EPA/600/A-96/053
Control of Mercury Emissions from Coal Combustors
S.V. Krishnan, Acurex Environmental Corporation, P.O. Box 13109, Research Triangle Park, NC
27709
Brian K. Gullett*, National Risk Management Research Laboratory, Air Pollution Prevention and
Control Division, Environmental Protection Agency, Research Triangle Park, NC 27711
Wojciech Jozewicz, Acurex Environmental Corporation, P.O. Box 13109, Research Triangle Park,
NC 27709
ABSTRACT
Injection of activated carbon sorbents has been shown in laboratory and field situations to
control total mercury (Hg) emissions from municipal waste combustors (MWCs). Efforts are
currently underway to extend this experience to control emissions of Hg from coal-based power
units. However, to achieve high removal of Hg (approximately 90 percent) from flue gas, the
activated carbon to Hg ratio (by weight) has been found to be significantly higher in coal-based
units than in MWCs. In order to optimize Hg removal for coal-based units, we performed bench-
scale experiments to study capture of three species of Hg: elemental (Hg°), mercuric chloride
(HgCl2), and mercuric oxide (HgO) by different activated carbon sorbents at temperatures (100
and 140 °C) and Hg concentrations (-1 ppb Hg) representative of applications in coal
combustors. This paper also reports a comparison of these results with our earlier investigations
of Hg control under MWC conditions.
INTRODUCTION
The Environmental Protection Agency (EPA) "mercury" report to Congress1 cites the largest
emitters of mercury (Hg) as coal-fired utilities, medical waste incinerators (MWIs), municipal
waste combustors (MWCs), chlor-alkali plants, copper and lead smelters, and cement
manufacturers. These sources account for 90 to 100 percent of anthropogenic Hg emissions. All
the sources listed above, except the coal-fired utilities, are subject to intended, proposed, or
existing regulations. The EPA "mercury" report suggests the likelihood of utility boilers' being
listed under Section 112 of the Clean Air Act Amendments (CAAA)2. This will enable EPA to
identify Hg emitting utility boilers and promulgate technology-based pollution control
requirements.
There are several methods of Hg control either being used or tested in waste incineration plants
(MWIs and MWCs). These include a variety of chemical, adsorption, and absorption
techniques3. Primarily, dry sorbent injection (DSI) followed by a fabric filter (FF), a spray dryer
(SD) followed by a FF, and wet scrubbing have been tested to determine Hg control. A database
is available4,s-6 on the levels of Hg control achieved in field tests employing these techniques.
-------
Apart from field tests, laboratory work has also been conducted by investigators
7,8,9,10 to
study
Hg capture by solid sorbents in MWCs. All tests show that existing flue gas desulfurization units
(SD/FF and wet scrubber systems) with the addition of DSI lead to consistently high (> 90
percent) Hg removal in MWCs.
The concentration of total Hg found in the flue gas of a coal-fired utility is typically at least an
order of magnitude lower than that found in a MWC flue gas. Moreover, the flue gas
composition [in particular, hydrogen chloride (HC1) concentration] of a coal-fired unit differs
from that of a MWC unit which may lead to the existence of different proportions of the species
of Hg [that is, the fractions of elemental mercury (Hg°), mercuric chloride (HgCl2), and mercuric
oxide (HgO) — the three species that exist at equilibrium10 in the flue gas]. Because of these
factors — mainly concentration and speciation of Hg — there is uncertainty in extending the
technology for MWC Hg capture to Hg control in coal-fired utilities.
Existing flue gas cleaning devices - particulate and sulfur dioxide (S02) removal - have shown
varying levels of Hg control11"19. Hg control achieved through various particulate and flue gas
desulfurization devices is summarized in Table 1. To resolve uncertainties regarding emissions,
the Pittsburgh Energy Technology Center (PETC) of the Department of Energy (DOE) has
initiated a program to characterize toxic emissions from coal-fired electric utilities. Table 2 lists
the capture of Hg in four of the eight power stations that are to be assessed in the DOE program.
The results shown in Tables 1 and 2 suggest the need for additional control processes in coal-
fired utilities to consistently achieve high (> 90 percent) Hg removal.
Duct injection of activated carbon has been proposed as an additional process for achieving high
levels of Hg removal in coal-fired utilities. This relatively simple add-on technology has been
tested by investigators1 1,16,20 in pilot-scale units.
In the pilot-plant studies of Felsvang et al.16, injection of activated carbon achieved Hg removals
greater than 90 percent. The high removals reported by Felsvang were for each of two types of
coals. However, to achieve 90 percent Hg removal by injection of activated carbon in each case,
the required amounts of activated carbon differed by a factor of five. This was thought to be due
to differences in Hg speciation in the flue gas. This conclusion would be contrary to the tests of
Chang et al.11 which showed no dependence of Hg removal on Hg species.
Chang et al.'s11 slipstream study concluded that injection of activated carbon before a FF is
capable of removing Hg. The results suggested a dependence on the flue gas temperature and the
amount of activated carbon used in determining the fraction of Hg removed. For temperatures
around 120 °C, they reported a carbon: Hg weight ratio of 3,000:1 for high Hg removal. Pilot
plant studies conducted at the University of North Dakota's Energy and Environmental Research
Center (UNDEERC)12 employing different coals, however, showed lower Hg removal (60
percent) with activated carbon injection at a similar carbon:Hg weight ratio.
At a carbon:Hg ratio of 3000:1 and an activated carbon cost of $ 1.125/kg, our estimates indicate
that the material cost would be $ 500,000 per year for a 500 MW power plant. Chang et al.11
report an annual cost of carbon injection of $ 100,000 - $ 1 million for Hg control in a 500 MW
-------
plant. These estimates can vary depending on equipment, installation, and disposal costs. The
significant additional costs associated with Hg removal in coal-fired utilities emphasize the need
for process optimization.
The factors found to influence Hg removal in the field studies were coal type (leading to different
Hg speciation), activated carbon injection rate or method (or both) and type, and flue gas
temperature. Bench-scale studies are convenient in order to study the individual effects of these
parameters on Hg capture in a controlled environment. Most of the bench-scale efforts have been
directed at studying Hg control in MWC situations (that is, relatively higher concentration of
Hg)7"10. Livengood et al.21 have studied Hg capture in bench-scale absorbers in nitrogen (N2)/air,
but, their studies were limited to Hg°. Moreover, the Hg° concentration in their experiments was
approximately 5-11 ppb, compared to approximately 1 ppb total Hg for most coal-fired utilities.
Morency22 studied bench-scale Hg capture in simulated flue gas to quantify Hg° and HgCl2
capture by activated carbons. Here, too, the inlet Hg concentration was much higher (roughly 13
ppb) than found typically in coal flue gas. Hg capture observed at the higher concentrations may
not be the same at lower Hg concentrations due to the non-linear effect of concentration on mass
transport and sorption kinetics. Also, the HgCl2 in Morency's study was produced by reacting Hg°
with HC1 which produced very small quantities of HgCl2(5-10 percent of total). This may lead to
errors in quantifying the separate capture of each species of Hg by activated carbons.
In this paper, we have studied the capture of low concentration Hg° (less than 2 ppb) and HgCl2
(approximately 1 ppb) by two types of activated carbons. The experiments were conducted with
only a single Hg species present in the gas stream in order to obtain information on the capture of
each individual species of Hg. The capture of the two Hg species was studied at two temperatures
— 100 and 140 °C. In addition, capture of HgO by the two activated carbons and hydrated lime
[Ca(OH)2] has been studied in this investigation. Finally, a comparison is made between low
concentration Hg control (coal-fired utility) and high concentration Hg control (MWC situation)
results obtained in our earlier studies7"10.
SORBENTS
Two types of activated carbons, namely PC-100 and FGD (manufactured by American Norit
Company, Inc.), were used in this study as in previous field studies23. PC-100 is bituminous-
coal-based activated carbon, and FGD is derived from lignite. Both the carbons are thermally
activated. PC-100 has a BET surface area of 964 m2/g compared to 547 m2/g for FGD. The
mass median diameters (obtained by gravity sedimentation) for PC-100 and FGD were 17.55 and
6.80 |im, respectively. The details of their physical characteristics, including the pore size
distributions, are given in an earlier study8. Apart from the two activated carbons, reagent grade
Ca(OH)2 was used in this study to compare its capture of HgO with the two activated carbons.
The Ca(OH)2 sample has a BET surface area of 13 m2/g.
EXPERIMENTAL PROCEDURES
Hg°
-------
Figure 1 shows the schematic of the bench-scale apparatus used to study capture of Hg° by PC-
100 and FGD. The apparatus has been used to study Hg° capture by sorbents in earlier studies7.
Unlike the previous studies, the concentration of Hg° used in this study (~ 1 ppb) was very near
the detection limit of the online Hg analyzer (Ametek, Model 400). The Hg analyzer was used
to provide only an estimate of the Hg° concentration in the gas stream. Since online analysis of
Hg° was not possible, the samples were directly analyzed for total Hg. Therefore, all experiments
were conducted in batches (differing in time of exposure of sorbent to Hg°).
The reactor (see Figure 1) was typically loaded with 100 mg of fresh test carbon in stage 1
followed by two stages of fresh PC-100 (100 mg each) as breakthrough traps (stages 2 and 3).
The three stages [i.e., the test (either PC-100 or FGD) and the two sections of PC-100] were
separated by glass wool.
After establishing a Hg° baseline (based on a reading from the online analyzer), the gas stream
was passed through the reactor where the activated carbons (maintained at a set temperature)
capture Hg° for 8 to 24 h as indicated in Figure 3. The experimental procedure (i.e., positioning
of valves) has been described earlier7,8. At the end of the fixed duration, the Hg° stream was
diverted away from the activated carbons. The three stages were then separately analyzed using
x-ray fluorescence (XRF) for total Hg. Two tests were run for a single datum.
The percentage Hg° capture by PC-100 or FGD after exposure to Hg° for time t, 4>t was
calculated as:
-------
XRF was employed to quantitatively determine total Hg captured in each of the three stages.
Standards were prepared for each sorbent type [the two activated carbons and Ca(OH)2] by
doping with known concentrations of Hg solution. The XRF machine (Siemens, Model SRS303)
was calibrated for each sorbent in the 0-9 and 0-50 ppm range using these standards. A linear
correlation for each sorbent was obtained in both ranges. The strength of the XRF signals
obtained from the analysis of sorbents from our experiments was matched with these correlations
to obtain the Hg species concentration in the solid samples.
The accuracy of XRF calibration was verified by comparing with the results of the online
analyzer (for Hg°). Here, a single stage of PC-100 (100 mg) was loaded in the reactor and
exposed to 30 ppb of Hg°. Integration of the signal from the online analyzer gave an estimate of
the Hg in PC-100. The sample was then unloaded from the reactor and analyzed employing
XRF. A very close match was seen between the two methods of analysis, indicating the
reliability of the XRF technique to measure Hg in samples. Periodically, samples used for
calibrating the XRF machine were also analyzed through cold vapor atomic absorption (CVAA)
as a further measure of data quality. Results from the CVAA analysis showed that the samples
used for calibrating the XRF machine were within 10 percent of the XRF value.
RESULTS AND DISCUSSION
Hg° Capture
Figure 3 shows the capture of 2 ppb Hg° (in N2 gas) by 100 mg of PC-100 and FGD at 100 and
140 °C. The face velocity is 5.5 m/min (18 ft/min) and the bed depth is roughly 0.3 cm.
It can be seen from Figure 3 that PC-100 captures a higher percentage of incoming Hg° compared
to FGD at both temperatures. PC-100 captures 80-90 percent of incoming Hg° up to an exposure
time of 24 h. On the other hand, FGD captures 30-50 percent of incoming Hg° at 100 aC and 10-
30 percent of incoming Hg° at 140 °C. Qualitatively, similar results were obtained when PC-100
and FGD were exposed to 30 ppb Hg° during MWC studies9.
The capture of Hg° by PC-100 does not seem to be affected by temperature; whereas, the capture
of Hg° by FGD is lower at 140 than at 100 °C. A similar trend in Hg° sorption with temperature
was seen for FGD when exposed to 30 ppb Hg° in N29. This is perhaps indicative of a
physisorpti ve mechanism of Hg° capture by FGD.
Despite the long exposure periods (24 h), the amount of Hg° captured per gram of PC-100 or
FGD is far lower than in our earlier study9. For example, after 24 h of exposure, the amount of
Hg° (mg) per gram of PC-100 is 0.06 (in this study) compared to 0.2 mg/g of PC-1009 after only
4 h of exposure. This is because the inlet Hg° concentration in the current study is lower by a
factor of 15. The lower Hg° concentration in this study is representative of conditions in a coal
flue gas and the higher concentrations in our earlier study9, that of conditions in a MWC. Hence,
it is veiy likely that activated carbon would be highly under-utilized with respect to its potential
of Hg° capture in a coal-fired unit.
-------
A high utilization of the activated carbon with respect to its potential for Hg° capture is critical
for an economical Hg removal process. Unlike the high HC1 concentrations present in a MWC
flue gas, the level of HC1 is lower in a coal flue gas. This possibly accounts for higher Hg° levels
in the latter10. Therefore, high capture of Hg° becomes important in a coal-fired utility. High
capture of Hg° with the least amount of activated carbon (thereby increasing the utilization of the
activated carbon) would reduce operating costs.
The carbon/Hg° exposure time is equal to the time between cleaning the particulate collection
devices. This time is controlled by the solids handling capacity (or pressure drop) of the device.
By injecting activated carbon in pulses downstream of the particulate collection device (and
upstream of a secondary collection device for carbon), the time that the carbon particle is
exposed to Hg° can be substantially increased. Hence, the amount of Hg° in the carbon particle
would be higher, thereby reducing the material cost. Moreover, this process would allow for
regeneration of the spent activated carbon as the secondary particulate collection device traps
only the carbon.
HgCl2 Capture
Hg speciation of MWC flue gas has shown Hg to exist in the oxidized form (Hg**), although the
exact distribution of the different Hg** species is unclear. However, our equilibrium studies10
have indicated that presence of minor quantities of HC1 in the flue gas leads to a significant
fraction as HgCl2.
Figure 4 shows the capture of HgCl2 by PC-100 and FGD at 100 and 140 °C. The concentration
of HgCl2 in the gas stream varied from 0.4 to 0.7 ppb. The experimental conditions (i.e., face
velocity, bed depth, sorbent amount) were the same as those for studying Hg° capture by these
sorbents.
Similar to Hg° capture results (see Figure 3), PC-100 captures a higher percentage of incoming
HgCl2 than FGD at each temperature. This behavior is in agreement with our earlier studies
involving high concentration (30 ppb) HgCl210. In the earlier study10, we had seen that both PC-
100 and FGD captured a higher fraction of HgCl2 at 100 than at 140 °C. This is also in
agreement with our current findings for PC-100 at the low concentrations. However, FGD seems
to capture higher amounts of HgCl2 at 140 °C. The percentage of HgCl2 captured by PC-100 is
around 55-65 percent at 100 °C and around 45 percent at 140 °C for the various exposure times
tested (see Figure 4). In our earlier study10, the percentage of HgCl2 captured by PC-100 was 90-
95 percent and 75-95 percent at 100 and 140 °C, respectively, for similar exposure periods. That
is, smaller fractions of HgCl2 are captured by PC-100 when exposed to lower concentrations of
HgCl2. The same conclusion is drawn for FGD from a comparison with our earlier results10. Two
factors that could possibly cause a non-linear sorption behavior (with respect to HgCl2
concentration) in FGD and PC-100 are intraparticle mass transport and the sorption isotherm.
Our calculations show that exposure of 100 mg of PC-100 to 0.5 ppb of HgCl2 for a duration of
24 h results in approximately 0.0135 mg HgCl2/ g of PC-100 (based on 70 percent capture of
incoming HgCl2 at 100 °C). From our past study10, we found that 100 mg of PC-100 captures
-------
0,17 mg of HgCl2, That is, similar to the results with Hg° capture, activated carbon would be
highly under-utilized with respect to its potential for HgCl2 capture. In order to effectively utilize
activated carbon's Hg capture potential, longer exposure times to Hg are desirable. This can be
achieved as explained in the section on Hg° capture by providing a secondary collection unit
solely for activated carbon.
HgO Capture
The two main species of Hg that exist in equilibrium in a flue gas are Hg° and HgCl2. A third, but
not significant species (according to equilibrium predictions at 600-800 °C), is HgO. There has
been no reported data in the literature on HgO capture by dry sorbent injection.
Figure 5 shows the capture of HgO (1 ppb in N2) by 100 mg of FGD at 100 and 140 °C. The data
show that roughly 35-45 percent of the incoming HgO is captured by FGD. This value is
approximately the same as that seen for capture of Hg° and HgCl2 by FGD at the two
temperatures. Unlike the capture of Hg° and HgCl2 by FGD, however, the capture of HgO
appears to be insensitive to temperature between 100 and 140 °C.
The temperature of the furnace (HgClj/HgO generator) encasing the diffusion vial (see Figure 2)
is the controlling factor in determining the concentration of HgCl2 or HgO in the gas stream.
Higher furnace temperatures were required to generate HgO than Hg€l2 for identical amounts of
each species in the vapor phase. This is indicative of HgO having a lower volatility than HgCl2.
However, the lower volatility of HgO did not result in higher capture (compared to the other two
Hg species) by FGD. Our earlier studies9 showed that certain active sites in carbons are
responsible for Hg° capture. Also, the affinity between Hg€l2 and activated carbon is probably
due to an acid/base interaction10. Here, HgCl2 is acidic in nature and, depending on the
manufacturing process, activated carbon can be basic in character. HgO, however, is reported to
be an extremely weak base when dissolved in water24. The same percentage of capture of HgO as
HgCl2 despite HgO's lower volatility may be explained by a weaker interaction between HgO
and FGD.
Ca(OH)2 has been shown to capture high fractions of incoming Hg€l210. The reason for the high
capture is postulated to be an acid/base reaction. Figure 6 shows the capture of 10 ppb of
incoming HgO by reagent grade Ca(OH)2 (the same material used in our earlier study). Roughly
5-20 percent of incoming HgO is captured. The fraction of HgCl2 captured by Ca(OH)2on the
other hand is much higher (45-90 percent)10. This is in agreement with the postulate of a weaker
interaction of HgO with the sorbents.
A comparison of the percentage of incoming HgO captured by PC-100, FGD, and reagent grade
Ca(OH)2 at 100 and 140 °C can be seen in Figure 6. Here, the concentration of HgO is 10 ppb in
N2. Similar to the results seen for HgCl2 capture, PC-100 exhibits the highest reactivity followed
by FGD. However, unlike the HgCl2 capture results, PC-100's and FGD's performance is not
significantly affected by temperature. This is in agreement with the results shown in Figure 5 (1
ppb of HgO).
-------
SUMMARY AND CONCLUDING REMARKS
Two types of activated carbons — PC-100 and FGD — were tested for capture of low
concentration (around 1 ppb) of Hg° and HgCl2. The two sorbents' potential for capture of the
two Hg species was tested at two temperatures — 100 and 140 °C. Experiments were also
conducted to determine capture of HgO (1 ppb) by FGD. Finally, a comparison of the capture of
HgO (approximately 10 ppb) by FGD, PC-100, and Ca(OH)2 was carried out.
In general, it was found that all sorbents [the two activated carbons and Ca(OH)2] capture higher
amounts of each species of Hg at the lower temperature. The same behavior was observed in our
earlier study910 with higher concentrations of Hg° and HgCl2. This is indicative of a
physisorptive mechanism.
Results from the study on higher Hg° and HgCl2 concentration capture10 by FGD and PC-100
showed that HgCl2 was captured with greater ease than Hg°. This, however, was not the case in
the current study. PC-100 was found to capture a higher fraction of the incoming Hg° than HgCl2
at both temperatures. Also, FGD was found to capture similar fractions of Hg° and HgCl2.
A comparison with our earlier study10 showed that the fractional capture of incoming HgCl2 was
lower in this study (at the lower concentration). The reason for lower HgCl2 (1 ppb case; current
study) capture by PC-100 and FGD is perhaps the non-linear behavior of HgCl2 capture by the
activated carbons. Assuming linear dependence of external mass transfer on HgCl2 capture, the
possible cause could be non-linear dependence of intraparticle mass transport or sorption kinetics
on HgCl2 concentration.
A higher capture of Hg (mg Hg° or HgCl2 per g of activated carbon) was seen in our earlier
studies compared to the current studies. The amount of Hg° or HgCl2 after only 4 h of exposure
to 30 ppb of Hg was greater by an order of magnitude compared to the amount of Hg in activated
carbons after 24 h of exposure at the lower concentration. This implies that employing the same
procedure (injection of activated carbon before a particulate collection device) would require a
larger activated carbon:Hg ratio in a coal-fired unit compared to MWC operation. The sorbent
and, therefore, operating costs are bound to be higher in a coal-fired utility for identical Hg
percentage removal. Activated carbon injected in a coal-fired flue gas can capture more Hg if left
exposed to the flue gas until the carbon becomes saturated with Hg. In current practice, the
time of exposure may be limited by the particulate collection device's solids handling capacity
(pressure drop). Longer exposure time for carbon may therefore be achieved by injecting the
particles after the main particulate collection device, but before a secondary particulate collection
device (primarily for activated carbon).
Studies on capture of solely HgO (1 ppb) were conducted for the first time in this study. Similar
capture as Hg° and HgCl2 was found for FGD at the two temperatures. Experiments were also
conducted to study HgO capture (10 ppb) by PC-100, FGD, and Ca(OH)2. At the higher HgO
concentration, it was found that the fraction of HgO captured by FGD and Ca(OH)2 was lower
than HgCl2 captured10. The capture of HgCl2 is postulated to be an acid/base reaction, with the
three sorbents possessing basic properties. HgO is classified as a weak base in water and, hence,
-------
the solid bases do not have the same capture mechanism for HgO as HgCl2.
ACKNOWLEDGEMENTS
The authors would like to acknowledge Richard E. Valentine (EPA/APPCD) for equipment
support and Lisa Adams (Acurex Environmental Corp.) for experimental assistance. This work
has been supported in part by the Electric Power Research Institute (EPRI). The technical advice
of Ramsay Chang (EPRI) is also appreciated.
REFERENCES
1. Mercury Study Report to Congress, Book 1 of 2, External Review Draft,
EPA/600/P-94/002Aa (NTIS PB95-167334), Environmental Criteria and
Assessment Office, Cincinnati, OH, January 1995.
2. Inside EPA's Clean Air Report, vol. VI, No. 3, February 9, 1995.
3. T.G. Brna and J.D. Kilgroe, "The Impact of Particulate Emissions Control on the
Control of Other MWC Air Emissions," J. Air & Waste Met. Assoc.. 40(9), 1324
(1990).
4. N.P. Getz, I.B. Thompson, and C.K. Amos, Jr., "Demonstrated and Innovative
Control Technologies for Lead, Cadmium and Mercury from Municipal Waste
Combustors," presented at the 85th Annual Meeting and Exhibition, Air & Waste
Management Association, Kansas City, MS, June 21-26, 1992.
5. T.L. Guest, "Mercury Control in Canada," presented at the 86th Annual Meeting
and Exhibition, Air & Waste Management Association, Denver, CO, June 13-18,
1993.
6. D.M. White, W.E. Kelly, M.J. Stucky, J.L. Swift, and M.A. Palazzolo, "Emission
Test Report: Field Test of Carbon Injection for Mercury Control, Camden County
Municipal Waste Combustor," EPA-600/R-93-181 (NTIS PB94-101540), September
1993.
7. B.K. Gullett and W. Jozewicz, "Bench-scale Sorption and Desorption of Mercury
with Activated Carbon," presented at 1993 International Conference on Municipal
Waste Combustion, Williamsburg, VA, March 30- April 2,1993.
8. W. Jozewicz, S.V. Krishnan, and B.K. Gullett, "Bench-scale Investigation of
Mechanisms of Elemental Mercury Capture by Activated Carbon," presented at the
Second International Conference on Managing Hazardous Air Pollutants,
Washington, D.C., July 13-15, 1993.
-------
9. S.V. Krishnan, B.K. Gullett, and W. Jozewicz, "Sorption of Elemental Mercury by
Activated Carbons," Environ. Sci. Technol. 28, 1506-1512 (1994).
10. S.V. Krishnan, B.K. Gullett, and W. Jozewicz, "Mercury Control by Injection of
Activated Carbon and Calcium-Based Sorbents," presented at The International
Conference on Solid Waste Management: Thermal Treatment & Waste to Energy
Technologies, Washington, D.C., April 18-21, 1995.
11. R. Chang, C.J. Bustard, G. Schott, T. Hunt, H. Noble, and J. Cooper, "Pilot Scale
Evaluation of Carbon Compound Additives for the Removal of Trace Metals at Coal-
Fired Utility Power Plants," presented at the Second International Conference on
Managing Hazardous Air Pollutants, Washington, D.C., July 13-15, 1993.
12. S.J. Miller, D.L. Laudal, R. Chang, and P.D. Bergman, "Laboratory-scale
Investigation of Sorbents for Mercury Control," presented at the AWMA Annual
Meeting, Paper No. 142, Cincinnati, OH, June 20-24, 1994.
13. J.D. Wesnor, "ABB's Investigations into the Utility Air Toxics Problem," in
Proceedings: 1993 S02 Control Symposium, Vol. 3, EPA-600/R-95-015c (NTIS
PB95-179248), pp. 61-1 thru 61-19, February 1995.
14. J.G. Noblett, Jr., F.B. Meserole, D.M. Seeger, and D.R. Owens, "Control of Air
Toxics from Coal-fired Power Plants Using FGD Technology," in Proceedings: 1993
S02 Control Symposium, Vol. 3, EPA-600/R-95-015c (NTIS PB95-179248), pp. 60-1
thru 60-15, February 1995.
15. K. Felsvang, R. Gleiser, G. Juip, and K.K. Nielsen," Control of Air Toxics by Dry
FGD Systems," presented at the Power-Generation Conference, Orlando, FL,
November 17-19, 1992.
16. K. Felsvang, R. Gleiser, G. Juip, and K.K. Nielsen, "Air Toxics Control by Spray
Dryer Absorption Systems," presented at the Second International Conference on
Managing Hazardous Air Pollutants, Washington, D.C., July 13-15, 1993.
17. H.S. Huang, C.D. Livengood, and S. Zaromb, "Emission of Airborne Toxics from
Coal-fired Boilers: Mercury," ANL/ESD/TM-35, NTIS: Springfield, VA, September
1991.
18. R. Meij, "The Fate of Mercury in Coal-fired Power Plants and the Influence of Wet
Flue-gas Desulphurization," J. Water. Air, and Soil Pollution. 56, 21 (1991).
19. T. Tomita et al., "Basic Studies on a Method for Fractional Determinations of Mercury
Compounds in Air," Aichi-ken Kogai Chosa Senta Shoho, 15:19-26 (in Chemical Abstracts,
112:41589y), 1987.
-------
20. R, Chang and C.J. Bustard, "Sorbent Injection for Flue Gas Mercury Removal,"
presented at the AWMA Annual Meeting, Cincinnati, OH, June 20-24, 1994.
21. C.D. Livengood, H.S. Huang, M.H. Mendelsohn, and J.M. Wu, "Mercury Capture in
Bench-Scale Absorbers," presented at the Tenth Annual Coal Preparation, Utilization,
and Environmental Control Contractors Conference, Pittsburgh, PA, July 18-21,
1994.
22. J.R. Morency, "Control of Mercury in Fossil Fuel-Fired Power Generation," presented
at the Tenth Annual Coal Preparation, Utilization, and Environmental Control
Contractors Conference, Pittsburgh, PA, July 18-21,1994.
23. K.L. Nebel, D.M. White, C.R. Parrish, T.G. Zirkle, M.A. Palazzolo, and M.W.
Hartman, "Emission Test Report: OMSS Field Test on Carbon Injection for Mercury
Control," EPA-600/R-92/192 (NTIS PB93-105518), September 1992.
24. F. Cotton and G. Wilkinson, Advanced Inorganic Chemistry, Interscience Publishers,
New York, NY, p 483,1962.
-------
TABLE 1. MERCURY CONTROL BY EXISTING DEVICES FOR PARTICULATE AND
SULFUR DIOXIDE CONTROL
Control Device Temperature °C % Hg (Total) Captured Reference No. /Cited By
Fabric Filter (FF)
99
107
121
135
204
50
33
27
20-80
10-60
8-30
11, Chang et al., 1993
it
M
12, Miller et al., 1994
Electrostatic
Precipitator (ESP)
Wet Scrubber (WS)
Hot ESP
ESP and WS
Spray Dryer (SD)
and ESP
SD and FF
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
26-40
94-98
30-95
10-90
36
0
10-70 (avg 50)
60 in Germany
75 in Sweden
88
57-79
5-25 (low CI
coal)
45-96 (high CI
coal)
13, Wesnor, 1995
14, Noblett et al., 1995
19, Tomita et al., 1987
PISCES database
13, Wesnor, 1995
13, Wesnor, 1995
18, Meij, 1991
17, Huang et al., 1991
15, Felsvang etal., 1992
16, Felsvang et al., 1993
TABLE 2. MERCURY CAPTURE AT FOUR OF THE EIGHT POWER PLANTS
COMMISSIONED TO BE CHARACTERIZED BY DOE/PETC
Power Plant Power (MW) Control Device % Hg (Total) Captured
Yates Unit No. 1 100 ESP and WS 50
(Georgia Power Co.)
Clay Boswell Energy Center 61 FF 70
Unit No. 2 (Minnesota Power Co.)
Baldwin Power Stations Unit No. 565-575 ESP 25
2 (Illinois Power Co.)
Springville Generation System 360 3 SD and 2 FF 15 across FF
(Tuscon Electric Power Co.) 25 across whole
-------
Hood
Trap
By-Pass
On-line
Analyzer
Generator
FURNACE
3 sorbent stages
(reactor)
PERSONAL
COMPUTER
CHART
RECORDER
Figure 1. Schematic diagram of bench-scale apparatus used for Hg sorption studies (v: valve).
Hood
By-Pass
3 sorbent stages (reactor)
/s
diffusion vial
FURNACE
Figure 2. Schematic diagram of bench-scale apparatus to study HgO or HgCl2 sorption (v: valve).
-------
100
80
60
40
20
0
PC-100
100*C
o
PC-100
140C
FGD
100-C
-•O-
FGD
140*C
O'*"
~,
-
0
10
15
TIME, h
20
25
30
Figure 3. Capture of low concentration Hg (2 ppb) by PC-100 and FGD at 100 and 140°C.
100
80
£ 60
40
20
0
PC-100
100«C
PC-100
14Q*C
FGD
10CMD
-•O"
FGD
14Q*C
*o
10
15
TIME, h
Figure 4. Capture of low concentration HgCI (0.4-0.7 ppb) by PC-100 and FGD at 100 and 140 C.
2
-------
100
0
O)
1
Q
LU
DC
D
t—
£L
<
O
I-
2
LU
o
DC
UJ
Q.
80
60
40
20
0
0
^
•O
10OC
140C
15
5 10
TIME, h
Figure 5. Capture of low concentration HgO (1 ppb) by FGD at 100 and 140° C.
20
100
o
o>
X
o
LU
cc
h-
0.
<
o
LU
O
CC
LU
Q.
80
60
40
20
FGD 10OC
-¦<>-
FGD 140«C
PC100100-C
-e-
PC100 140C
lime ioo»c
LIME 140C
*****••••••
0
0
10
TIME, h
15
20
Figure 6. Capture of HgO (10 ppb) by PC-100, FGD, and Ca(OH^ at 100 and 140°C.
-------
mdtv/tdt DTD T3 noo TECHNICAL REPORT DATA
iNKIVijn,J_j_rwll x UZ,3 (Please read Instructions on the reverse before compte
1. REPORT NO. 2.
EPA/600/A-96/053
3,
4. TITLE AND SUBTITLE
Control of Mercury Emissions from Coal Combustors
5. REPORT DATE
6. PERFORMING ORGANIZATION CODE
7, AUTHOR
------- |