Mercury Oxidation by UV Irradiation: Effect of Contact Time, UV

May 9, 2017 - A separate study of mercury oxidation at 39 and 138 °C also used 254 nm .... This study determined the variability of mercury oxidation...
0 downloads 0 Views 709KB Size
Article pubs.acs.org/IECR

Mercury Oxidation by UV Irradiation: Effect of Contact Time, UV Wavelength, and Moisture Content Alexander F. Gruss,†,∥ Regina Rodriguez,*,‡,§ and David W. Mazyck‡,⊥ †

Siemens Industry Inc., 4800 North Point Parkway, Alpharetta, Georgia 30022, United States Environmental Engineering Sciences, University of Florida, Gainesville, Florida 32611, United States



ABSTRACT: Elemental mercury was irradiated with UV at 185, 254, and 365 nm wavelengths over a range of water vapor concentrations (0.001−800 ppmv) and temperatures (27−149 °C) to assess what conditions favored its oxidation. There was no measurable mercury oxidation with the 365 nm wavelength at any conditions. Oxidation was higher with 185 nm irradiation over the range of temperatures and water vapor concentrations compared to 254 nm irradiation. With little water vapor in the air stream, oxidation at the 254 nm wavelength was very low (35%). However, as water vapor concentration increased, mercury oxidation increased to 80% at 93 and 149 °C under 254 nm irradiation. Mercury oxidation at 185 nm was steadier over the range of water vapor concentrations tested (80−90%), with decreasing levels of oxidation at the upper range of water vapor concentrations for all temperatures. Oxidation was tested at various contact times at both 254 and 185 nm wavelengths, with optimal oxidation with a contact time of 1.2 s.



INTRODUCTION Mercury (Hg) is classified as a hazardous air pollutant (HAP) because it has adverse effects on human health.1 The 1990 Amendments to the Clean Air Act directed the Environmental Protection Agency (EPA) to study the health and environmental impacts of HAPs,2 and the EPA concluded that mercury emitted into the atmosphere bioaccumulates in the environment and can cause impaired neurological development in fetuses, infants, and children.3 In the U.S., coal-fired power plants account for over onethird of the total anthropogenic mercury emissions,4 approximately 50 tons of mercury annually.1 In 2000, the EPA announced its intent to regulate HAP emissions from coal- and oil-fired power plants, and in April 2015 this legislation was enacted. The expectation is about a 90% reduction in mercury emissions across the U.S. coal fleet. Currently, activated carbon injection is considered the most reliable technology by the industry. However, a major drawback is the cost, as the injection rates to achieve high mercury removal can be cost prohibitive and mercury adsorption is dependent on flue gas conditions. To improve activated carbon performance, bromine can be added either to the coal (e.g., as CaBr2) or to the carbon (i.e., brominated activated carbon). However, although costs of the sorbent are reduced because the amount required is less, overall compliance costs may increase from corrosion of plant equipment.5 Corrosion of activated carbon storage silos, air-heater baskets, etc. can easily cost a power plant millions of dollars. Another strategy is multipollutant control, where technologies (wet flue gas desulfuriza© 2017 American Chemical Society

tion (WFGD) scrubbers) designed to remove other pollutants (e.g., sulfur dioxide) are also used to additionally remove mercury.6,7 This approach works well with high-rank coals (e.g., eastern bituminous) that contain a higher chlorine content, which causes a higher percentage of oxidized mercury, a form more water-soluble than elemental mercury (Hg0) and thus more easily removed by a WFGD. For low-rank coals, such as lignite or sub-bituminous coals, elemental mercury is the dominant species in the flue gas resulting in lower mercury capture by a WFGDs alone. Therefore, an approach where mercury is oxidized, for example, via UV irradiation, to convert elemental Hg to oxidized Hg just prior to the WFGD could offer a viable technology for coal-fired power plants burning low rank coals. Simply recoating SCR catalysts, when installed, is in excess of $10 M let alone the cost of installing a new SCR to the flue gas air pollution control system. Although UV requires constant power, activated carbon injection likewise requires constant energy to pneumatically convey the PAC from the silo to the flue duct. Therefore, UV technology offers a feasible solution. It is envisioned that UV lamps could be placed perpendicular to the flue gas to maximize irradiation. UV lamps can be purchased commercially up to 3 m in length; therefore, two UV lamps in a transparent (e.g., quartz) sleeve could be envisioned for this application just upstream of the Received: Revised: Accepted: Published: 6131

January 3, 2017 May 6, 2017 May 9, 2017 May 9, 2017 DOI: 10.1021/acs.iecr.7b00032 Ind. Eng. Chem. Res. 2017, 56, 6131−6135

Article

Industrial & Engineering Chemistry Research

Figure 1. Test stand for monitoring impact and variables on Hg oxidation.

WFGD (i.e., in the flue gas duct feeding the WFGD). The number of UV lamps required is a function of inherent oxidation efficiency, flue gas flow rate, and flue gas duct dimensions. Briefly, in previous work,8 UV plus silica−titania coatings on ceramic packing material proved capable of 90% Hg removal in a pilot-scale flue gas demonstration. The intent of the work herein is to begin to understand whether the packing material can be removed to simplify the solution (i.e., rely solely on UV irradiation). Typical flue gas temperatures prior to the WFGD are approximately 100−150 °C. The typical water vapor concentrations (WVCs) prior to the WFGD and after the particulate collection device is 6−17% WVC.9 Studying a range of WVCs was of interest particularly since water vapor may have the biggest impact on Hg oxidation as discussed below. Indeed, some have studied approaches to oxidize Hg0 including nonthermal plasma,10 corona discharge,11 ozone,12 or UV irradiation.13,14 UV irradiation employed to oxidize mercury was studied at 254 nm with simulated flue gas, at temperatures between 27 and 177 °C,13 and focused on the effects of gas components (SO2, NO) as well as light intensity. This study found that mercury was most easily oxidized at temperatures below 149 °C and that nitric oxide (NO) gas negatively affected mercury oxidation. A separate study of mercury oxidation at 39 and 138 °C also used 254 nm UV light and a simulated flue gas.14 This study also found that a temperature increase and the presence of NO had negative effects on mercury oxidation. In both studies, the UV wavelength and WVC were not varied. These two variables are both modified herein because (1) 365 and 185 nm wavelengths would likely result in different outcomes and (2) we wanted to understand the impact of varying moisture contents. Natural gas environments have shown lower moisture contents in the gas stream;9 therefore, very low WVCs were also studied in this work. As previous literature provided insight regarding flue gas constituents, the focus of this work was on the impact of UV wavelength, particularly 185 nm UV that results in some ozone production. The reaction occurring between mercury and oxygen in the presence of 254 nm light is given by equation1 Hg 0 + 2O2 + 254 nm light → HgO + O3

Due to Hg0 serving as a sensitizer for ozone formation, the reaction products are mercuric oxide as well as ozone.15−17 Secondary reactions studied further in this work include ozone’s capability of oxidizing mercury.13 Oxidized Hg is the conversion of elemental Hg (Hg0) to Hg2+, and whether in the presence of oxygen or chlorine it can react to form HgO or HgCl2, oxidized Hg is highly soluble in water and would be easily captured in the WFGD via absorption. Herein, the objective was to investigate UV wavelength, contact time, temperature, and water vapor concentrations on mercury oxidation. Other works17,18 have looked at additional variables including contact time and reactor volumes, but here the primary interest was to understand the oxidation behaviors of Hg and how it can be oxidized in the presence of various UV wavelengths and without the use of sorbents or catalyst.



EXPERIMENTAL SECTION The test stand used in this study can be seen in Figure 1. Air from a compressed air tank (Airgas) is run through a Pyrex tube (“main line”) wrapped in heat tape with the temperature controlled by a power controller. Both gas flow rate (3−6 LPM) and the flow rate of water (varied flow rates) into the system were controlled by rotameters (Aarlborg, various models). Although the concentration of oxygen in this study (21% of O2 in air) is likely greater than that in a coal-fired power plant flue gas, the intent was to ensure excess oxygen for eq 1. It is not expected that oxygen concentration between, for example, 4% and 21% would impact the results. Previous findings13,14 support that increases in O2 concentrations beyond 6% had minimal impact on Hg oxidation. They also showed that under UV radiation, O2 can produce strong oxidizing agents O3 and O− to oxidize Hg0. Flow rates for the employed gases were determined based on typical flue gas contact times which were translated to the lab scale reactor.17 Elemental mercury was introduced into the system by passing nitrogen gas (0.2 LPM) over an elementalmercury-containing bubbler in a heated water bath (maintained at 41 °C) to minimize fluctuation in mercury delivery. To control the water vapor concentration, deionized (DI) water was injected from a pressure vessel (pressurized by high-purity helium at 0.7 atm, Airgas) via a rotameter into the main line where it immediately vaporized.

(1) 6132

DOI: 10.1021/acs.iecr.7b00032 Ind. Eng. Chem. Res. 2017, 56, 6131−6135

Article

Industrial & Engineering Chemistry Research The 100 mL Pyrex annular reactor with a quartz sleeve for UV irradiation was sufficiently spaced from the main line to allow mixing of the gas constituents to occur. The inside of the reactor was approximately 14 cm long, and the quartz sleeve was 2.54 cm in diameter. There was a space of approximately 0.76 cm between the outside of the quartz sleeve and the inside of the reactor. The reactor, tubing leading to it, and the tubing following it were all wrapped in heat tape to maintain constant desired temperatures in the test stand. Teflon tubing was used on all lines that were not Pyrex to avoid adsorption of mercury. A junction after the reactor allowed for the effluent air to be directed to the point of analysis while the excess air was exhausted (see Figure 1). The method of analysis used was an Ohio Lumex Zeeman mercury analyzer (model RA-915+). The Zeeman mercury analyzer (“Zeeman”) is an instrument that measures gaseous elemental mercury in real time using the principle of atomic absorption spectrometry to detect and quantify mercury. EPA method 30b was used to confirm full oxidation of elemental mercury and establish a mass balance of total mercury in the system. The sampling interval of the Zeeman was every 5 s, and at 15 min, the Hg effluent averages were recorded. The Hg oxidation in these experiments was examined at various contact times (0.3−1.5 s) to understand oxidation efficiency. The mercury concentration was maintained at 10 ppb (89 μg m−3), which is like concentrations observed under actual flue gas conditions.6 Wavelengths of 365, 254, and 185 nm (lamps manufactured by Atlantic Ultraviolet Corporation) were chosen, the last two lower than the 365 nm wavelength examined in existing literature.19,20 The lamps were heated for 30 min prior to experiments, and temperatures during experiments did not fluctuate. UV lamp manufacturers use 185 nm and ozone lamps synonymously as the 185 nm is known to create ozone in the presence of oxygen.21

Figure 2. Mercury oxidation by 254 nm UV at 149 °C and WVC of 26 ppmv with varying contact times.

Figure 3. Mercury oxidation by 185 nm UV at 149 °C and WVC of 26 ppmv with varying contact times.



RESULTS AND DISCUSSION Effect of Contact Time. Contact time is an important design factor in designing purification systems, as a shorter contact time allows for a smaller reactor volume lowering costs. In this study, it was necessary to determine the minimum contact time required to achieve at least 91% mercury oxidation, as this is the mercury removal required by the United States Environmental Protection Agency (EPA).22 The assumption is that if greater than 91% of the elemental fraction is oxidized, it would be captured in the WFGD. Contact times in this study were varied from 0.3 to 1.5 s, similar to those found in treatment systems of coal combustion power plants. These times were chosen based on the assumption that UV lamps would be placed perpendicular to the flue gas and irradiate the flue gas in all three dimensions of the flue duct. Full-scale testing would be required to verify this assumption. Oxidation was expected to increase as the contact time increased, which was confirmed (see Figures 2 and 3). Figure 2 shows mercury oxidation via 254 nm irradiation at 149 °C and a WVC of 26 ppmv. Oxidation reached 86% at a 1.2 s contact time but leveled off as contact time increased. At a UV wavelength of 185 nm and the same WVC and temperature (Figure 3), there was a greater level of mercury oxidation at contact times under 1 s than at the same contact times with 254 nm UV. However, mercury oxidation did not exceed 91% until a contact time of 1.2 s. The Reynolds numbers vary from 440 down to 90 for the 0.3 and 1.5 s contact times, respectively, indicating laminar flow. This might explain the low oxidation at

the shorter contact times. Alternatively, it is known that Hg2+ will convert back to Hg0 in the presence of UV and water vapor as studied by Byrne, 2009,23 and as discussed in more detail below. Effect of Temperature, UV Wavelength, and Water Vapor. Water vapor was hypothesized to negatively impact Hg oxidation as described by Byrne, 2009,23 and because it could absorb photons limiting Hg irradiation. Therefore, it was important to determine the change in mercury oxidation by the UV bulbs as WVC was varied. When the gas was irradiated with 365 nm UV, oxidation was minimal over a range of temperatures and WVCs (Figures 4 and 5) likely because it lacks sufficient energy to excite mercury electrons to the conduction band.20 The 365 nm light lacks the energy to permit mercury to serve as a photosensitizer (eq 1) because mercury can only absorb radiation at less than 254 nm. At temperatures of 93 and 149 °C (Figures 4 and 5), there are two notable trends. First, at lower water vapor concentrations, irradiation with 185 nm produces more oxidized Hg than with 254 nm UV. While both 185 and 254 nm can directly oxidize Hg (eq 1) and generate hydroxyl radicals,23 the additional oxidant (O3) (present only with 185 nm irradiation) results in greater oxidation of Hg as detailed by Hall, 1995.12

6133

O2 + OH− + 185 nm UV light → O3 + H+

(2)

O3 + Hg 0 → HgO + O2

(3) DOI: 10.1021/acs.iecr.7b00032 Ind. Eng. Chem. Res. 2017, 56, 6131−6135

Article

Industrial & Engineering Chemistry Research

seen by others13,14 in previous work where temperatures above 138 °C started to decrease the oxidation of Hg. A plausible explanation for the temperature difference is the lack of photon efficiency of the UV lamp as the gas temperature increases from 93 to 149 °C. This confirms that even though Hg oxidation is a fast photochemical process, temperature and WVC can significantly impact the oxidation reactions.



CONCLUSION Mercury oxidation of at least 90% was only achievable under certain scenarios (WVC of 100 ppmv) with 185 nm UV irradiation, where the water vapor concentrations were high enough to aid in the formation of OH radicals but not sufficient to impede ozone formation. This study determined the variability of mercury oxidation under UV irradiation, where the Hg oxidation will prevail in an environment exhibiting as high water vapor concentrations, low temperatures, and UV irradiation consistent with 254 nm wavelengths. It was also concluded that this method could be efficient in removing mercury in conjunction with a capturing mechanism, such as a WFGD scrubber.

Figure 4. Mercury oxidation in comparison to increasing water vapor concentrations at varying UV wavelengths (185, 254, and 365 nm) at 93 °C and 1.2 s contact time.



AUTHOR INFORMATION

Corresponding Author

*R.R.: e-mail, reggie17r@ufl.edu. ORCID

Regina Rodriguez: 0000-0002-8268-7867 Present Address §

R.R.: Environmental Engineering Services, University of Florida, Gainesville, FL 32611, United States.

Notes

The authors declare no competing financial interest. ∥ A.F.G.: e-mail, [email protected]. ⊥ D.W.M.: e-mail, dmazyck@ufl.edu

Figure 5. Mercury oxidation in comparison to increasing water vapor concentrations at varying UV wavelengths (185, 254, and 365 nm) at 149 °C and 1.2 s contact time.



Second, there is a decline in oxidation of Hg for 185 nm UV light with increasing WVC. It is understood that ozone is produced by UV when an oxygen molecule absorbs the energy from 185 nm UV light and dissociates whereby oxygen atoms react with oxygen molecules to form ozone (eq 2). The resulting ozone can then engage in reaction from eq 3 to oxidize Hg.12 Perhaps there is a decline in ozone production from the excess WVC impeding ozone formation by the bulb.24 In this scenario, water vapor aids in oxidation through formation of OH radicals by UV irradiation25 while not inhibiting the interaction of the radicals with Hg0. WVC of 100 ppmv was found to be low enough to not interfere with ozone formation but at the same time sufficient to form OH radicals to aid in Hg oxidation of 90% (Figure 4). These OH radicals are formed from hydroxyl ions present in the water vapor by UV irradiation, and with more water vapor there are more hydroxyl radicals. The radicals subsequently aid in the oxidation of mercury as depicted in eq 4.26 OH* + Hg 0 → Hg 2 + + OH−

REFERENCES

(1) Keating, M. H.; Beauregard, D.; Benjey, W. G.; Driver, L.; Maxwell, W. H.; Peters, W. D. Mercury study report to congress. Volume 2: An inventory of anthropogenic mercury emissions in the United States. EPA-452/R-97-004; U.S. Environmental Protection Agency, 1997. (2) S.1630: Clean Air Act Amendments of 1990. Washington, D.C., 1990. (3) Study of Hazardous Air Pollutant Emissions from Electric Utility Steam Generating UnitsFinal Report to Congress. U.S. Environmental Protection Agency: Washington, D.C., 1998. (4) Pai, P.; Niemi, D.; Powers, B. A North American inventory of anthropogenic mercury emissions. Fuel Process. Technol. 2000, 65−66, 101−115. (5) The Electric Power Research Institute (EPRI). 2014 Update on EPRI’s Balance-of-Plant Effects Study of Bromine-Based Mercury Controls. http://www.epri.com/abstracts/Pages/ProductAbstract. aspx?ProductId=000000003002003404, 2014. (6) Pavlish, J. H.; Hamre, L. L.; Zhuang, Y. Mercury control technologies for coal combustion and gasification systems. Fuel 2010, 89, 838−847. (7) Tabatabaie-Raissi, A.; Muradov, N. Z.; Peng, P. H. U.S. Patent 5,842,110, 1998. (8) Casasus, A. I.; Gruss, A.; Baun, D.; Morales, M.; Mazyck, D. Silica-Titania Coated Packing: A Novel Solution Capable of 90% Hg Capture with Low Operation and Maintenance Costs. J. Environ. Eng. 2013, 139, 86−94. (9) Levy, E.; Bilirgen, H.; Jeong, K.; Kessen, M.; Samuelson, C.; Whitcombe, C. Recovery of water from boiler flue gas. DOE, 2008.

0

(4)

At 93 °C, oxidation is also declining with increasing WVC in the 254 nm irradiated system, and the plausible explanation is provided by Byrne, 2009,23 who demonstrated the reduction of oxidized Hg in water with 254 nm light. At 93 °C, Hg oxidation was improved over those experiments conducted at 149 °C as 6134

DOI: 10.1021/acs.iecr.7b00032 Ind. Eng. Chem. Res. 2017, 56, 6131−6135

Article

Industrial & Engineering Chemistry Research (10) Byun, Y.; Ko, K. B.; Cho, M.; Namkung, W.; Shin, D. N.; Lee, J. W.; Koh, D. J.; Kim, K. T. Oxidation of elemental mercury using atmospheric pressure non-thermal plasma. Chemosphere 2008, 72, 652−658. (11) Ko, K. B.; Byun, Y.; Cho, M.; Namkung, W.; Hamilton, I. P.; Shin, D. N.; Koh, D. J.; Kim, K. T. Pulsed corona discharge for oxidation of gaseous elemental mercury. Appl. Phys. Lett. 2008, 92, 251503. (12) Hall, B. The gas-phase oxidation of elemental mercury by ozone. Water, Air, Soil Pollut. 1995, 80, 301−315. (13) Granite, E. J.; Pennline, H. W. Photochemical removal of mercury from flue gas. Ind. Eng. Chem. Res. 2002, 41, 5470−5476. (14) Jia, L.; Dureau, R.; Ko, V.; Anthony, E. J. Oxidation of Mercury under Ultraviolet (UV) Irradiation. Energy Fuels 2010, 24, 4351−4356. (15) Dickinson, R. G.; Sherrill, M. S. Formation of ozone by optically excited mercury vapor. Proc. Natl. Acad. Sci. U. S. A. 1926, 12, 175− 178. (16) Volman, D. H. Reaction of Optically Excited Mercury Vapor with Oxygen. J. Chem. Phys. 1953, 21, 2086. (17) Li, Y.; Lee, S. R.; Wu, C. Y. UV-Absorption-Based Measurements of Ozone and Mercury: An Investigation on Their Mutual Interferences. Aerosol Air Qual. Res. 2006, 6, 418−429. (18) Pitoniak, E.; Wu, C. Y.; Mazyck, D. W.; Powers, K. W.; Sigmund, W. Adsorption enhancement mechanisms of silica-titania nanocomposites for elemental mercury vapor removal. Environ. Sci. Technol. 2005, 39, 1269−1274. (19) Lee, T. G.; Biswas, P.; Hedrick, E. Overall kinetics of heterogeneous elemental mercury reactions on TiO2 sorbent particles with UV irradiation. Ind. Eng. Chem. Res. 2004, 43, 1411−1417. (20) Li, Y.; Wu, C. Y. Kinetic study for photocatalytic oxidation of elemental mercury on a SiO2-TiO2 nanocomposite. Environ. Eng. Sci. 2007, 24, 3−12. (21) Jeong, J.; Sekiguchi, K.; Lee, W.; Sakamoto, k. Photodegradation of gaseous volatile organic compounds (VOCs) using TiO2 photoirradiated by an ozone-producing UV lamp: decomposition characteristics, identification of by-products and water-soluble organic intermediates. J. Photochem. Photobiol., A 2005, 169, 279−287. (22) Harder, A. EPA to Issue Power-Plant Emissions Rules This Summer. Wall Street J. 2015 (Jan 7), http://www.wsj.com/articles/ epa-to-issue-power-plant-emissions-rules-this-summer-1420650958. (23) Byrne, H.; Borello, A.; Bonzongo, J.; Mazyck, D. Investigations of Photochemical Transformations of Aqueous Mercury: Implications for Water Effluent Treatment Technologies. Water Res. 2009, 43, 4278−4284. (24) Lukes, P.; Appleton, A. T.; Locke, B. R. Hydrogen peroxide and ozone formation in hybrid gas-liquid electrical discharge reactors. IEEE Trans. Ind. Appl. 2004, 40, 60−67. (25) Caren, R. P.; Ekchian, J. A. Method for Using Hydroxyl Radical to Reduce Pollutants in the Exhaust Gases from the Combustion of a Fuel. U.S. Patent 5,863,413, Jan 26, 1999. (26) Zhang, H.; Lindberg, S. E. Sunlight and iron (III)-induced photochemical production of dissolved gaseous mercury in freshwater. Environ. Sci. Technol. 2001, 35, 928−935.

6135

DOI: 10.1021/acs.iecr.7b00032 Ind. Eng. Chem. Res. 2017, 56, 6131−6135