- Email: [email protected]
Mercury emissions from coal-fired power stations: The current state of the art in the Netherlands
Science of the Total Environment 368 (2006) 393 – 396 www.elsevier.com/locate/scitotenv
Mercury emissions from coal-fired power stations: The current state of the art in the Netherlands Ruud Meij *, Henk te Winkel KEMA Power Generation and Sustainables, Building M05, P.O. Box 9035, NL-6800 ET Arnhem, The Netherlands Received 16 December 2004; received in revised form 24 August 2005; accepted 26 September 2005 Available online 11 November 2005
Abstract About 30% of the electricity produced in the Netherlands is generated by coal, all of which is imported. At the same time, the co-combustion of biomass is becoming increasingly important. For the last 25 years, the fate of the elements/trace elements in general and of mercury in particular has been studied in great detail. It appears that on average 50% of the mercury is removed in the ESP (particulate control) and 50% of the remainder is removed in the flue gas desulphurization (FGD), resulting in a total mercury removal of 75%. If a high dust selective catalytic reduction (SCR for NOx reduction) is present, the total removal can be up to 90%. The results indicate that on average the removal of mercury during the cocombustion of biomass is similar to that found for full coal-firing. The conclusion is that a modern coal-fired power station with the above-mentioned flue gas cleaning equipment also removes mercury up to 90%. These cleaning devices are being installed to reduce the emission of particulates, sulphur dioxide and nitrogen oxides. This means that mercury abatement can be increased while meeting the EU regulation for SO2 and NOx . The application of Best Available Technique (BAT) for coalfired installations by 1-1-2008 will lead to a further increase in the construction and operation of FGD and DeNOx installations. D 2005 Elsevier B.V. All rights reserved. Keywords: Mercury; Emissions; Coal-fired power stations; Co-combustion; ESP; FGD
1. Introduction Approximately 30% of all the electricity produced in the Netherlands is generated by coal. Only imported bituminous coal from all over the world is fired in pulverised dry-bottom boilers (low NOx -burners, tangential-fired and wall-fired) equipped with high efficiency cold side electrostatic precipitators (ESP) and wet flue-gas desulphurization plants (FGD). Thirty
* Corresponding author. Tel.: +31 26 3390559; fax: +31 26 4454659. E-mail address: [email protected] (R. Meij). 0048-9697/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.scitotenv.2005.09.083
percent of the installed coal-fired capacity is equipped with a DeNOx ; it is expected that in the near future this figure will rise to 100%. At the same time, the cocombustion of biomass is becoming increasingly important. For the last 25 years, the fate of the elements/ trace elements in general and of mercury in particular has been studied in great detail. Mercury enters the power station first and foremost in the fuel and leaves the power station mainly in the ash and flue gases; only a minor part is present in the gypsum, sludge of the waste water treatment plant and effluent. In this paper, an overview is given of the distribution of mercury over the various flows, as measured in the last 25 years in the Netherlands.
394
R. Meij, H. te Winkel / Science of the Total Environment 368 (2006) 393–396
2. Experimental Flue gases for mercury analyses were sampled by different methods (Meij, 1991b). The first method is by using activated carbon. Flue gases were sucked at a flow of 1 l min 1; lead on activated carbon in a cartridge with two compartments, each with a diameter of 10 mm and a length of 50 mm, at a temperature of about 60 8C. Aerosols were separated by a plane Teflon filter. Spherical activated carbon of KUREHA was used (grade: BAC MP, diameter 0.1 to 1 mm). The second method is according to NEN-EN 13506 using bDe GraaffQ washbottles with a content of about 100 ml. Flue gases were sucked at a flow of 1–2 l min 1 through two washbottles containing 4% K2Cr2O7 in 4 M HNO3 or 1.5% KMnO4 in 10% H2SO4. These two wash-bottles were preceded by a wash-bottle containing 2% H2O2 in order to remove SO2. If speciation of mercury was measured, a fourth wash-bottle, filled with HCl, was used as sec-
ond one for sampling Hg2+, the last two bottles are then used for sampling Hg0. A continuously emissions monitoring (CEM) of PSA (Sir Galahad) is additionally used in 1999, 2003 and 2004 (Corns, 1999). The solid samples were digested using HNO3–HF at 180 8C in a pressurised Teflon-lined decomposition vessel under O2 atmosphere according to an USA standard (ASTM-D3684) or by a Dutch standard (NVN 2507). Atomic fluorescence spectroscopy (AFS) was used for chemical analysis. The activated carbon was analysed by Instrumental Neutron-Activation Analyses (INAA). 3. Results and discussion KEMA has measured the mercury content in bituminous steam coal originating from all over the world. The weighted average value for the coal, as fired in Dutch power stations and regardless of variations per
Table 1 Mass balances of mercury around boiler + ESP Test seriesa IV XI XII XIIIb XIV XVI C-XX C-XXVI C-XXX C-XXXI C-XXXII
Power station
Year
AC 8 1981 AC 8 1986 CG 13 1986 AC 8 1987 AC 8 1987 MC 5f 1991 BS-12 1998 CG 13g 1999 MV-1 2001 MV-1 2001 MV-1 2001 MP 1f 2002 2002 MP 3f MP 2f 2002 MP 1f 2002 2002 MP 2f C-XXXIII AC 9 2002 XVII HW-8 2003 C-XXXIXa MV-1 2003 C-XXXIXb MV-1 2003 C-XLIXa AC-8g 2004 C-XLIXb AC-8g 2004 Mean for 20 mass balances Averaged figure as used for model calculations, based on 55 field a b c d e f g h
Origin of coal Australia Eastern USA Eastern USA Australia Colombia Eastern USA Poland Blend Blend Blend Blend Indonesia South Africa Australia Colombia Eastern USA Blend Blend Blend Blend blend blend tests
C means co-combustion. FBA= furnace bottom ash. PFA= pulverized fuel ash or coal ash collected in the ESP. Out/in is outgoing flow of FBA, PFA and flue gases divided by ingoing flow of fuel. Division normalized to 100%. High-temperature of ESP, N140 8C. DeNOx (SCR). Coal analyses adopted.
FBAb
PFAc
Flue gases
Out/Ind
8 41 10 43 27 6 54 77 61 51 59 48 60 32 32 13 30 27 87 92 48 34 43 49
91 59 90 57 73 94 46 21 39 48 39 51 39 67 67 85 70 73 13 8 50 46 56 50
0.8 0.7 0.9 1.2 0.8 1.2 1.3 0.8 1.0 1.0 0.6 0.7 0.9 0.6 0.7 0.9 0.8 1.0 0.8 1.2 1.0h 1.0h 0.9 F 0.2 1.00
e
Share (%) 1 b1 b1 b1 b1 0.2 b1 b2 0.3 0.5 1.1 b1 b1 b1 b1 b1 b0.4 0.02 0.3 0.2 1.4 20.5 b1 1
R. Meij, H. te Winkel / Science of the Total Environment 368 (2006) 393–396
year of origin, has been constant for the last 10 years at a very low level of 0.1 mg kg 1 (see Meij, 1991a; Meij et al., 2002). Much effort has been put into developing a reliable sampling system for gaseous mercury in flue gases and the ambient air (Meij, 1991b). Proof of the accurate figures of measurements in fuel, ashes and flue gases is given by mass balances. The results of such mass balances around boiler + ESP are given in Table 1, and those around FGD + Waste Water Treatment Plant (WWTP) in Table 2. It appears that in general the closure of the mass balances around ESP + boiler and around the FGD in general is very good, considering the low concentrations in the flows. The results from 55 field tests show that the behaviour of mercury in the boiler and ESP is affected by several factors. First, the temperature of the ESP is an important parameter. Second, the chlorine content of the fuel appears to have a profound effect on the speciation of mercury, which in turn greatly affects the emissions (Meij, 1991a). High calcium levels reduce the effect of chlorine, whereas sulphur reduces the available calcium. However, calcium compounds can also promote the adsorption of Hg2Cl2 on fly ash particles (Gullit et al., 1999), thus increasing the removal in the ESP. Because in the Netherlands steam coal from all over the world is fired, it has been possible to study the influence of coal composition. Mercury chloride can in fact be removed better than metallic mercury. An SCR type DeNOx also affects the speciation of mercury, as a result of which removal increases (campaign C-XXVI). However, extreme low chlorine contents in the fuel (0.006% and 0.004%) explains why, although an SCR type of DENOx is present in campaign C-XLIX, the
395
mercury removal is at an average level. Also unburned carbon has a positive effect on the removal of mercury. These factors are discussed in more detail in Meij and Te Winkel (2003). Making mass balances for an FGD is more difficult because the long residence time of the scrubber liquid demands constant conditions. However, the quantities of the flows in the FGD are lower than those around the boiler, so that the total mass balances usually close (see Table 2). A statistical survey of 25 years of measuring mercury at Dutch coal-fired power stations is presented in Table 3. Only relative figures in the form of removal percentages in the ESP, FGD and the total of the flue gas cleaning equipment are provided. Roughly 50% of the mercury is removed in the ESP and the remaining 50% is removed for 50% in the wet FGD, resulting in a total removal of 75%. A distinction is made between full coal-firing and co-combustion. It appears that in general, co-combustion does not influence the behaviour of mercury. Mercury removals for bituminous coal are found in the US EPA’s Mercury Information Collection Request (ICR) and are 35% in ESPs and 70% for a combination of ESP + wet FGD (Wayland, 2001). Recent measurements in the US show that the combination of SCR + ESP + wet FGD gives rise to mercury removals between 84% and 89% (Withum et al., 2005). In Japan, mercury removals are reported for ESP of 33% and the combination of ESP and wet FGD of 70% (Yokoyama, 1999). These mercury removal rates in the USA and Japan are about similar to the Dutch findings, except for the ESP. The lower flue gas temperature in the Netherlands (~ 120 8C) with respect to Japan and the USA
Table 2 Mass balances of mercury around FGD + WWTP together with the total mass balance Test series
Power station
Year
XII CG 13 1986 C-XX BS-12 1998 C-XXVI CG 13 1999 C-XXVIII AC 9 2002 XVII HW-8 2003 C-XXXIX MV-1 2003 C-XXXIX MV-1 2003 C-XLIXa AC-8 2004 C-XLIXb AC-8 2004 Mean Averaged figure as used for model calculations a
n.m. = not measured.
Shares in %, normalized to 100% In
Out/In Out
Process water
Flue gases
Lime stone
Flue gases
Gypsum
Sludge
Effluent
2 n.m. b0.2 0.1 0.02 0.1 0.2 0.01 0.01 0.3 0.5
66 100 85 99 97 98 98 97 98 93 99
b31 n.m. 14 b1 3 2 2 2 2 4 0.5
23 88 16 49 17 47 36 43 38 40 50
b62 9 75 37 70 37 41 50 56 47 33
9 4 8 14 12 16 23 7 6 11 17
6 0.02 0.20 0.00 0.17 0.05 0.12 0.01 0.00 0.7 0.1
FGD
Total
0.9 0.4 0.3 1.1 1.3 1.8 2.0 1.2 1.2 1.1 1.0
n.m.a 0.9 1.1 0.9 1.2 0.9 1.3 1.1 1.1 1.1 1.0
396
R. Meij, H. te Winkel / Science of the Total Environment 368 (2006) 393–396
Table 3 Statistical survey of mercury removal in Dutch coal-fired power stations
Mean S.D. V (%) Min Max n
Removal in ESP (%)
Removal in FGD (%)
Total removal (%)
100% coal
Co-combustion
100% coal
Co-combustion
100% coal
Co-combustion
59 27 46 0 99 41
58 23 40 23 89 9
48 23 49 8 72 7
51 12 24 28 64 8
78 8 11 57 85 10
78 11 15 50 92 18
S.D. = standard deviation, V = coefficient of variation or relative standard deviation, min = lowest value, max = highest value and n = number of measurements.
(z 140 8C) explains why more mercury condenses on the fly ash particles. The mercury concentrations in the coal fly ash are still low (b 0.5 ppm) and do not posses an environmental hazard. It should be noted that the removal of Hg in a wet FGD cannot be considered independently, because the process is influenced by other flue gas cleaning systems. If HgCl2 has already been effectively removed in an ESP, the removal in the following FGD is lower. In the end, their combined removal will have improved (see Table 3). 4. Conclusions Extensive measurements at Dutch coal-fired power stations over a period of 25 years indicate that on average the following percentages of mercury are removed when these flue gas cleaning devices are present: a) cold side ESP 50% b) cold side ESP + wet FGD 75% c) cold side ESP + wet FGD + DENOx (SCR) up to 90% These cleaning devices are being installed to reduce the emission of particulates, sulphur dioxide and nitrogen oxides. This means that mercury abatement can be increased while meeting the EU regulation for SO2 and NOx . The application of BAT for coal-fired installations by 1-1-2008 will lead to a further increase in the construction and operation of FGD and DeNOx installations. Acknowledgements The projects described in this paper were carried out by KEMA within the framework of the Technical Service Agreement between it and the Netherlands’ six
electricity generating companies, namely Delta, Electrabel Nederland, E.ON Benelux, EPZ, Essent Energie Productie and Nuon Power Generation. References Corns W. The development of a continuous emission monitor for mercury speciation and its utilization in mercury removal studies. Presented at the conference on air quality II: mercury, trace elements, and particulate matter, McLean, VA, Sept 19–21, 1999. Paper; 1999, vol. A3-1. Gullit BK, Ghorishi B, Huggins FE. Mercury chloride capture by alkaline sorbents. In the Proceedings of the Air and Waste Management Association 93rd Annual Meeting in Salt Lake City, UT on June 18–22, 1999. Paper 259; 1999. Meij R. The fate of mercury in coal-fired power plants and the influence of wet flue-gas desulphurisation. Water Air Soil Pollut 1991a;(56):21 – 33. Meij R. A sampling method based on activated carbon for sampling gaseous mercury in ambient air and flue gases. Water Air Soil Pollut 1991b;56:117 – 29. Meij R, Te Winkel BH. Mercury emissions of coal-fired power stations, firing bituminous coal, as a function of plant parameters, coal composition and co-combustion. Proceedings of the International Conference on Air Quality IV (Mercury, Trace Elements and Particulate Matter), 22–24 September 2003, Arlington, VA, USA; 2003. Meij R, Vredenbregt LHJ, Te Winkel BH. The fate of mercury in coalfired power plants. J Air Waste Manage Assoc 2002;52:912 – 7. Wayland RJ. Mercury and utilities: current control technologies. Presented at the Northeast Midwest Institute/ECOS Meeting in Washington, DC, July 31 2001; 2001. Withum JA, Tseng SC, Locke JE. Mercury sampling at Power Plants with SCR-FGD Combinations. Presented at the MEC2 Workshop Mercury Emissions from Coal in Ottawa, Canada, May 24–25, 2005; 2005. Yokoyama T. Trace element emissions from coal-fired power plants in Japan. Legislation, emissions, environmental control technologies. Proceedings of CEM 99, International Conference on Emissions Monitoring, held at the University of Warwick, Coventry, UK, September 6–8, 1999. The paper was presented at the TraceElements Workshop organised by IEA on 9 September 1999; 1999.
Mercury emissions from coal-fired power stations: The current state of the art in the Netherlands Ruud Meij *, Henk te Winkel KEMA Power Generation and Sustainables, Building M05, P.O. Box 9035, NL-6800 ET Arnhem, The Netherlands Received 16 December 2004; received in revised form 24 August 2005; accepted 26 September 2005 Available online 11 November 2005
Abstract About 30% of the electricity produced in the Netherlands is generated by coal, all of which is imported. At the same time, the co-combustion of biomass is becoming increasingly important. For the last 25 years, the fate of the elements/trace elements in general and of mercury in particular has been studied in great detail. It appears that on average 50% of the mercury is removed in the ESP (particulate control) and 50% of the remainder is removed in the flue gas desulphurization (FGD), resulting in a total mercury removal of 75%. If a high dust selective catalytic reduction (SCR for NOx reduction) is present, the total removal can be up to 90%. The results indicate that on average the removal of mercury during the cocombustion of biomass is similar to that found for full coal-firing. The conclusion is that a modern coal-fired power station with the above-mentioned flue gas cleaning equipment also removes mercury up to 90%. These cleaning devices are being installed to reduce the emission of particulates, sulphur dioxide and nitrogen oxides. This means that mercury abatement can be increased while meeting the EU regulation for SO2 and NOx . The application of Best Available Technique (BAT) for coalfired installations by 1-1-2008 will lead to a further increase in the construction and operation of FGD and DeNOx installations. D 2005 Elsevier B.V. All rights reserved. Keywords: Mercury; Emissions; Coal-fired power stations; Co-combustion; ESP; FGD
1. Introduction Approximately 30% of all the electricity produced in the Netherlands is generated by coal. Only imported bituminous coal from all over the world is fired in pulverised dry-bottom boilers (low NOx -burners, tangential-fired and wall-fired) equipped with high efficiency cold side electrostatic precipitators (ESP) and wet flue-gas desulphurization plants (FGD). Thirty
* Corresponding author. Tel.: +31 26 3390559; fax: +31 26 4454659. E-mail address: [email protected] (R. Meij). 0048-9697/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.scitotenv.2005.09.083
percent of the installed coal-fired capacity is equipped with a DeNOx ; it is expected that in the near future this figure will rise to 100%. At the same time, the cocombustion of biomass is becoming increasingly important. For the last 25 years, the fate of the elements/ trace elements in general and of mercury in particular has been studied in great detail. Mercury enters the power station first and foremost in the fuel and leaves the power station mainly in the ash and flue gases; only a minor part is present in the gypsum, sludge of the waste water treatment plant and effluent. In this paper, an overview is given of the distribution of mercury over the various flows, as measured in the last 25 years in the Netherlands.
394
R. Meij, H. te Winkel / Science of the Total Environment 368 (2006) 393–396
2. Experimental Flue gases for mercury analyses were sampled by different methods (Meij, 1991b). The first method is by using activated carbon. Flue gases were sucked at a flow of 1 l min 1; lead on activated carbon in a cartridge with two compartments, each with a diameter of 10 mm and a length of 50 mm, at a temperature of about 60 8C. Aerosols were separated by a plane Teflon filter. Spherical activated carbon of KUREHA was used (grade: BAC MP, diameter 0.1 to 1 mm). The second method is according to NEN-EN 13506 using bDe GraaffQ washbottles with a content of about 100 ml. Flue gases were sucked at a flow of 1–2 l min 1 through two washbottles containing 4% K2Cr2O7 in 4 M HNO3 or 1.5% KMnO4 in 10% H2SO4. These two wash-bottles were preceded by a wash-bottle containing 2% H2O2 in order to remove SO2. If speciation of mercury was measured, a fourth wash-bottle, filled with HCl, was used as sec-
ond one for sampling Hg2+, the last two bottles are then used for sampling Hg0. A continuously emissions monitoring (CEM) of PSA (Sir Galahad) is additionally used in 1999, 2003 and 2004 (Corns, 1999). The solid samples were digested using HNO3–HF at 180 8C in a pressurised Teflon-lined decomposition vessel under O2 atmosphere according to an USA standard (ASTM-D3684) or by a Dutch standard (NVN 2507). Atomic fluorescence spectroscopy (AFS) was used for chemical analysis. The activated carbon was analysed by Instrumental Neutron-Activation Analyses (INAA). 3. Results and discussion KEMA has measured the mercury content in bituminous steam coal originating from all over the world. The weighted average value for the coal, as fired in Dutch power stations and regardless of variations per
Table 1 Mass balances of mercury around boiler + ESP Test seriesa IV XI XII XIIIb XIV XVI C-XX C-XXVI C-XXX C-XXXI C-XXXII
Power station
Year
AC 8 1981 AC 8 1986 CG 13 1986 AC 8 1987 AC 8 1987 MC 5f 1991 BS-12 1998 CG 13g 1999 MV-1 2001 MV-1 2001 MV-1 2001 MP 1f 2002 2002 MP 3f MP 2f 2002 MP 1f 2002 2002 MP 2f C-XXXIII AC 9 2002 XVII HW-8 2003 C-XXXIXa MV-1 2003 C-XXXIXb MV-1 2003 C-XLIXa AC-8g 2004 C-XLIXb AC-8g 2004 Mean for 20 mass balances Averaged figure as used for model calculations, based on 55 field a b c d e f g h
Origin of coal Australia Eastern USA Eastern USA Australia Colombia Eastern USA Poland Blend Blend Blend Blend Indonesia South Africa Australia Colombia Eastern USA Blend Blend Blend Blend blend blend tests
C means co-combustion. FBA= furnace bottom ash. PFA= pulverized fuel ash or coal ash collected in the ESP. Out/in is outgoing flow of FBA, PFA and flue gases divided by ingoing flow of fuel. Division normalized to 100%. High-temperature of ESP, N140 8C. DeNOx (SCR). Coal analyses adopted.
FBAb
PFAc
Flue gases
Out/Ind
8 41 10 43 27 6 54 77 61 51 59 48 60 32 32 13 30 27 87 92 48 34 43 49
91 59 90 57 73 94 46 21 39 48 39 51 39 67 67 85 70 73 13 8 50 46 56 50
0.8 0.7 0.9 1.2 0.8 1.2 1.3 0.8 1.0 1.0 0.6 0.7 0.9 0.6 0.7 0.9 0.8 1.0 0.8 1.2 1.0h 1.0h 0.9 F 0.2 1.00
e
Share (%) 1 b1 b1 b1 b1 0.2 b1 b2 0.3 0.5 1.1 b1 b1 b1 b1 b1 b0.4 0.02 0.3 0.2 1.4 20.5 b1 1
R. Meij, H. te Winkel / Science of the Total Environment 368 (2006) 393–396
year of origin, has been constant for the last 10 years at a very low level of 0.1 mg kg 1 (see Meij, 1991a; Meij et al., 2002). Much effort has been put into developing a reliable sampling system for gaseous mercury in flue gases and the ambient air (Meij, 1991b). Proof of the accurate figures of measurements in fuel, ashes and flue gases is given by mass balances. The results of such mass balances around boiler + ESP are given in Table 1, and those around FGD + Waste Water Treatment Plant (WWTP) in Table 2. It appears that in general the closure of the mass balances around ESP + boiler and around the FGD in general is very good, considering the low concentrations in the flows. The results from 55 field tests show that the behaviour of mercury in the boiler and ESP is affected by several factors. First, the temperature of the ESP is an important parameter. Second, the chlorine content of the fuel appears to have a profound effect on the speciation of mercury, which in turn greatly affects the emissions (Meij, 1991a). High calcium levels reduce the effect of chlorine, whereas sulphur reduces the available calcium. However, calcium compounds can also promote the adsorption of Hg2Cl2 on fly ash particles (Gullit et al., 1999), thus increasing the removal in the ESP. Because in the Netherlands steam coal from all over the world is fired, it has been possible to study the influence of coal composition. Mercury chloride can in fact be removed better than metallic mercury. An SCR type DeNOx also affects the speciation of mercury, as a result of which removal increases (campaign C-XXVI). However, extreme low chlorine contents in the fuel (0.006% and 0.004%) explains why, although an SCR type of DENOx is present in campaign C-XLIX, the
395
mercury removal is at an average level. Also unburned carbon has a positive effect on the removal of mercury. These factors are discussed in more detail in Meij and Te Winkel (2003). Making mass balances for an FGD is more difficult because the long residence time of the scrubber liquid demands constant conditions. However, the quantities of the flows in the FGD are lower than those around the boiler, so that the total mass balances usually close (see Table 2). A statistical survey of 25 years of measuring mercury at Dutch coal-fired power stations is presented in Table 3. Only relative figures in the form of removal percentages in the ESP, FGD and the total of the flue gas cleaning equipment are provided. Roughly 50% of the mercury is removed in the ESP and the remaining 50% is removed for 50% in the wet FGD, resulting in a total removal of 75%. A distinction is made between full coal-firing and co-combustion. It appears that in general, co-combustion does not influence the behaviour of mercury. Mercury removals for bituminous coal are found in the US EPA’s Mercury Information Collection Request (ICR) and are 35% in ESPs and 70% for a combination of ESP + wet FGD (Wayland, 2001). Recent measurements in the US show that the combination of SCR + ESP + wet FGD gives rise to mercury removals between 84% and 89% (Withum et al., 2005). In Japan, mercury removals are reported for ESP of 33% and the combination of ESP and wet FGD of 70% (Yokoyama, 1999). These mercury removal rates in the USA and Japan are about similar to the Dutch findings, except for the ESP. The lower flue gas temperature in the Netherlands (~ 120 8C) with respect to Japan and the USA
Table 2 Mass balances of mercury around FGD + WWTP together with the total mass balance Test series
Power station
Year
XII CG 13 1986 C-XX BS-12 1998 C-XXVI CG 13 1999 C-XXVIII AC 9 2002 XVII HW-8 2003 C-XXXIX MV-1 2003 C-XXXIX MV-1 2003 C-XLIXa AC-8 2004 C-XLIXb AC-8 2004 Mean Averaged figure as used for model calculations a
n.m. = not measured.
Shares in %, normalized to 100% In
Out/In Out
Process water
Flue gases
Lime stone
Flue gases
Gypsum
Sludge
Effluent
2 n.m. b0.2 0.1 0.02 0.1 0.2 0.01 0.01 0.3 0.5
66 100 85 99 97 98 98 97 98 93 99
b31 n.m. 14 b1 3 2 2 2 2 4 0.5
23 88 16 49 17 47 36 43 38 40 50
b62 9 75 37 70 37 41 50 56 47 33
9 4 8 14 12 16 23 7 6 11 17
6 0.02 0.20 0.00 0.17 0.05 0.12 0.01 0.00 0.7 0.1
FGD
Total
0.9 0.4 0.3 1.1 1.3 1.8 2.0 1.2 1.2 1.1 1.0
n.m.a 0.9 1.1 0.9 1.2 0.9 1.3 1.1 1.1 1.1 1.0
396
R. Meij, H. te Winkel / Science of the Total Environment 368 (2006) 393–396
Table 3 Statistical survey of mercury removal in Dutch coal-fired power stations
Mean S.D. V (%) Min Max n
Removal in ESP (%)
Removal in FGD (%)
Total removal (%)
100% coal
Co-combustion
100% coal
Co-combustion
100% coal
Co-combustion
59 27 46 0 99 41
58 23 40 23 89 9
48 23 49 8 72 7
51 12 24 28 64 8
78 8 11 57 85 10
78 11 15 50 92 18
S.D. = standard deviation, V = coefficient of variation or relative standard deviation, min = lowest value, max = highest value and n = number of measurements.
(z 140 8C) explains why more mercury condenses on the fly ash particles. The mercury concentrations in the coal fly ash are still low (b 0.5 ppm) and do not posses an environmental hazard. It should be noted that the removal of Hg in a wet FGD cannot be considered independently, because the process is influenced by other flue gas cleaning systems. If HgCl2 has already been effectively removed in an ESP, the removal in the following FGD is lower. In the end, their combined removal will have improved (see Table 3). 4. Conclusions Extensive measurements at Dutch coal-fired power stations over a period of 25 years indicate that on average the following percentages of mercury are removed when these flue gas cleaning devices are present: a) cold side ESP 50% b) cold side ESP + wet FGD 75% c) cold side ESP + wet FGD + DENOx (SCR) up to 90% These cleaning devices are being installed to reduce the emission of particulates, sulphur dioxide and nitrogen oxides. This means that mercury abatement can be increased while meeting the EU regulation for SO2 and NOx . The application of BAT for coal-fired installations by 1-1-2008 will lead to a further increase in the construction and operation of FGD and DeNOx installations. Acknowledgements The projects described in this paper were carried out by KEMA within the framework of the Technical Service Agreement between it and the Netherlands’ six
electricity generating companies, namely Delta, Electrabel Nederland, E.ON Benelux, EPZ, Essent Energie Productie and Nuon Power Generation. References Corns W. The development of a continuous emission monitor for mercury speciation and its utilization in mercury removal studies. Presented at the conference on air quality II: mercury, trace elements, and particulate matter, McLean, VA, Sept 19–21, 1999. Paper; 1999, vol. A3-1. Gullit BK, Ghorishi B, Huggins FE. Mercury chloride capture by alkaline sorbents. In the Proceedings of the Air and Waste Management Association 93rd Annual Meeting in Salt Lake City, UT on June 18–22, 1999. Paper 259; 1999. Meij R. The fate of mercury in coal-fired power plants and the influence of wet flue-gas desulphurisation. Water Air Soil Pollut 1991a;(56):21 – 33. Meij R. A sampling method based on activated carbon for sampling gaseous mercury in ambient air and flue gases. Water Air Soil Pollut 1991b;56:117 – 29. Meij R, Te Winkel BH. Mercury emissions of coal-fired power stations, firing bituminous coal, as a function of plant parameters, coal composition and co-combustion. Proceedings of the International Conference on Air Quality IV (Mercury, Trace Elements and Particulate Matter), 22–24 September 2003, Arlington, VA, USA; 2003. Meij R, Vredenbregt LHJ, Te Winkel BH. The fate of mercury in coalfired power plants. J Air Waste Manage Assoc 2002;52:912 – 7. Wayland RJ. Mercury and utilities: current control technologies. Presented at the Northeast Midwest Institute/ECOS Meeting in Washington, DC, July 31 2001; 2001. Withum JA, Tseng SC, Locke JE. Mercury sampling at Power Plants with SCR-FGD Combinations. Presented at the MEC2 Workshop Mercury Emissions from Coal in Ottawa, Canada, May 24–25, 2005; 2005. Yokoyama T. Trace element emissions from coal-fired power plants in Japan. Legislation, emissions, environmental control technologies. Proceedings of CEM 99, International Conference on Emissions Monitoring, held at the University of Warwick, Coventry, UK, September 6–8, 1999. The paper was presented at the TraceElements Workshop organised by IEA on 9 September 1999; 1999.