Mercury species in smelter and power plant plumes

J. Y. Lu, C. M. Banic and W. H. Schroeder

Meteorological Service of Canada, Environment Canada, 4905 Dufferin Street, Toronto, Ontario, Canada M3H 5T4

ABSTRACT

Information on mercury speciation is required to determine the transport, transformation and fate of mercury in the environment. As part of a study on Physical and Chemical Evolution of Aerosols in Smelter and Power Plant Plumes mercury species, i.e., total elemental mercury in the gas phase (TGM), reactive gaseous mercury (RGM) and total particulate mercury (TPM), were determined in stack plumes of a coal-burning power plant and a copper smelter and in ambient air up-wind of these stacks.

INTRODUCTION

Smelters and coal-burning power plants can be important anthropogenetic sources of mercury (Kerfoot et al., 1999). Mercury emission from these sources is mainly reported as total mercury. Because different chemical forms of mercury have different reactivities, bioavailabilities and toxicities (Mahaffey, 1999), information on mercury speciation is required both for determining its transport, transformation and fate in the environment (Schluter, 2000) and for assessing its toxicity and the influence on human health (Braeckman, 1998). In this study, TGM, TPM and RGM in stack emissions and in ambient air were measured during January and February, 2000. Two sources were studied: the Nanticoke coal-fired power generating station operated by Ontario Power Generation in Ontario, Canada and the Horne Smelter (copper) operated by Noranda Metallurgy in Quebec, Canada. The plumes were sampled at aging times of up to 1 hour using the National Research Council (NRC) DHC-6 Twin Otter aircraft. Ambient air was sampled by the aircraft and at a surface-based mobile laboratory (hereafter referred to as the ground site) located on a farm 19 km northwest of the Nanticoke facility and then on a small rise in the town of Rouyn 1.5 km southwest of the Horne Smelter. This study is further described in a paper by Banic et al. in these proceedings.

METHODS

Total gaseous mercury (TGM) and “elemental” mercury – Automatic measurements: TGM was measured using a Mercury Vapor Analyzer (Model 2537A, Tekran, Toronto). The sampling integration time was 5 minutes at the ground site and 2.5 minutes for the aircraft measurements. Under normal ambient sampling conditions the reactive fraction of the mercury in the gas phase is small (<10%). Thus at a ground site we consider that we are determining TGM.

The plume gases contain an unknown fraction of reactive to elemental gaseous mercury. They also may contain acidic and organic gases which can degrade the performance of the Tekran Mercury Vapor Analyzer. Because of this the instrument on the aircraft was protected by a trap containing soda lime, XAD and KCl to filter acidic gases, organics and reactive mercury species from the air stream. The intent for the aircraft Tekran was to measure elemental mercury only.

Total particulate-phase mercury (TPM) – Manual measurements: The aircraft samples were collected by filtering air stored in a 0.5 m³ conductive bag used to grab-sample at specific points in the plume. It took 2 seconds to fill the bag with plume air. The bag was repeatedly filled during a flight and sampled continuously for time periods from 0.5 to 1.5 hours for a total volume sampled through the filter up to about 0.5 m³. The technique developed by Lu et al. (1998) was modified and applied for sampling and analysis of total particulate-phase mercury in these studies. Particulate matter was collected on a quartz filtration medium situated in a quartz tube with a diameter of 20 mm. The mercury species associated with the particulate matter were thermally released and converted to Hg⁰ at 900^oC in a tube furnace (Revco Scientific, NC, USA). The elemental mercury released was carried by zero air at a flow rate of 1.5 L min^-1 to a Mercury Vapor Analyzer (Model 2537A, Tekran, Toronto) for quantification.

Reactive mercury was not scrubbed from the air stream before the particulate collection. Any RGM carried through the bag and inlet to the collection surface may be collected by the quartz surface or the particles themselves and be detected as particulate mercury.

Reactive gaseous mercury (RGM) – Manual measurements: The KCl-coated denuders have been developed and used for measurement of RGM in ambient air (Sommar et al., 1999; Feng et al., 2000). In this study, quartz annular denuders (University Research Glass, NC, USA) coated with KCl were used for adsorbing reactive gaseous mercury species from air. The denuders at the ground site were housed in a metal box and the temperature inside the box was maintained at about 30^oC (ambient temperatures ranged from –20 to 0 ^oC). The mercury species collected by the denuders were thermally released and converted to Hg⁰ at 550^oC in a tube furnace (Revco Scientific, NC, USA). The released Hg⁰ was carried by zero air at a flow rate of 1.5 L min^-1 to a Mercury Vapor Analyzer ((Model 2537A, Tekran, Toronto) for quantification.

The denuder tubes exposed on the aircraft were housed in a case that protruded out the roof of the aircraft forward of the prop line. The denuder was heated by a flow of cabin air (approximately 20^oC) over its exterior while sampling air near 0^oC or colder. The inlet to the denuder tube was a rear-facing 4mm inner diameter quartz tube about 1 inch in overall length with a 90^o bend. This orientation was used to keep all but the finest particles from entering the denuder.

Field Blanks for TPM and RGM: Field blanks for both ground site and aircraft measurements were processed just as actual samples, but the exposure time for the blank traps to the ambient air was about 2 minutes with no air flow.

Calibration: The mercury vapor analyzers were calibrated using both internal permeation mercury sources and a common external calibration source. In both cases a known volume of air saturated with elemental mercury vapor at a pre-selected temperature is injected to the analyzer.

RESULTS AND DISCUSSION

1. Blanks

Instrumental blank: Instrumental blank was obtained by flowing zero air through the Tekran analyzer instruments. The values obtained were below the limit of detection of the instrument, which corresponds to ~0.8 pg mercury for 5-minute sample integration time.

Aircraft TPM and RGM blank: Tables 1 and 2 list aircraft TPM and RGM blank values from Nanticoke and Rouyn-Noranda. Comparison of the results in the Tables reveals that the blank values for both TPM and RGM measurements were relatively high at the beginning and decreased at the end of the first field study (Table 1). The average blank values obtained from the second field study, Table 2, agree well with those obtained at the end of the first field study (the average values after Jan. 26 in Table 1). These results suggest improvement of the experimental procedures.

Table 1. Aircraft TPM and RGM blank values, Nanticoke, Ontario, Jan. 18-28

Flight # (date)	TPM blank pg	RGM blank pg
1 (Jan.18)	22.7	23.5
	3.1
	3.5
6 (Jan.22, pm)	61.4	22.7
7 (Jan.23)	7.4	18.1
	10.0	26.5
	10.5
9 (Jan.24, pm)	8.9	8.0
		14.0
11(Jan.26)	1.0	4.3
	3.9	1.3
	4.5	2.4
Average	12.4 ± 16.5	13.4 ± 9.2
Aver. before Jan.26	15.9 ± 18.1	18.8 ± 6.3
Aver. on Jan.26	3.1 ± 1.5	2.7 ± 1.2

Table 2. Aircraft TPM and RGM blank values, Rouyn-Noranda, Quebec, Feb. 14-24

Flight # (date)	TPM blank pg	RGM blank pg
15 (Feb.14)	5.5
16 (Feb.15)	4.7
17 (Feb.16)	4.6	3.0
		1.7
		3.2
18 (Feb.17)	5.7	0.6
	4.7	12.5
20 (Feb. 19)	3.2	6.5
21 (Feb. 21)	0.9
22 (Feb. 21)	2.0	5.4
23 (Feb. 22)	2.0
Average	3.7 ± 1.6	4.7 ± 3.7

Ground site TPM and RGM blank: Table 3 lists the blank values from ground sites of both locations. There is no significant difference in the average TPM and RGM blank values between ground site (Table 3) and aircraft (Table 2). These values are in the same range obtained by USEPA researchers (Stevens, personal communication) and suggest that the sampling and sample analysis protocols have been well established.

Table 3. Ground site TPM and RGM blank values

TPM blank, pg	RGM blank, pg
5.8	15.6
0.1	0.6
4.7	0.8
	3.9
	7.4
	1.2
	1.4
3.5 ±2.5	4.4 ± 5.1

2. Mercury species in plume gases and background air

Table 4 shows concentrations of mercury species in plume gases and background air at both sites for selected flight times. The aircraft intercepted the emissions at distances of 10 to 20 km from the stacks on these flights. For the flights considered in Table 4, the source emissions are characterized by higher concentrations of mercury measured as elemental mercury than reactive and particulate forms. Both sites show that TPM and RGM concentrations in plume gases are higher than in air at ground level. Higher concentrations of TPM and RGM observed at the ground site on Feb. 21 were a result of the wind carrying air from the vicinity of the sources towards the ground site. The TPM concentrations reported in Table 4 are generally lower (estimated at least 10%) than the actual values. The high resistance to flow of the particulate traps affects the performance of the mercury vapor analyzer when the traps are heated for more than 10 minutes. Thus particulate mercury not released during a 10-minute heating time is not detected.

Table 4. Mercury species in plume gases and background air

Sampling location/date	TGM or “elemental” Hg ng m^-3		TPM ng m^-3		RGM ng m^-3
	Ground site	Aircraft in plume	Ground site	Aircraft	Ground site	Aircraft
Nanticoke
Jan. 26	2.0	2.6	0.007±0.003	0.045	0.003±0.0003	0.10
Jan. 27	1.9	2.5	0.0009±0.0001	0.039	0.0009±0.0003	0.057
Jan. 28	2.0	£ 7 (variable)	0.002±0.0001	0.049	0.001±0.0004	0.010
Rouyn-Noranda
Feb. 19	1.7	£ 2.2 (variable)	0.003	0.16	0.002±0.0005	0.037
Feb. 21	1.6	1.4	0.025	0.54	0.037	0.51

CONCLUSIONS

The data shown here demonstrate that it is possible to use an aircraft platform to determine 3 different species of mercury in emissions from anthropogenic sources at distances of several tens of km from the source. Analysis of the winter data set will continue, with a focus on the influence of sunlight and aging time on the speciation of mercury. A summer field study is to follow in 2000.

Acknowledgment

Funding for this project was provided by the Toxic Substances Research Initiative, the Mining Association of Canada and Ontario Power Generation Inc. We thank the staff of Noranda Inc.-Horne Smelter and Ontario Power Generation for their assistance. Advice and discussions with Stéphane Robert of the Horne Smelter were invaluable in planning and executing the experiment. We also like to thank the pilots and staff of NRC-IAR, Steve Bacic and John Deary of MSC for their support.

REFERENCES

Braeckman B, Cornelis R, Rzeznik U, Raes H (1998), Environmental Research 79(1):33.

Feng XB, Sommar J, Gardfeldt K, Lindqvist O (2000), Fresenius J. Anal. Chem. 366(5):423.

Lu JY, Schroeder WH, Berg T, Munthe J, Schneeberger D, Schaedlich F (1998), Anal. Chem. 70:2403.

Kerfoot WC, Harting S, Rossmann R, Robbins JA (1999), J. Gr. Lakes Res. 25(4):663-682.

Mahaffey KR (1999), Public Health Rep. 114(5):397.

Schluter K (2000), Environ. Geol. 39(3-4):249.

Sommar J, Feng XB, Gardfeldt K, Lindqvist O (1999), J. Environ. Monit. 1(5):435.

Stevens, R. K., USEPA-RTP.