- Email: [email protected]
Mercury emissions into the atmosphere from a chlor-alkali complex measured with the lidar technique
AtmosphericEnvironmentVol.26A,No. 7,pp. 12531258,1992.
0(0)06981/92S5.00+0.00 © 1992PergamonPresspie
Printedin GreatBritain.
MERCURY EMISSIONS INTO THE ATMOSPHERE FROM A CHLOR-ALKALI COMPLEX MEASURED WITH THE LIDAR TECHNIQUE R. FERRARA a n d B. E. MASERT! C. N. R. -Istituto di Biofisica, Via San Lorenzo 26, 1-56127 Pisa, Italy and H. EDNER, P. RAGNARSON, S. SVANBERG a n d E. WALLINDER Department of Physics, Lund Institute of Technology P. O. Box 118, S-221 00 Lund, Sweden (First received 1 June 1991 and in final form 12 September 1991)
Abstract--Data are reported on atmospheric mercury fluxmeasured by a lidar systemat a chlor-alkali plant located in central Italy. Two mercury sources were identifiedover the electrolyticcell rooms. A flux value of 36 gh-1 was determined during the daytime while at night the value increase to 56 gh-1. The mercury emitted into the atmosphere was found to be 4 g per 1000 kg of chlorine produced. Atmospheric mercury concentrations were supplemented with some determinations made with the point monitor technique. Key word index: Mercury, atmosphere, chlor-alkali plant, lidar.
INTRODUCTION Mercury and its compounds are widely used for industrial and agricultural applications due to the unusual physico-chemical properties of these materials (Nriagu, 1979). One of the principal uses of mercury (which accounts for about 20% of the total consumption) has been in the production of caustic soda and chlorine. The chlor-alkali processes make use of mercury in the simultaneous production of chlorine and caustic soda by electrolysis of brine solution using a flowing cathode of metallic mercury, Twenty years ago it was estimated that 200-250 g of mercury could be lost to the environment for every 1000 kg of chlorine produced, In the most advanced chlor-alkali installations losses of mercury to the environment should be less than 3 g per 1000 kg of chlorine (Frei and Hutzinger, 1985). Mercury is released into the environment mainly in waste water (a), in the atmosphere (b) and in sludges (c): (a) Mercury discharged into waste water should not surpass a value of about 1 g per 1000 kg of chlorine produced; (b) Mercury emitted into the atmosphere (especially from electrolytic cell rooms) should be of the order of 1.5-2 g per 1000 kg of chlorine; (c) Mercury in sludges, old cell room parts, etc., is not easy to estimate, In the literature, data are available on mercury discharged into waste water from some chlor-alkali complexes (Renzoni, 1977; E1-Rayis et al., 1986), but data on atmospheric mercury are scarce and refer only
to the area around the complex (Lindqvist et al., 1991). Rather than dealing with direct measurements of atomic mercury concentration in air, many authors are discussing mercury content in vegetative organisms, used as biological indicators of atmospheric mercury pollution (Wallin, 1976; Lodenius and Herranen, 1981; Lodenius and Tulisalo, 1984; Hynninen and Lodenius, 1986; Gonzales, 1991). In the present paper we report a study on mercury emission into the atmosphere from a chlor-alkali complex located in central Italy. The chlor-alkali complex of Rosignano Solvay & Cie (Livorno) produces 120,000 tons y - 1 of chlorine. Consequently, the minimum loss of mercury into the environment should be about 360 kg y - l, if the most modern techniques are applied. From our previous work (Maserti and Ferrara, 1991) mercury discharged to the waste water in the coastal area has been estimated to be of the order of 150-180kgy -1. Few data are available on mercury concentration in the atmosphere around this chlor-alkali complex (Maserti and Ferrara, 1991), while no information is present on mercury emitted directly into the atmosphere by this complex. Research on mercury concentrations in plants and lichens collected in this area reveals that atmospheric background values (3-5 ng m - 3) are reached within a radius of 4-5 km from the emission point with an asymmetry in the spatial distribution due to the presence of prevailing winds. In order to evaluate the mercury concentrations in the atmosphere above the complex and the total flux of the metal emitted, a measurement campaign was performed in 1990 using a lidar system developed at the Lund Institute of Technology (Sweden).
1253
1254
R. FERRARAet al.
This remote-sensing technique allows elemental mercury mapping in the atmosphere at distances ranging up to 1000 m from the lidar position and also an evaluation of the mercury flux from a source if the speed and direction of the wind are known. The first study of atmospheric mercury emissions with the lidar technique was reported by Edner et al. (1989). In addition to measurements carried out with the lidar system, point monitors were used for an accurate evaluation of the mercury levels in the atmospheric layers close to the ground and for a further comparison of the results obtained with the two analytical techniques, as has been reported elsewhere (Ferrara et al., 1992).
EXPERIMENTAL Atmospheric mercury measurements by the lidar technique
The lidar remote-sensing system employed in the present study has been extensively described by Edner et al. (1987, 1989). The system is housed in a van and can work autonomously by means of a 20-kV diesel generator towed by the van. This laser-radar system is capable of generating pulse energies of up to 5 mJ with a linewidth of 0.OO1 nm at the mercury resonance line (253.6 nm), and with a repetition rate of 10 Hz. The laser beam is transmitted into the atmosphere using quartz prisms and a large plane mirror which can be rotated around both the horizontal and the vertical axes. Back-scattered radiation is reflected by the same mirror and collected with a Newtonian telescope to reach the detection system. The laser is tuned on and off the resonance line of mercury every laser shot, allowing differential absorption measurements. A transient recorder performs A / D conversion of the signal with a time resolution of 10 ns, giving a range resolution of about 1 m . The digital signals are averaged on a computer and stored on floppy disks. A single scan is accomplished in a time that is considerably shorter than lOOms; for a given measurement direction a few thousand scans are then averaged during a few minutes of integration time to obtain a satisfactory signal-to-noise ratio,
The detection limit is of the order of 2 ng m- 3 of elemental mercury. The mercury compounds present in the atmosphere cannot be detected with the lidar technique, but these compounds do not represent the dominant fraction, constituting generally only 5-10% of the total atmospheric mercury. Atmospheric mercury measurements by point monitors
The point monitors used by us in atmospheric mercury determinations have been described elsewhere (Ferrara et al., 1992). Gaseous mercury is collected on gold traps at a flow rate of 1 E min- ~ by means of a battery-operated membrane pump, for a period of time set on a programmable timer. Mercury, electrothermally desorbed, is determined by atomic absorption spectroscopy or atomic fluorescence spectroscopy (detection limit 0.01 ng of mercury). Determinations of atmospheric mercury levels carried out with the lidar system and with point monitors during an extensive field study provided comparable results (Ferrara et al., 1992). The speed and direction of the wind were measured by means of a weather station located near the electrolytic cells.
RESULTS AND DISCUSSION A measurement campaign was performed at the Rosignano Solvay chlor-alkali complex from 19 to 21 September 1990. In Fig. I the mobile lidar system is shown against the background of the chlor-alkali complex. A first horizontal screening carried out over the industrial scanning area allowed us to identify the main sources of atmospheric mercury. In Fig. 2 a typical horizontal map of the mercury distribution over the study area is given. D a t a refer to a height of about 5-10 m above the ground and were obtained on 21 September from 4 p.m. to 7 p.m. with a wind speed of 7 m s -1 and a wind direction of 220 °. The highest mercury concentration (up to 903 ng m - 3) was measured over the electrolytic cell rooms
Fig. I. The mobile lidar system against the background of the chlor-alkali complex of Rosignano Solvay (Livorno, Italy).
Mercury emissions measured by a lidar system
1
1255
,
~/ki IHI-'.:.~.
II
.............
Fig. 2. Horizontal map of the mercury distribution above the chlor-alkali complex measuCedby the lidar. Along the given measurement directions, mean mercury concentrations were evaluated as presented in the circles for the integration intervals marked along the lines. The length of the integration intervals were selected depending on the concentration level of mercury. The results of point monitors are included in square boxes. and downwind from this mercury source. A smaller mercury source with a concentration reaching 351 ng m - 3 seems to be present at about 200 m from the iidar van position. Until now the cause of this emission is unclear, even though the material removed from the settling tanks, formerly dumpedin the nearby field, can be suspected, The mercury levels measured in the atmosphere decrease rapidly a few hundred meters from the plant, reaching values quite comparable to the background ones outside the industrial area of the complex, Mercury determinations with point monitors were performed outside the complex. They give comparable results to those obtained with lidar, with the exception of the value (68 ng m-3) measured southwards from the electrolytic cells, probably due to the falling mercury vapour. In Fig. 3 vertical profiles of atmospheric mercury concentrations measured in two directions are plotted; the presence oftwo plumes reveals the existence oftwo distinct mercury sources as stated above. A maximum concentration value greater than 1000 ng m-3 was determined at a height of 10 m in the plume from the cell rooms. An example of a horizontal scan of the atmospheric mercury levels over the same area is given in Fig. 4, where the lidar curve showing the ratio between the on and off resonance line of mercury for
one of the directions is inserted. A horizontal ratio curve is obtained in the absence of mercury while its presence is evinced from the slope of the curve, which is directly related to the atmospheric mercury concentration. In Table 1 the results of atmospheric mercury measurements performed with the lidar system are reported together with the speed and direction of the wind. In the last column of the same table the flux values from the mercury sources are also indicated. Measurements were carried out downwind from the cell-room area, where the two plumes were located. The flux values have been calculated by integrating the mercury concentration over the area of the plume cross-section and then multiplying the integrated content by the component of the wind speed perpendicular to the scan. The two lowest values were observed in a position {C) upwind from the main emission source. Analysing our data we find that the average nighttime value (56gh -~) is higher than the daytime average (36gh 1). From 10p.m. to 6a.m. the chlor-alkali complex increases the chlorine production because of the lower cost of electricity and we observe this fact reflected in a higher mercury emission. It is important to point out that the data reported in Table 1 are the first direct measurements aimed at an evaluation of the mercury flux to the atmosphere of a
1256
R. FERRARAet al.
Fig. 3. Vertical profiles of atmospheric mercury concentrations measured in two directions upwind and downwind from the electrolytic cell room.
Table 1. Day, time, measurement direction, speed and direction of wind and mercury flux in the atmosphere above the chlor-alkali complex of Rosignano Solvay (Italy). The horizontal directions A-C for the vertical scans are as shown in Fig. 2
Day 19 Sept. 20 Sept. 21 Sept. 21 Sept. 21 Sept. 21 Sept. 21 Sept. 21 Sept. 21 Sept. 21 Sept. 21 Sept. 21 Sept.
1990 1990 1990 1990 1990 1990 1990 1990 1990 1990 1990 1990
Time
Measurement direction
Wind speed (m s- 1)
Wind direction
Mercury flux (g h - 1)
6.43 p.m. 11.34 p.m. 12.24 a.m. 12.54 a.m. 01.36 a.m. 02.09 a.m. 12.29 p.m. 01.13 p.m. 01.26 p.m. 05.06 p.m. 06.24 p.m. 06.37 p.m.
A B B B A B A A C A A C
2.52.5 7.5 9.0 9.0 9.0 9.0 8.0 9.0 6.0 6.0 6.0
220 ° 220 ° 220 ° 220 ° 220 ° 220 ° 220 ° 220 ° 220 ° 240 ° 230 ° 250 °
34 40 57 55 65 64 42 42 3 31 31 2
chlor-alkali plant. T h e only d a t u m (30 g h - 1) available in the literature was r e p o r t e d by E d n e r et al. (1989), a n d it was determined at a Swedish chlor-alkali p l a n t d u r i n g a n evaluation of the lidar system's capability for a t m o s p h e r i c mercury mapping.
T h e low flux values (3 a n d 2 g h - 1) observed o n 21 S e p t e m b e r 1990 at 1.26 p.m. a n d 6.37 p.m. u p w i n d from the m a i n cell house refer to a smaller source inside the chlor-alkali plant. U s i n g the collected d a t a a n d assuming they reflect
Mercury emissions measured by a lidar system
1257
Fig. 4. Horizontal scan of atmospheric mercury above the electrolytic cell room. The ratio of on and off resonance lidar curve for one of the directions is inserted.
normal complex activity, an average diurnal mercury flux of 43 g h - 1, corresponding to a mercury amount of 400 kg y-1 emitted into the atmosphere, has been calculated. A careful examination of the vertical profiles of the atmospheric mercury concentration determined with the lidar reveals that a large part of the mercury vapor emitted from the cell rooms falls down close to the ground. Depending on the surrounding buildings it is not always possible to scan from the ground level, From the measured plume cross-sections we have estimated that 20% of emitted mercury goes undetected. Consequently, the mercury amount discharged into the atmosphere from the chlor-alkali complex is of the order of 500 kg y - 1. Taking into account the chlorine production of this complex the mercury emission into the atmosphere is 4 g per 1000 kg of chlorine. This finding shows that the chlor-alkali plant of Rosignano Solvay uses quite good production techniques, even though they can be
further improved by applying the most modern technology. CONCLUSION From the atmospheric mercury determination performed with the lidar above the chlor-alkali plant of Rosignano Solvay two distinct sources were located. The main mercury source is the emission from the electrolytic cells with a flux of 31-65 g h - 1; the other source, of minor importance, presents a flux value of 2-4 g h-1 and is located over old deposits of solid wastes removed from the settling tanks. The distribution of the mercury concentrations in the atmosphere obtained with the lidar evidences the presence of these hot spots and a sharp decrease of the mercury levels reaching background values about 1-2 km from the complex. From the vertical scan of the mercury concentrations in the plume it appears evident that the
1258
R. FERRARAet al.
maximum concentration is reached just above the ventilation outlet of the cell rooms (25 m) and that the
Gonzales H. (1991) Mercury pollution caused by a chloralkali plant. Wat. Air Soil Pollut. 56, 83-93. Hynninen V. and Lodenius M. (1986) Mercury pollution near mercury vapor falling down rapidly results in elevated an industrial source in southwest Finland. Bull. Envir. values in a restricted area at ground level. Contain. Toxicol. 36, 294-298. Lindqvist O., Johansson K., Aastrup M., Andersson A., Bringmark L., Hovsenius G., Hakanson L., Iverfeldt A., Meili M. and Timm B. (1991) Mercury in the Swedish environment Wat. Air Soil Pollut. 55. REFERENCES Lodenius M. and Herranen M. (1981) Influence of a chloralkali plant on the mercury contents of fungi. Chemosphere Edner H., Fredriksson K., Sunesson A., Svanberg S., Uneus 10, 313-318. L. and Wendt W. (1987) Mobile remote sensing system for Lodenius M. and Tulisalo E. (1984) Environmental mercury atmospheric monitoring. Appl. Opt. 26, 4330-4338. contamination around a chlor-alkali plant. Bull. Envir. Edner H., Faris G. W., Sunesson A. and Svanberg S. (1989) Contam. Toxicol. 32, 439-444. Atmospheric atomic mercury monitoring using differential Maserti B. E, and Ferrara R. (1991) Mercury in plants, soil absorption lidar techniques. Appl. Opt. 28, 921-930. and atmosphere near a chlor-alkali complex. Wat. Air Soil El-Rayis O. A., Halim Y. and Aboul-Dahab O. (1986) Total Pollut. 56, 15-20. mercury in the coastal marine ecosystem west of Alexan- Nriagu J. O. (1977) The production and uses of mercury. dria. FAO Fish. Rep. 325 (Suppl.), 58-73. In The Biogeochemistry of Mercury in the Environment Ferrara R., Maserti B. E., Edner H., Ragnarson P., Svanberg (edited by Nriagu J. O.). Elsevier/North Holland BiomedS. and Wallinder E. (1992) Atmospheric mercury deterical Press, Amsterdam, New York, Oxford. minations by Lidar and Point Monitors in environmental Renzoni A. (1977) A case of mercury abatement along the studies. Toxicol. envir. Chem. (in press). Tuscan coast. J. Etud. Pollut. CIESM 2, 95-97. Frei R. W. and Hutzinger O. (eds) (1985) Analytical Aspects of Wallin T. (1976) Deposition of airborne mercury from six Mercury and Other Heavy Metals in the Environment. Swedish chior-alkali plants surveyed by moss analysis. Gordon and Breach, London, New York. Envir. Pollut. 10, 102-114.
0(0)06981/92S5.00+0.00 © 1992PergamonPresspie
Printedin GreatBritain.
MERCURY EMISSIONS INTO THE ATMOSPHERE FROM A CHLOR-ALKALI COMPLEX MEASURED WITH THE LIDAR TECHNIQUE R. FERRARA a n d B. E. MASERT! C. N. R. -Istituto di Biofisica, Via San Lorenzo 26, 1-56127 Pisa, Italy and H. EDNER, P. RAGNARSON, S. SVANBERG a n d E. WALLINDER Department of Physics, Lund Institute of Technology P. O. Box 118, S-221 00 Lund, Sweden (First received 1 June 1991 and in final form 12 September 1991)
Abstract--Data are reported on atmospheric mercury fluxmeasured by a lidar systemat a chlor-alkali plant located in central Italy. Two mercury sources were identifiedover the electrolyticcell rooms. A flux value of 36 gh-1 was determined during the daytime while at night the value increase to 56 gh-1. The mercury emitted into the atmosphere was found to be 4 g per 1000 kg of chlorine produced. Atmospheric mercury concentrations were supplemented with some determinations made with the point monitor technique. Key word index: Mercury, atmosphere, chlor-alkali plant, lidar.
INTRODUCTION Mercury and its compounds are widely used for industrial and agricultural applications due to the unusual physico-chemical properties of these materials (Nriagu, 1979). One of the principal uses of mercury (which accounts for about 20% of the total consumption) has been in the production of caustic soda and chlorine. The chlor-alkali processes make use of mercury in the simultaneous production of chlorine and caustic soda by electrolysis of brine solution using a flowing cathode of metallic mercury, Twenty years ago it was estimated that 200-250 g of mercury could be lost to the environment for every 1000 kg of chlorine produced, In the most advanced chlor-alkali installations losses of mercury to the environment should be less than 3 g per 1000 kg of chlorine (Frei and Hutzinger, 1985). Mercury is released into the environment mainly in waste water (a), in the atmosphere (b) and in sludges (c): (a) Mercury discharged into waste water should not surpass a value of about 1 g per 1000 kg of chlorine produced; (b) Mercury emitted into the atmosphere (especially from electrolytic cell rooms) should be of the order of 1.5-2 g per 1000 kg of chlorine; (c) Mercury in sludges, old cell room parts, etc., is not easy to estimate, In the literature, data are available on mercury discharged into waste water from some chlor-alkali complexes (Renzoni, 1977; E1-Rayis et al., 1986), but data on atmospheric mercury are scarce and refer only
to the area around the complex (Lindqvist et al., 1991). Rather than dealing with direct measurements of atomic mercury concentration in air, many authors are discussing mercury content in vegetative organisms, used as biological indicators of atmospheric mercury pollution (Wallin, 1976; Lodenius and Herranen, 1981; Lodenius and Tulisalo, 1984; Hynninen and Lodenius, 1986; Gonzales, 1991). In the present paper we report a study on mercury emission into the atmosphere from a chlor-alkali complex located in central Italy. The chlor-alkali complex of Rosignano Solvay & Cie (Livorno) produces 120,000 tons y - 1 of chlorine. Consequently, the minimum loss of mercury into the environment should be about 360 kg y - l, if the most modern techniques are applied. From our previous work (Maserti and Ferrara, 1991) mercury discharged to the waste water in the coastal area has been estimated to be of the order of 150-180kgy -1. Few data are available on mercury concentration in the atmosphere around this chlor-alkali complex (Maserti and Ferrara, 1991), while no information is present on mercury emitted directly into the atmosphere by this complex. Research on mercury concentrations in plants and lichens collected in this area reveals that atmospheric background values (3-5 ng m - 3) are reached within a radius of 4-5 km from the emission point with an asymmetry in the spatial distribution due to the presence of prevailing winds. In order to evaluate the mercury concentrations in the atmosphere above the complex and the total flux of the metal emitted, a measurement campaign was performed in 1990 using a lidar system developed at the Lund Institute of Technology (Sweden).
1253
1254
R. FERRARAet al.
This remote-sensing technique allows elemental mercury mapping in the atmosphere at distances ranging up to 1000 m from the lidar position and also an evaluation of the mercury flux from a source if the speed and direction of the wind are known. The first study of atmospheric mercury emissions with the lidar technique was reported by Edner et al. (1989). In addition to measurements carried out with the lidar system, point monitors were used for an accurate evaluation of the mercury levels in the atmospheric layers close to the ground and for a further comparison of the results obtained with the two analytical techniques, as has been reported elsewhere (Ferrara et al., 1992).
EXPERIMENTAL Atmospheric mercury measurements by the lidar technique
The lidar remote-sensing system employed in the present study has been extensively described by Edner et al. (1987, 1989). The system is housed in a van and can work autonomously by means of a 20-kV diesel generator towed by the van. This laser-radar system is capable of generating pulse energies of up to 5 mJ with a linewidth of 0.OO1 nm at the mercury resonance line (253.6 nm), and with a repetition rate of 10 Hz. The laser beam is transmitted into the atmosphere using quartz prisms and a large plane mirror which can be rotated around both the horizontal and the vertical axes. Back-scattered radiation is reflected by the same mirror and collected with a Newtonian telescope to reach the detection system. The laser is tuned on and off the resonance line of mercury every laser shot, allowing differential absorption measurements. A transient recorder performs A / D conversion of the signal with a time resolution of 10 ns, giving a range resolution of about 1 m . The digital signals are averaged on a computer and stored on floppy disks. A single scan is accomplished in a time that is considerably shorter than lOOms; for a given measurement direction a few thousand scans are then averaged during a few minutes of integration time to obtain a satisfactory signal-to-noise ratio,
The detection limit is of the order of 2 ng m- 3 of elemental mercury. The mercury compounds present in the atmosphere cannot be detected with the lidar technique, but these compounds do not represent the dominant fraction, constituting generally only 5-10% of the total atmospheric mercury. Atmospheric mercury measurements by point monitors
The point monitors used by us in atmospheric mercury determinations have been described elsewhere (Ferrara et al., 1992). Gaseous mercury is collected on gold traps at a flow rate of 1 E min- ~ by means of a battery-operated membrane pump, for a period of time set on a programmable timer. Mercury, electrothermally desorbed, is determined by atomic absorption spectroscopy or atomic fluorescence spectroscopy (detection limit 0.01 ng of mercury). Determinations of atmospheric mercury levels carried out with the lidar system and with point monitors during an extensive field study provided comparable results (Ferrara et al., 1992). The speed and direction of the wind were measured by means of a weather station located near the electrolytic cells.
RESULTS AND DISCUSSION A measurement campaign was performed at the Rosignano Solvay chlor-alkali complex from 19 to 21 September 1990. In Fig. I the mobile lidar system is shown against the background of the chlor-alkali complex. A first horizontal screening carried out over the industrial scanning area allowed us to identify the main sources of atmospheric mercury. In Fig. 2 a typical horizontal map of the mercury distribution over the study area is given. D a t a refer to a height of about 5-10 m above the ground and were obtained on 21 September from 4 p.m. to 7 p.m. with a wind speed of 7 m s -1 and a wind direction of 220 °. The highest mercury concentration (up to 903 ng m - 3) was measured over the electrolytic cell rooms
Fig. I. The mobile lidar system against the background of the chlor-alkali complex of Rosignano Solvay (Livorno, Italy).
Mercury emissions measured by a lidar system
1
1255
,
~/ki IHI-'.:.~.
II
.............
Fig. 2. Horizontal map of the mercury distribution above the chlor-alkali complex measuCedby the lidar. Along the given measurement directions, mean mercury concentrations were evaluated as presented in the circles for the integration intervals marked along the lines. The length of the integration intervals were selected depending on the concentration level of mercury. The results of point monitors are included in square boxes. and downwind from this mercury source. A smaller mercury source with a concentration reaching 351 ng m - 3 seems to be present at about 200 m from the iidar van position. Until now the cause of this emission is unclear, even though the material removed from the settling tanks, formerly dumpedin the nearby field, can be suspected, The mercury levels measured in the atmosphere decrease rapidly a few hundred meters from the plant, reaching values quite comparable to the background ones outside the industrial area of the complex, Mercury determinations with point monitors were performed outside the complex. They give comparable results to those obtained with lidar, with the exception of the value (68 ng m-3) measured southwards from the electrolytic cells, probably due to the falling mercury vapour. In Fig. 3 vertical profiles of atmospheric mercury concentrations measured in two directions are plotted; the presence oftwo plumes reveals the existence oftwo distinct mercury sources as stated above. A maximum concentration value greater than 1000 ng m-3 was determined at a height of 10 m in the plume from the cell rooms. An example of a horizontal scan of the atmospheric mercury levels over the same area is given in Fig. 4, where the lidar curve showing the ratio between the on and off resonance line of mercury for
one of the directions is inserted. A horizontal ratio curve is obtained in the absence of mercury while its presence is evinced from the slope of the curve, which is directly related to the atmospheric mercury concentration. In Table 1 the results of atmospheric mercury measurements performed with the lidar system are reported together with the speed and direction of the wind. In the last column of the same table the flux values from the mercury sources are also indicated. Measurements were carried out downwind from the cell-room area, where the two plumes were located. The flux values have been calculated by integrating the mercury concentration over the area of the plume cross-section and then multiplying the integrated content by the component of the wind speed perpendicular to the scan. The two lowest values were observed in a position {C) upwind from the main emission source. Analysing our data we find that the average nighttime value (56gh -~) is higher than the daytime average (36gh 1). From 10p.m. to 6a.m. the chlor-alkali complex increases the chlorine production because of the lower cost of electricity and we observe this fact reflected in a higher mercury emission. It is important to point out that the data reported in Table 1 are the first direct measurements aimed at an evaluation of the mercury flux to the atmosphere of a
1256
R. FERRARAet al.
Fig. 3. Vertical profiles of atmospheric mercury concentrations measured in two directions upwind and downwind from the electrolytic cell room.
Table 1. Day, time, measurement direction, speed and direction of wind and mercury flux in the atmosphere above the chlor-alkali complex of Rosignano Solvay (Italy). The horizontal directions A-C for the vertical scans are as shown in Fig. 2
Day 19 Sept. 20 Sept. 21 Sept. 21 Sept. 21 Sept. 21 Sept. 21 Sept. 21 Sept. 21 Sept. 21 Sept. 21 Sept. 21 Sept.
1990 1990 1990 1990 1990 1990 1990 1990 1990 1990 1990 1990
Time
Measurement direction
Wind speed (m s- 1)
Wind direction
Mercury flux (g h - 1)
6.43 p.m. 11.34 p.m. 12.24 a.m. 12.54 a.m. 01.36 a.m. 02.09 a.m. 12.29 p.m. 01.13 p.m. 01.26 p.m. 05.06 p.m. 06.24 p.m. 06.37 p.m.
A B B B A B A A C A A C
2.52.5 7.5 9.0 9.0 9.0 9.0 8.0 9.0 6.0 6.0 6.0
220 ° 220 ° 220 ° 220 ° 220 ° 220 ° 220 ° 220 ° 220 ° 240 ° 230 ° 250 °
34 40 57 55 65 64 42 42 3 31 31 2
chlor-alkali plant. T h e only d a t u m (30 g h - 1) available in the literature was r e p o r t e d by E d n e r et al. (1989), a n d it was determined at a Swedish chlor-alkali p l a n t d u r i n g a n evaluation of the lidar system's capability for a t m o s p h e r i c mercury mapping.
T h e low flux values (3 a n d 2 g h - 1) observed o n 21 S e p t e m b e r 1990 at 1.26 p.m. a n d 6.37 p.m. u p w i n d from the m a i n cell house refer to a smaller source inside the chlor-alkali plant. U s i n g the collected d a t a a n d assuming they reflect
Mercury emissions measured by a lidar system
1257
Fig. 4. Horizontal scan of atmospheric mercury above the electrolytic cell room. The ratio of on and off resonance lidar curve for one of the directions is inserted.
normal complex activity, an average diurnal mercury flux of 43 g h - 1, corresponding to a mercury amount of 400 kg y-1 emitted into the atmosphere, has been calculated. A careful examination of the vertical profiles of the atmospheric mercury concentration determined with the lidar reveals that a large part of the mercury vapor emitted from the cell rooms falls down close to the ground. Depending on the surrounding buildings it is not always possible to scan from the ground level, From the measured plume cross-sections we have estimated that 20% of emitted mercury goes undetected. Consequently, the mercury amount discharged into the atmosphere from the chlor-alkali complex is of the order of 500 kg y - 1. Taking into account the chlorine production of this complex the mercury emission into the atmosphere is 4 g per 1000 kg of chlorine. This finding shows that the chlor-alkali plant of Rosignano Solvay uses quite good production techniques, even though they can be
further improved by applying the most modern technology. CONCLUSION From the atmospheric mercury determination performed with the lidar above the chlor-alkali plant of Rosignano Solvay two distinct sources were located. The main mercury source is the emission from the electrolytic cells with a flux of 31-65 g h - 1; the other source, of minor importance, presents a flux value of 2-4 g h-1 and is located over old deposits of solid wastes removed from the settling tanks. The distribution of the mercury concentrations in the atmosphere obtained with the lidar evidences the presence of these hot spots and a sharp decrease of the mercury levels reaching background values about 1-2 km from the complex. From the vertical scan of the mercury concentrations in the plume it appears evident that the
1258
R. FERRARAet al.
maximum concentration is reached just above the ventilation outlet of the cell rooms (25 m) and that the
Gonzales H. (1991) Mercury pollution caused by a chloralkali plant. Wat. Air Soil Pollut. 56, 83-93. Hynninen V. and Lodenius M. (1986) Mercury pollution near mercury vapor falling down rapidly results in elevated an industrial source in southwest Finland. Bull. Envir. values in a restricted area at ground level. Contain. Toxicol. 36, 294-298. Lindqvist O., Johansson K., Aastrup M., Andersson A., Bringmark L., Hovsenius G., Hakanson L., Iverfeldt A., Meili M. and Timm B. (1991) Mercury in the Swedish environment Wat. Air Soil Pollut. 55. REFERENCES Lodenius M. and Herranen M. (1981) Influence of a chloralkali plant on the mercury contents of fungi. Chemosphere Edner H., Fredriksson K., Sunesson A., Svanberg S., Uneus 10, 313-318. L. and Wendt W. (1987) Mobile remote sensing system for Lodenius M. and Tulisalo E. (1984) Environmental mercury atmospheric monitoring. Appl. Opt. 26, 4330-4338. contamination around a chlor-alkali plant. Bull. Envir. Edner H., Faris G. W., Sunesson A. and Svanberg S. (1989) Contam. Toxicol. 32, 439-444. Atmospheric atomic mercury monitoring using differential Maserti B. E, and Ferrara R. (1991) Mercury in plants, soil absorption lidar techniques. Appl. Opt. 28, 921-930. and atmosphere near a chlor-alkali complex. Wat. Air Soil El-Rayis O. A., Halim Y. and Aboul-Dahab O. (1986) Total Pollut. 56, 15-20. mercury in the coastal marine ecosystem west of Alexan- Nriagu J. O. (1977) The production and uses of mercury. dria. FAO Fish. Rep. 325 (Suppl.), 58-73. In The Biogeochemistry of Mercury in the Environment Ferrara R., Maserti B. E., Edner H., Ragnarson P., Svanberg (edited by Nriagu J. O.). Elsevier/North Holland BiomedS. and Wallinder E. (1992) Atmospheric mercury deterical Press, Amsterdam, New York, Oxford. minations by Lidar and Point Monitors in environmental Renzoni A. (1977) A case of mercury abatement along the studies. Toxicol. envir. Chem. (in press). Tuscan coast. J. Etud. Pollut. CIESM 2, 95-97. Frei R. W. and Hutzinger O. (eds) (1985) Analytical Aspects of Wallin T. (1976) Deposition of airborne mercury from six Mercury and Other Heavy Metals in the Environment. Swedish chior-alkali plants surveyed by moss analysis. Gordon and Breach, London, New York. Envir. Pollut. 10, 102-114.