 30% Hg0, 65% Hg2þ and 5% Hgp. For the smelters without CAT, data cited from Pacyna and Pacyna (2002) were used, i.e. 80% Hg0, 15% Hg2þ and 5% Hgp. Previous studies used a sigmoid curve to simulate the dynamics of technology changes in estimating historical and future emissions of carbon aerosol (Bond et al., 2007; Streets et al., 2004) and mercury (Streets et al., 2011) on global scales. However, the inventories suffered from large uncertainties due to regional variations of technologies and related EFs. In this paper, we used the following equation to reflect the variation of equivalent EFs of all processes over time (Streets et al., 2004, 2011; Bond et al., 2007):

ðtt0 Þ 2s2

2

yt ¼ ða  bÞe

þb

(1)

where yt is the equivalent mercury EF in t year, a is the equivalent EF of the worst technology in history, b is the equivalent EF of the best technology available today, t0 is the base year, and s is the curve parameter. Based on the equivalent EFs (Table 1), the values of parameters a, b, t0 and s can be calculated, and the results are provided in Table 2. Then, we can estimate the values of equivalent EFs at any point of time with Eq. (1).

2.2. Production data and emissions calculation Metal productions of all large plants (regarded as point sources) in 2010 can be calculated using the percentage of processes applied in large plants (Table 1) and the provincial-level production of primary zinc, lead and copper smelting. The data of provincial-level metal production were cited from the Yearbook of Nonferrous Metals Industry of China (2011). Meanwhile, the metal production of all small plants (regarded as non-point sources) can be obtained by subtracting the corresponding production of large plants from the total production. The mercury speciation profile changed over time with the development of CAT. We calculated the proportion of CAT application on an annual basis according to the following equations:

Table 2 Parameter values of the sigmoid curve to simulate the dynamics of EFs changes by year.

Zinc Lead Copper

a

b

s

t0

23.7 8.4 19.2

0.37 0.041 0.23

67.2 13.36 23.7

1847 1976 1959

334

X. Ye et al. / Atmospheric Environment 103 (2015) 331e338

3. Results and discussion

ZðTÞ ¼ 1  XðTÞ

(2) 3.1. Mercury emissions in 2010





ðTT0 Þ 2S2

XðTÞ ¼ X0  Xf e

2

þ Xf

(3)

where Z(T) is the percentage of CAT adopted in the smelting process in year T, X(T) is the percentage of CAT not adopted in the smelting process in year T, X0 is the initial ratio (usually 1), Xf is the final ratio (usually 0), T0 is the base year, and S is the curve parameter. Eq. (3) can be simplified to the following equation:

XðTÞ ¼ e

ðTT0 Þ 2S2

2

(4)

The value of parameter S was calculated based on percentages of CAT in 2010 (Wu et al., 2012) and is listed in Table S2. Then, the curve of the percentage of CAT not adopted in the smelting process during 1949e2010 can be fitted using SPSS software (Fig. S1). Finally, the annual adopted percentages of CAT in zinc, lead and copper smelting and the mercury speciation fraction can be obtained. The mercury emissions from all large and small plants at the provincial level in 2010 were obtained by multiplying the corresponding production and the equivalent EFs of point and non-point sources by province. Then, the point source emissions were allocated to each plant by using the capacity data (USGS, 2010), while the non-point emissions were allocated spatially at 0.1  0.1 using the population density. The calculations were conducted for all grids to generate spatial distributions at the province level. The historical emissions were derived based on temporally varied technologies and EFs.

2.3. Uncertainty analysis The uncertainties of the inventories were evaluated by Monte Carlo simulation. Variations in metal production, mercury content in concentrates and equivalent EFs were considered, and the technology-based EFs (MEP, 2010) were assumed to be constant in this study. The input parameters randomly selected from the corresponding statistical distributions were used in the calculations. The activity level is generally assumed to be a normal distribution of 5% coefficient variation for the metal production (Zhao et al., 2011). The mercury contents in the concentrates consumed in each province are considered to have lognormal distributions (Wu et al., 2012). For the historical emissions during 1949e2010, the equivalent EFs are generally described by triangular distributions, with the lower and upper bounds being the highest and lowest values of EFs (Zhao et al., 2011). By performing 10,000 Monte Carlo simulation iterations, the range of mercury emissions (with a 90% confidence interval) was derived. Sensitivity analysis was also conducted using Monte Carlo simulations for the 2010 mercury emissions estimation. The degree of influence can be determined by the correlation coefficient, and a higher coefficient value indicates higher influence on the inventory output (Chen et al., 2013). There are three factors associated with the uncertainties of mercury emissions, i.e. metal production amounts, mercury content in the consumed concentrates in each province, and equivalent EFs. In this study, the equivalent EFs were calculated based on the mercury content in concentrates consumed in each province and the average mercury content in concentrates. Therefore, two parameters were considered in the sensitivity analysis, i.e. metal production amounts and mercury content in the consumed concentrates by provinces.

The production and mercury emissions from China's primary zinc, lead and copper smelting in 2010 are shown in Table 3. The total mercury emission originating from nonferrous metal smelting was 14.65 Mg, of which the emission from zinc, lead and copper smelting was 7.49, 6.05 and 1.10 Mg, respectively. The results are much lower than the estimations made in previous studies (Streets et al., 2005; Wu et al., 2006; Hylander and Herbert, 2008; Wu et al., 2012). This disparity can be attributed mainly to the EFs used in different studies. In previous studies on mercury emissions from China's nonferrous metal smelting, the EFs used were measured in developed countries. For example, Pirrone et al. (1996) considered that the mercury EFs for zinc and lead smelting in developing countries were 25 and 3 g/Mg, respectively. Streets et al. (2005) and Wu et al. (2006) used EFs measured in Europe, i.e., 86.6 g/Mg for zinc, 43.6 g/Mg for lead and 9.6 g/Mg for copper. However, due to the differences in processes and pollution-control techniques adopted in China, the mercury EFs may differ significantly from those measured in the developed countries. In addition, because the mercury removal effect of APCDs was not considered in these EFs, the mercury emissions were over estimated to some extent (Wang et al., 2010; Zhang et al., 2012). To promote research on mercury emission inventories, China's Ministry of Environment Protection investigated the amounts of mercury emissions from various industries, and a technical report on the national survey of mercury pollution sources was published in 2010. The EFs of mercury from various nonferrous metal smelting processes were recommended in this report (MEP, 2010). The EFs were 0.37e23.7, 0.23e19.2 and 0.041e8.4 g/Mg for zinc, lead and copper, respectively. These recommended EFs were used in this study, which should be more suitable in estimating the mercury emission inventories in China. Hylander and Herbert (2008) estimated that China's nonferrous metal smelting emitted approximately 83 Mg in 2005, making it the largest contributor to the global mercury emissions. The mercury contents in the concentrates used by Hylander and Herbert (2008) were from the major mines throughout the world. No data on China's mines were used in their study. Meanwhile, as the application of APCDs is growing rapidly, their EFs could not be applied to the current emissions estimates in China. In the study by Wu et al. (2012), the mercury emission from China's nonferrous metal smelting was estimated to be 72.5 Mg in 2010. This is an upto-date mercury emission inventory from nonferrous metal smelting in China, using a technology-based method and considering the mercury content in nationwide concentrates and mercury removal efficiencies of APCDs. The EFs used in Wu's estimation were based on survey results (Wu et al., 2012), and the values (i.e. 5.17e45.75, 1.19e29.35, 0.28e14.96 g/Mg for zinc, lead and copper, respectively) were higher than the recommended EFs in the MEP 2010 report (MEP, 2010). In this study, the mercury emissions were calculated based on the MEP recommended EFs, and the differences in smelting processes (including APCDs) and mercury contents of concentrates were comprehensively considered (Table 1). Therefore, the mercury emissions estimated in this study were lower than the literature studies and could be closer to the actual situation in China. Results of this study show that the mercury emissions from point and non-point sources accounted for 23.3% and 76.7%, respectively. The atmospheric mercury emissions from China's nonferrous metal smelting were mainly from the non-point sources, but the metal production of the small-scale plants accounts for only 21.5% of the national total. Therefore, the mercury emissions

X. Ye et al. / Atmospheric Environment 103 (2015) 331e338

335

Table 3 The provincial production and mercury emissions from primary nonferrous metal smelting in 2010. Zinc

Lead

Production (kiloton)a

Mercury emissions (kg) P

Beijing Tianjin Hebei Shanxi Inner Mongolia Liaoning Jilin Heilongjiang Shanghai Jiangsu Zhejiang Anhui Fujian Jiangxi Shandong Henan Hubei Hunan Guangdong Guangxi Hainan Chongqing Sichuan Guizhou Yunnan Tibet Shaanxi Gansu Qinghai Ningxia Xinjiang Total a b

0.0 0.0 0.0 0.4 366.2 383.7 0.0 0.0 0.0 0.0 33.3 1.8 10.2 1.8 0.0 273.4 0.0 1185.2 258.5 485.8 0.0 0.0 270.8 21.3 863.3 0.0 547.9 235.1 93.5 0.0 1.4 5033.6

b

0.0 0.0 0.0 0.0 39.3 27.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 39.0 0.0 94.5 76.1 45.0 0.0 0.0 0.0 0.0 135.4 0.0 358.1 842.0 0.0 0.0 0.0 1656.9

b

Production (kiloton)

b

N

T

0.0 0.0 0.0 0.9 37.5 26.2 0.0 0.0 0.0 0.0 7.4 1.9 1.4 0.7 0.0 37.2 0.0 90.2 72.6 42.9 0.0 0.0 3986.3 52.3 129.2 0.0 341.7 803.6 199.1 0.0 5.8 5836.9

0.0 0.0 0.0 0.9 76.8 53.7 0.0 0.0 0.0 0.0 7.4 1.9 1.4 0.7 0.0 76.2 0.0 184.7 148.7 87.9 0.0 0.0 3986.3 52.3 264.6 0.0 699.8 1645.6 199.1 0.0 5.8 7493.8

0.0 0.0 0.0 0.0 55.1 32.1 0.0 0.0 0.0 0.0 0.0 77.7 6.4 93.1 0.0 951.2 0.0 758.6 108.4 149.8 0.0 12.6 0.0 5.7 379.0 0.0 89.6 26.9 5.8 28.4 13.5 2794.0

Copper Mercury emissions (kg) P

N

T

0.0 0.0 0.0 0.0 0.0 49.8 0.0 0.0 0.0 0.0 0.0 14.6 0.0 75.1 0.0 688.5 0.0 397.8 158.4 37.9 0.0 0.0 0.0 0.0 210.9 0.0 0.0 10.6 0.0 0.0 0.0 1643.6

0.0 0.0 0.0 0.0 513.7 88.1 0.0 0.0 0.0 0.0 0.0 25.8 19.1 132.8 0.0 1216.9 0.0 703.0 279.9 67.0 0.0 37.8 0.0 17.2 372.8 0.0 607.3 18.8 0.5 264.9 40.6 4406.2

0.0 0.0 0.0 0.0 513.7 137.9 0.0 0.0 0.0 0.0 0.0 40.4 19.1 207.9 0.0 1905.4 0.0 1100.8 438.3 105.0 0.0 37.8 0.0 17.2 583.7 0.0 607.3 29.4 0.5 264.9 40.6 6049.8

Production (kiloton)

0.0 8.3 4.4 80.6 178.8 39.3 0.4 0.4 65.8 226.3 105.9 401.9 11.1 491.1 308.0 39.4 186.1 0.0 0.0 0.0 0.0 7.7 3.0 0.1 290.1 1.5 1.0 465.4 0.0 0.0 4.1 2920.7

The production of zinc, lead and copper in China in 2010 was sourced from CNMIA (2011). P is the point source, N is the non-point source, T is the total (P þ N).

Fig. 2. Mercury emissions from non-point sources of nonferrous metal smelting in China in 2010 (0.1  0.1 resolution).

Mercury emissions (kg) P

N

T

0.0 0.3 0.0 0.8 1.9 1.1 0.0 0.0 0.0 8.6 0.0 5.7 0.0 20.7 4.7 1.4 4.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 44.0 0.0 0.0 12.7 0.0 0.0 0.0 106.1

0.0 1.6 15.9 4.0 9.7 5.7 1.4 1.2 186.9 44.0 171.7 29.3 31.4 105.6 24.1 6.9 21.0 0.0 0.0 0.0 0.0 21.8 8.3 0.4 224.0 3.9 1.9 64.5 0.0 0.0 11.6 996.8

0.0 1.9 15.9 4.8 11.6 6.8 1.4 1.2 186.9 52.7 171.7 35.0 31.4 126.3 28.8 8.2 25.1 0.0 0.0 0.0 0.0 21.8 8.3 0.4 268.0 3.9 1.9 77.2 0.0 0.0 11.6 1102.9

336

X. Ye et al. / Atmospheric Environment 103 (2015) 331e338

per unit output from non-point sources were twelve times higher than those from point sources. Obviously, it was caused by the high EFs of the backward smelting processes applied in small plants. For the different mercury species, approximately 8.10, 6.16 and 0.75 Mg were emitted as Hg2þ, Hg0 and Hgp, respectively. For zinc, lead and copper smelting, Hg2þ emissions accounted for 58.1%, 34.0% and 7.8%, Hg0 emissions accounted for 47.9%, 45.8% and 6.3%, and Hgp emissions accounted for 53.6%, 39.1% and 7.2%, respectively. 3.2. Regional distribution The provincial mercury emissions from nonferrous metal smelting in 2010 were listed in Table 3, and the corresponding contributions were shown in Fig. S2. The top six provinces with mercury emissions included Sichuan, Henan, Gansu, Shanxi, Hunan and Yunnan, accounting for 78.2% of the national total. Sichuan province had the largest contribution (27.3%) to the national emissions. The spatial distributions of point source and non-point source mercury emissions in 2010 are shown in Figs. 1 and 2, respectively. Point source emissions illustrated two zonal distributions in the eastewest direction, and the mercury emissions were mainly from Gansu, Henan, Hunan, Yunnan, Shanxi and Guangdong provinces, accounting for 90.1% of the national total point source emissions. For non-point sources, the mercury emissions were mainly from Sichuan (including Chongqing), Henan, Shaanxi, Gansu, Hunan and Yunnan provinces, accounting for 76.6% of the national total non-point source emissions; Sichuan (including Chongqing) contributed 35.5% of the total non-point emissions. In addition, there were significant mercury emissions in Shanghai (shown in Fig. 2), which were mainly from non-point source emissions of the local copper smelters. For zinc smelting, the total mercury emitted into the atmosphere in 2010 was 7.49 Mg, of which the point source emissions accounted for 22.1%. Mercury emissions from zinc smelting were high in the provinces of Sichuan, Gansu, Shaanxi, Yunnan, Qinghai and Hunan (Fig. S2), accounting for 53.2%, 22.0%, 9.3%, 3.5%, 2.7% and 2.5% of the national emission from zinc smelting, respectively. Gansu province and Shaanxi province have the largest point source emissions. The main cause for Gansu's large emissions was the high mercury content in the concentrates, approximately ten times higher than the national average, and its point source emissions were more than half (50.8%) of the national point source emissions. For non-point source emissions, Sichuan, Gansu and Shaanxi were the largest three provinces. Sichuan accounted for 68.3% of the national non-point source emissions and 53.2% of the national total mercury emissions. Its emissions were mostly emitted from nonpoint sources because it had no large-scale zinc smelter. Shaanxi's mercury emissions from both point and non-point sources were lower than Gansu because its mercury content in the concentrates was less than 1/5 of that in Gansu, although its zinc production was twice as high. Hunan province is famous for its zinc production, accounting for 23.7% of the national zinc output. However, the low mercury content of its concentrates (less than 1/4 of the national average) made its emissions insignificant. In Hunan province, the mercury emissions from point and non-point sources accounted for 2.5% of the national total. Mercury emission from lead smelting was 6.05 Mg in 2010. The proportion of non-point source emissions was 72.8%, while the corresponding lead production accounted for 47.8% of the national total. The mercury emission from lead smelting was mainly from Henan, Hunan, Shaanxi, Yunnan, Inner Mongolia and Guangdong provinces (Fig. S2), contributing approximately 31.5%, 18.2%, 10.4%, 9.7%, 8.5% and 7.2%, respectively. The large lead productions in Henan, Hunan and Yunnan provinces (accounting for 34.0%, 27.1% and 13.6% of the national output) resulted in their high mercury

emissions. Inner Mongolia and Guangdong's high emissions were due to the high mercury content of the lead ore. Shaanxi's emission was mainly from non-point sources. In 2010, mercury emission from copper smelting was 1.1 Mg, and approximately 90.4% was from non-point sources. The copper production in non-point sources accounted for only 6.9% of the national output. The mercury emission from copper smelting was mainly from Yunnan, Shanghai, Zhejiang, Jiangxi and Gansu provinces (Fig. S2). The high mercury content in copper concentrates consumed in Yunnan province resulted in an EF approximately four times higher than the national average. Yunnan's point source, nonpoint source and total emissions were all the largest, accounting for 41.5%, 22.5% and 24.3% of the corresponding national emissions, respectively. In Shanghai and Zhejiang provinces, almost all copper smelters were small-scale, and their mercury emissions were mainly related to the applied processes. Although the production was small, their mercury emissions were large due to the high EFs. Jiangxi and Gansu provinces had significant mercury emissions because they produced approximately one-third of the national copper production (Table S1).

3.3. Historical mercury emissions during 1949e2010 The production of zinc, lead and copper increased from 0.1, 2.1 and 1.9 kiloton in 1949e5034, 2794 and 2921 kiloton in 2010, respectively. Over the last two decades, the annual average growths for zinc, lead and copper production was 16.9%, 15.5% and 13.8%, respectively. Fig. 3 shows the dynamic variation of mercury EFs from nonferrous metal smelting during 1949e2010. The variation of mercury EFs reflected the transition from old, small-scale and uncontrolled processes to modern and large-scale processes with emission controls. The historical trends of mercury emissions from China's nonferrous metal smelting are shown in Fig. 4. Cumulative atmospheric mercury emission during 1949e2010 was 323.0 Mg, and the emission from zinc, lead and copper smelting accounted for 32.8%, 44.6% and 22.6%, respectively. During the period of 1949e2010, the proportion of mercury emissions from zinc smelting in the three types of smelters increased from 1.4% in 1949 to 53.7% in 2010. The proportion of mercury emissions from lead smelting declined from 70.7% in 1949 to 39.1% in 2010. The proportion of mercury emissions from copper smelting reached a maximum (40.1%) in 1987 and then decreased rapidly. In this study, the temporal trend of mercury emissions from zinc, lead and copper smelting was calculated using the EFs and

Fig. 3. Metal production and equivalent EFs of zinc, lead and copper in China during 1949e2010.

X. Ye et al. / Atmospheric Environment 103 (2015) 331e338

337

to 41% in 2010, while the proportion of Hg2þ increased from 15% in 1949 to 54% in 2010. During 1949e2010, the cumulative emissions of Hg2þ, Hg0 and Hgp from China's nonferrous metals smelting were 205.4, 101.5 and 16.1 Mg, respectively (Fig. S3). For mercury emissions during 1949e2010, the percentage of Hg2þ emission from zinc, lead and copper smelting was 43.9%, 36.8% and 19.3%; Hg0 emission was 27.3%, 48.4% and 24.3%; and Hgp emission was 32.8%, 44.6% and 22.6%, respectively. 3.4. Uncertainty analysis

Fig. 4. Historical mercury emissions from zinc, lead and copper smelting in China during 1949e2010.

metal production in each year. The equivalent EFs during 1949e2010 ranged 1.60e7.74, 2.10e19.2 and 0.37e8.4 g/Mg for zinc, lead and copper smelting, respectively (shown in Fig. 3). For lead and copper smelting, the APCD technology was adopted in 1959 and 1976, respectively. Therefore, in the periods of 1949e1959 and 1949e1976, the EFs for lead and copper smelting were respectively higher than those in the periods thereafter. The historical mercury emissions from zinc, lead and copper smelting are presented in Fig. 4. From 1960 to 1969, mercury emissions from zinc, lead and copper smelting showed two significant declines. From China's Nonferrous Metal Statistics Yearbook, we can see two rapid reductions of the metal production (up to more than 50%) due to the well-known political campaigns during this period in China. During the period of 2000e2010, the mercury emission from copper smelting decreased gradually, while the emission from zinc smelting showed a continuous growth. For lead smelting, the mercury emission generally showed an increasing trend. The mercury emission amounts and percentage of different mercury species from nonferrous metal smelting during 1949e2010 are shown in Fig. 5. Similar to the total mercury, emissions of Hg2þ, Hg0 and HgP showed an overall increasing trend. In particular, Hg2þ emissions grew rapidly and became the largest component after 2006, exceeding Hg0. With the application of CAT in the acid plant, the proportion of Hg0 decreased from 80% in 1949

In this inventory, the uncertainty of mercury emissions was originated mainly from the metal production amounts, mercury content in the consumed concentrates in each province and equivalent EFs. The metal production data were cited from the Yearbook of Nonferrous Metals Industry of China, and the mercury contents in the concentrates were sourced from Wu et al. (2012). The equivalent EFs during 1949e2010 were calculated using a sigmoid curve. By performing 10,000 Monte Carlo simulation iterations, the variation of the total atmospheric mercury emissions from nonferrous metal smelting in 2010 was [67%, 134%], indicated by a 90% confidence interval of the mean value. For the historical emissions during 1949e2010, the range of mercury emissions indicated an overall increasing trend, in which the bottom bound of the range is 9% to 61% and the top bound is 18%e 79% (Fig. 6). Sensitivity analysis was conducted using Monte Carlo simulation for mercury emissions in 2010, and the results are presented in Table 4. The sensitivity analysis showed that the main contributor was the uncertainty of mercury content in the concentrates, which could be influenced significantly by geological conditions and ore compositions in different regions. In particular, the contribution of mercury content in zinc concentrates to the variance of emissions was 71.2% and 71.4% from point and non-point sources, respectively. The mercury content in lead concentrates also showed a notable influence, with the uncertainty contributions of 23.9% and 19.0% from point and non-point sources, respectively. The sensitivity analysis illustrated that improving the accuracy of mercury content in concentrates can reduce the uncertainty in mercury emission inventories for nonferrous metal smelting. 4. Conclusion In this paper, we presented an inventory of mercury emissions

Fig. 5. Historical emissions of various mercury species from nonferrous metal smelting during 1949e2010.

Fig. 6. Range of mercury emissions from nonferrous metal smelting in China during 1949e2010 (The top and bottom lines represent the 95th and 5th percentiles, respectively).

338

X. Ye et al. / Atmospheric Environment 103 (2015) 331e338

Table 4 The uncertainty contributions for mercury emissions in 2010. Point source (%) Hg contents in concentrates Zinc 71.2 Lead 23.9 Copper 3.2

Non-point source (%) Production Hg contents in concentrates 0.9 0.7 0.0

71.4 19.0 8.3

Production 0.9 0.3 0.0

from nonferrous metal smelting in China during 1949e2010. The total mercury emissions from China's nonferrous metal smelting in 2010 was estimated to be 14.65 Mg, lower than the values reported in previous studies. The mercury emission from primary zinc, lead and copper smelting was 7.49, 6.05 and 1.10 Mg, respectively. The mercury emissions were mainly originated from non-point sources, accounting for 76.7% of the total amount. This was mainly caused by the backward smelting process applied in the small-scale plants. For spatial distribution, the provinces with large mercury emission in 2010 included Sichuan, Henan, Gansu, Shaanxi, Hunan and Yunnan, accounting for 78.2% of the national total. The mercury emission of point sources was mainly from Gansu province, and the emission of non-point sources was mainly from Sichuan province. For zinc smelting, the largest mercury emission was originated from Sichuan province, and for lead and copper smelting, the largest mercury emission was from Henan and Yunnan provinces, respectively. During 1949e2010, the cumulative mercury emission from nonferrous metal smelting in China was 323.0 Mg, with 32.8%, 44.6% and 22.6% from zinc, lead and copper smelting, respectively. For mercury speciation, the cumulative emission of Hg2þ, Hg0 and HgP was 205.4, 101.5, 16.1 Mg, respectively. During 1949e2010, the contribution of Hg0 decreased from 80% to 41%, while Hg2þ increased from 15% to 54%. The changes of species fraction were mainly due to the application of CAT in the acid plants. Acknowledgment This study was funded by the National Natural Science Foundation of China (41130535, 41471403), and the China Geological Survey (Grant No. 12120113015200). Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.atmosenv.2014.12.062. References Bond, T.C., Bhardwaj, E., Dong, R., Jogani, R., Jung, S.K., Roden, C., Streets, D.G., Trautmann, N.M., 2007. Historical emissions of black and organic carbon aerosol from energy-related combustion, 1850e2000. Glob. Biogeochem. Cycles 21, GB2018,. http://dx.doi.org/10.1029/2006GB002840. Chen, C., Wang, H.H., Zhang, W., Hu, D., Chen, L., Wang, X.J., 2013. High-resolution inventory of mercury emissions from biomass burning in China for 2000e2010 and a projection for 2020. J. Geophys. Res. Atmos. 118 (21), 12248e12256. Chinese Nonferrous Metal Industry Association (CNMIA), 2008e2011. The Yearbook of Nonferrous Metals Industry of China (2008e2011). Chinese Nonferrous Metal Industry Association Press, Beijing. Chinese Nonferrous Metal Industry Association (CNMIA), 2011. The Yearbook of

Nonferrous Metals Industry of China (2011). Chinese Nonferrous Metal Industry Association Press, Beijing. Corbitt, E.S., Jacob, D.J., Holmes, C.D., Streets, D.G., Sunderland, E.M., 2011. Global sourceereceptor relationships for mercury deposition under present-day and 2050 emissions scenarios. Environ. Sci. Technol 45 (24), 10477e10484. Habashi, F., 1978. Metallurgical plants e how mercury pollution is Abated. Environ. Sci. Technol. 12 (13), 1372e1376. Hylander, L.D., Herbert, R.B., 2008. Global emission and production of mercury during the pyrometallurgical extraction of nonferrous sulfide ores. Environ. Sci. Technol 42, 5971e5977. Hu, D., Zhang, W., Chen, L., Chen, C., Ou, L.B., Tong, Y.D., Wei, W., Long, W.J., Wang, X.J., 2012. Mercury emissions from waste combustion in China from 2004 to 2010. Atmos. Environ. 62, 359e366. Kamata, H., Ueno, S., Naito, T., Yukimura, A., 2008. Mercury oxidation over the V2O5(WO3)/TiO2 Commercial SCR catalyst. Ind. Eng. Chem. Res. 47 (21), 8136e8141. Lee, W.J., Bae, G.N., 2009. Removal of elemental mercury (Hg(0)) by nanosized V2O5/ TiO2 catalysts. Environ. Sci. Technol. 43 (5), 1522e1527. Li, G.H., Feng, X.B., Li, Z.G., Qiu, G.L., Shang, L.H., Liang, P., Wang, D.Y., Yang, Y.K., 2010. Mercury emission to atmosphere from primary Zn production in China. Sci. Total Environ. 408 (20), 4607e4612. Ministry of Environmental Protection of China (MEP), 2010. A Technical Guideline About National Survey of Mercury Pollution Sources. Available at: http://www. zhb.gov.cn/gkml/hbb/bgt/201104/W020110408402621585604.pdf (accessed 20.12.14.). Pacyna, E.G., Pacyna, J.M., 2002. Global emission of mercury from anthropogenic sources in 1995. Water Air Soil Pollut. 137 (1e4), 149e165. Pacyna, E.G., Pacyna, J.M., Sundseth, K., Munthe, J., Kindbom, K., Wilson, S., Steenhuisen, F., Maxson, P., 2010. Global emission of mercury to the atmosphere from anthropogenic sources in 2005 and projections to 2020. Atmos. Environ. 44 (20), 2487e2499. Pacyna, E.G., Pacyna, J.M., Steenhuisen, F., Wilson, S., 2006. Global anthropogenic mercury emission inventory for 2000. Atmos. Environ. 40 (22), 4048e4063. Pan, L., Woo, J.H., Carmichael, G.R., Tang, Y.H., Friedli, H.R., Radke, L.F., 2006. Regional distribution and emissions of mercury in east Asia: a modeling analysis of Asian Pacific Regional Aerosol Characterization Experiment (ACE-Asia) observations. J. Geophys. Res. Atmos. 111 http://dx.doi.org/10.1029/ 2005JD006381. D07109. Pirrone, N., Keeler, G.J., Nriagu, J.O., 1996. Regional differences in worldwide emissions of mercury to the atmosphere. Atmos. Environ. 30 (17), 2987. Pirrone, N., Mason, R.P. (Eds.), 2009. Hg Fate and Transport in the Global Atmosphere, Measurements and Models. Springer, Geneva. Pirrone, N., Cinnirella, S., Feng, X., Finkelman, R.B., Friedli, H.R., Leaner, J., Mason, R., Mukherjee, A.B., Stracher, G.B., Streets, D.G., Telmer, K., 2010. Global mercury emissions to the atmosphere from anthropogenic and natural sources. Atmos. Chem. Phys. 10 (13), 5951e5964. Straube, S., Hahn, T., Koeser, H., 2008. Adsorption and oxidation of mercury in tailend SCR-DeNOx plantsdbench scale investigations and speciation experiments. Appl. Catal. B Environ. 79, 286e295. Streets, D.G., Bond, T.C., Lee, T., Jang, C., 2004. On the future of carbonaceous aerosol emissions. J. Geophys. Res. Atmos. 109, D24212,. http://dx.doi.org/10.1029/ 2004JD004902. Streets, D.G., Hao, J.M., Wu, Y., Jiang, J.K., Chan, M., Tian, H.Z., Feng, X.B., 2005. Anthropogenic mercury emissions in China. Atmos. Environ. 39 (40), 7789e7806. Streets, D.G., Devane, M.K., Lu, Z., Bond, T.C., Sunderland, E.M., Jacob, D.J., 2011. Alltime releases of mercury to the atmosphere from human activities. Environ. Sci. Technol. 45 (24), 10485e10491. USGS (US Geological Survey), 2010. Minerals Information. Available at: http:// minerals.usgs.gov/minerals/pubs/commodity/ (accessed 20.12.14.). Wang, S.X., Song, J.X., Li, G.H., Wu, Y., Zhang, L., Wan, Q., Streets, D.G., Chin, C.K., Hao, J.M., 2010. Estimating mercury emissions from a zinc smelter in relation to China's mercury control policies. Environ. Pollut. 158 (10), 3347e3353. Wu, Y., Wang, S.X., Streets, D.G., Hao, J.M., Chan, M., Jiang, J.K., 2006. Trends in anthropogenic mercury emissions in China from 1995 to 2003. Environ. Sci. Technol. 40 (17), 5312e5318. Wu, Q.R., Wang, S.X., Zhang, L., Song, J.X., Yang, H., Meng, Y., 2012. Update of mercury emissions from China's primary zinc, lead and copper smelters, 2000e2010. Atmos. Chem. Phys. 12 (22), 11153e11163. Zhang, L., Wang, S., Wu, Q., Meng, Y., Yang, H., Wang, F., Hao, J., 2012. Were mercury emission factors for Chinese non-ferrous metal smelters overestimated? Evidence from onsite measurements in six smelters. Environ. Pollut. 171C, 109e117. Zhao, Y., Nielsen, C.P., Lei, Y., McElroy, M.B., Hao, J., 2011. Quantifying the uncertainties of a bottom-up emission inventory of anthropogenic atmospheric pollutants in China. Atmos. Chem. Phys. 11, 2295e2308.