Microsoft Word - 476hernandez.docx CHEMICAL ENGINEERING TRANSACTIONS VOL. 43, 2015 A publication of The Italian Association of Chemical Engineering Online at www.aidic.it/cet Chief Editors: Sauro Pierucci, Jiří J. Klemeš Copyright © 2015, AIDIC Servizi S.r.l., ISBN 978-88-95608-34-1; ISSN 2283-9216 A Comparison among Different Sorbents for Mercury Adsorption from Flue Gas Dino Musmarraa, Despina Karatzaa, Amedeo Lanciab, Marina Prisciandaroc, Giuseppe Mazziotti di Celsod aDipartimento di Ingegneria Civile, Design, Edilizia e Ambiente, Seconda Università degli Studi di Napoli, Real Casa dell’Annunziata, Via Roma 29, 81031 Aversa (CE), Italy bDipartimento di Ingegneria Chimica, dei Materiali e della Produzione Industriale, Università di Napoli “Federico II”, P.le Tecchio 80, 80125 Napoli, Italy cDipartimento di Ingegneria Industriale e dell’Informazione e di Economia, Viale Giovanni Gronchi 18, 67100 L’Aquila, Italy dDipartimento di Scienze degli Alimenti, Università degli Studi di Teramo, Via Carlo R. Lerici, 64023 Mosciano S. Angelo (TE), Italy marina.prisciandaro@univaq.it The adverse effects of mercury on human health are a well-known problem, as mercury is a persistent toxic pollutant. Among the anthropogenic sources, the coal fired utility boilers, the combustion of municipal solid waste (MSW) and the cement production process are the main sources of mercury emission. The problem of mercury abatement is particularly tricky, because during combustion up to 98% of mercury is rapidly volatized without being captured by means of particulate control devices. The present work is focused on the removal of mercury compounds in flue gas by dry adsorption process on solid sorbents, which is considered a BAT (Best Available Technology) for mercury pollution control. In this paper, a comparison among different commercial solid sorbents (i.e. several commercial activate carbons) used by this research group to adsorb metallic and divalent mercury from combustion flue gas is presented. The tested activated carbon are HGR carbon and Darco G60 carbon, impregnated or not. The comparison is carried out in terms of isotherm curves, for the temperatures ranging from 90°C to 200°C, which is an interval typical of the exhaust gas treatment line downstream an industrial combustor. 1. Introduction Mercury emissions can cause serious problems to human health. This is the reason why European Commission in 2006 established strict mercury concentration limits in some aquatic food (European Commission, 2006) reserved to human consumption, where mercury most efficiently accumulates (Pavlish, 2009). Mainly, mercury emissions can be found in exhaust gas from coal fired utility boilers (Uddin et al., 2008; Granite et al., 2000; Mazziotti di Celso et al., 2013), from the combustion of municipal solid waste (MSW) (Serre and Silcox, 2000; Jurng et al., 2002) and from the cement production process (Zheng et al., 2012; Prisciandaro et al., 2003). Among these three processes, the highest values are imputed to MSW, as showed by Pacyna et al. (2006), who declared from 200 to 1000 μg of Hg/Nm3 mercury concentrations in exhaust gas, while in coal fired utility boilers and cement production process they varies between 5 and 20 μg of Hg/Nm3 (Zheng et al., 2012). In addition, not only mercury emission values, but also mercury species have to be considered: elemental mercury (Hg0) is very hard to remove, due to its insolubility in water, low melting point and scarce reactivity with other species. Owing to these features, elemental mercury treatment has been widely studied in literature (Zheng et al., 2012; Karatza et al., 2011; Yao et al., 2014). Anyway in the flue gases from MSW incinerators, mercury is mainly found as HgCl2, due to the relatively high concentration of HCl (Heidel et al., 2014), while elemental mercury is the prevailing form in emissions from coal combustion processes (Pacyna et al., 2006). DOI: 10.3303/CET1543411 Please cite this article as: Musmarra D., Karatza D., Lancia A., Prisciandaro M., Mazziotti Di Celso G., 2015, A comparison among different sorbents for mercury adsorption from flue gas, Chemical Engineering Transactions, 43, 2461-2466 DOI: 10.3303/CET1543411 DOI: 10.3303/CET1543411 Please cite this article as: Musmarra D., Karatza D., Lancia A., Prisciandaro M., Mazziotti Di Celso G., 2015, A comparison among different sorbents for mercury adsorption from flue gas, Chemical Engineering Transactions, 43, 2461-2466 DOI: 10.3303/CET1543411 2461 The removal of Hg0 and HgCl2 is still a topical research. Many authors proposed different cleanup technologies, like mercury capture by means of sorbent injection (Pavlish et al., 2003; Yang et al., 2007; Karatza et al., 1996, 1998; Lancia et al., 1996): this is a promising one, which permits to incisiveness remove elemental mercury, with the main drawbacks due to the excessive amount of sorbent required, which is expensive, and the hard loading of particulate to the filter baghouse (Bustard et al., 2004). Better performances can be obtained with adsorption technique onto active carbons (Musmarra et al., 2013). Particularly, active carbons with a large surface area allow mercury physical adsorption. Furthermore when they are impregnated with sulfur, iodine, chlorine or bromine, active carbons have demonstrated to be very efficient as regard mercury abatement. This remarkable behavior is due to the ability to activate, on the carbon surface, a lot of sites, where very strong chemical bounds with mercury are established (Choi et al., 2009). However, at the moment, in spite of the engagement of the scientific community, kinetic and thermodynamic mechanisms which lay the foundations for active carbon adsorption are not well known. In this paper a comparison of mercury adsorption results obtained by using different commercial sorbents, impregnated or not, is reported. In particular experimental runs were performed with the following commercial activated carbons: HGR (from Calgon-Carbon Corp.) (Karatza et al., 2000), and Darco G60 (from BDH) alone (Karatza et al., 1996a) and impregnated with Na2S in different percentage (Karatza et al., 1996b). The study was carried out in an apparatus at laboratory scale, in which simulated flue gas containing Hg0 and HgCl2, separately, was contacted with the above mentioned active carbons. Experiments were performed at a temperature of interest in the removal process (90−200°C) and with the concentration of Hg0 and HCl2 in the gas stream varying in the range of 1−7 mg/m3. Finally, equilibrium data were fitted with Langmuir isotherm, to obtain Langmuir parameters. 2. Experimental In Figure 1 the apparatus used for the experiments with elemental mercury is sketched. A simulated flue gas, at the required temperature and Hg0 concentration, was produced and mercury adsorption on fixed bed was carried out. The simulated flue gas was obtained by evaporating the liquid mercury contained in a thermostated glass saturator into a stream of pure nitrogen. The saturator was made of a horizontally placed string of 10 empty glass spheres 30 mm ID, connected by short and narrow glass tubes (about 1 mm ID, 5 mm long), into which about 350 g of liquid mercury (reagent grade Hg0 from BDH) were deposited. This design allowed both a relatively large gas-liquid contact area (in the spheres) and a good mixing of the gas (in the tubes). The mass flow rate of nitrogen into the saturator was kept constant by a hot wire flow rate controller. The Hg0 concentration in the gas stream fed to the reactor was controlled by varying the temperature of the saturator and by diluting in the junction N the saturated stream with one of pure nitrogen, whose flow rate was controlled by another hot wire flow rate controller. Figure 1 – Experimental apparatus for elemental mercury RS 232 A W R TIC WTB N2 PIFRC PIFRC N2/ O2 N H S Liquid metallic mercury V C H FRC, Flow Rate Control PI, Pressure Indicator TIC, Temperature Indicator Control WTB, Water Thermostated Bath N, Junction V, Valve H Heater S, Saturator C, Owen R, Reactor W, Fan A, Analyzer Gas exit 2462 The diluted stream was fed to a glass reactor containing the sorbent. The reactor was 35 mm ID and 60 mm high, and was kept in a thermostated oven equipped with a PID temperature controller. The adsorbing bed was made of adsorbing particles (10 – 30 mg) mixed with 100 μm glass beads (5 g) as inert, and its length (L) was 6 mm. This arrangement was chosen in order to avoid channeling, while keeping a small reactive surface. Figure 2 shows the apparatus used for the experiment of HgCl2. The simulated flue gas, at the required temperature and HgCl2 concentration, was produced and the mercury chloride adsorption on fixed bed was carried out. The simulated flue gas was obtained by sublimating reagent grade solid HgCl2 contained in a stainless steel (AISI 316) cylindrical saturator into a stream of pure nitrogen. The saturator was kept at a fixed temperature by a heating tape driven by a PID temperature controller and the mass flow rate of nitrogen stream was measured by a hot wire flow rate controller. The HgCl2 concentration in the gas stream fed to the reactor was controlled by varying the temperature of the saturator and by diluting the saturated stream with a stream of pure nitrogen in the junction N. The diluted stream was fed to a glass reactor containing the sorbent. The reactor was 8.85 mm I.D. and 60 mm high, and was kept into a thermostated oven equipped with a PID temperature controller. The bed was made of 12 mg of adsorbing particles mixed with 3.0 g of inert glass beads of the same size, and its length (L) was 40 mm. This arrangement was chosen with the aim of operating with a bed long enough to avoid channeling, while keeping a small reactive surface. For both experimental apparatuses a downward flow was used in the reactor, in order to avoid losses of powder in the gas stream. Different commercially activated carbon was used as sorbent: HGR by Calgon-Carbon Corp., Darco G60 from BHD alone and impregnated with Na2S, too. Tab. 1 reports the physical characteristics of the adsorbing materials. Figure 2 − Experimental apparatus for mercury chloride The two impregnated activated carbons were obtained by sinking the raw activated carbon into concentrated aqueous solutions of Na2S and then by drying the filtered solid. Namely, the two solutions used had 5% and 15% (w/w) concentrations of Na2S and, correspondingly, the impregnated carbons had Na2S content of 7.8% and 18.7% (w/w). As reported in Table 1 the impregnation process considerably reduces the surface area of the solid and increases its bulk density. Table 1 – Physical characteristics of the adsorbing materials Material Average diameter (μm) BET surface area m2/g Bulk density kg/m3 Darco G60 3.5 230 4.50×102 Darco G60 + 7.8% Na2S 3.9 184 4.97×102 Darco G60 + 18.7% Na2S 3.9 123 5.62×102 HGR 4.5 628 5.92×102 In the experimental runs the temperature of the adsorbing bed (T) was varied between 90 to 200°C, while the Hg0 or HgCl2 concentration (c0) in the gas stream entering bed was changed from 1 to 7 mg/m3. The Hg0 concentration in the gas stream out-flowing from the reactor was measured, as a function of time, by using the mercury continuous analyzer MONITOR 2000 by Seefelder Messtechnik. The HgCl2 concentration was determined by adsorbing the gas by means of gas washing bottles containing a KMnO4 aqueous solution acidified by H2SO4 and then by analyzing the samples by means of Cold Vapor Atomic Adsorption (CVAA), 2463 using NaBH4 as reducing agent. Such procedure was considered accurate, since no mercury was found in the gas exiting the washing bottle, when another bottle was placed in series to the first. Besides, the quantity of mercuric chloride and elemental mercury adsorbed on the bed was measured by leaching after each run the bed material with aqua regia (HNO3 + 3HCl) and then analyzing the solution by means of CVAA, and a difference not larger than 8% was found in the mercury material balance. 3. Results and Comments The phenomenon investigated is an adsorption process: hence the adsorbate loading at equilibrium is a function of gas phase concentration. So, if the Langmuir isotherm is used to describe the sorbent-sorbate interaction, the rate of the process (r) can be expressed as the difference between the adsorption rate and the desorption rate, as stated by the following equation: = − − (1) where c is the Hg0 concentration in the gas phase, ω is its concentration as adsorbate on the solid, ωmax is the asymptotic adsorbate concentration, and k1 and k2 are the kinetic constants of the adsorption and of the desorption reaction, respectively. At equilibrium (r=0) such rate equation leads to the Langmuir isotherm: ∗ = ∗1 + ∗ (2) where K=k1⁄k2 is the equilibrium constant. The equilibrium data showed in the following figures were used to evaluate Langmuir parameters as a function of the temperatures, using a non-linear regression. Table 2 shows Langmuir parameters for the adsorbing materials. Table 2 – Langmuir parameters for the adsorbing materials Material ωmax (-) K m3/g T °C HGR 8.4 x 10-2 581 120 2.2 x 10-2 421 150 1.1 x 10-2 200 200 Darco G60, Hg0 6.9 x 10-4 1202 90 1.9 x 10-4 668 120 1.05 x 10-4 420.8 150 Darco G60, HgCl2 1.13 x 10-1 594 120 3.87 x 10-2 477 150 2.53 x 10-2 168 200 Darco G60 + 7.8% Na2S 1.62 x 10-1 376 120 1.06 x 10-1 250 150 3.97 x 10-2 129 200 Figure 3 – Adsorption isotherms for HGR activated carbon at different temperatures. Figure 4 – Adsorption isotherms for Darco G60 activated carbon at different temperatures. 2464 Figure 3 and 4 show isotherm curves for two different solids, HGR and Darco G60, both not impregnated, as a function of elemental mercury concentration, for three different temperatures. These isotherms are of favorable kind and they show, apart from solid sorbent, that the higher the temperature, the lower the adsorption capacity, confirming the exothermic nature of the adsorption process under consideration. However, looking at y-axis range, the behavior of HGR solid seems to be much more favorable than Darco G60 one, because its adsorption capacity is about two order of magnitude higher than Darco G60. This peculiarity is probably due to the surface area value of HGR, which is about three times as Darco G60 alone, as showed in Table 1. Figure 5 – Adsorption isotherms for Darco G60 activated carbon at different temperatures. Figure 6 – Adsorption isotherms for Darco G60 impregnated with 7.8% Na2S activated carbon at different temperatures. Figure 5 and 6 make the comparison between isotherm curves for the same solid, Darco G60 is on the left, Darco G60 impregnated with 7.8% Na2S on the right. The curves are still showed for three different temperature, but the mercury specie investigated is the chloride type. The comparison between results relative to raw carbon and those obtained for impregnated one indicates that the addition of Na2S increases the adsorption capacity: this behavior is due to the formation of bonds between the adsorbate molecule and active sites distributed on the surface of the adsorbent, which favor the adsorption of mercury molecules. Finally, increasing the percentage of Na2S until 18.7%, not relevant gains in terms of mercury adsorption were detected. 4. Conclusions In this paper a comparison among different commercial solid sorbents, alone and impregnated, used in adsorption process, aimed to remove mercury compounds in flue gas, was showed. Results obtained demonstrated that adsorption process is always promoted in case of low temperature values. Furthermore HGR seems to be a more efficient solid sorbent in mercury removal than Darco G60. This last improves its performance, when impregnated with 7.8% Na2S, due to formation of bonds between the adsorbate molecule and active sites, even if higher Na2S percentage shows have poor effects in increasing adsorption process. References Bustard J., Durham M., Starns T., Lindsey C., Martin C., Schlager R., Baldrey K., 2004, Full-scale evaluation of sorbent injection for mercury control on coal-fired power plants, Fuel Processing Technology 85, 549- 562. Choi H.K., Lee S.H., Kim S.S., 2009, The effect of activated carbon injection rate on the removal of elemental mercury in a particulate collector with fabric filters, Fuel Processing Technology 90, 107-112. European Commission, Reg. CE n. 1881/2006. Granite E.J., Pennline H.W., Hargis R.A., 2000, Novel sorbents for mercury removal from flue gas, Ind. Eng. Chem. Res. 39, 1020–1029. Heidel B., Hilber M., Scheffknecht G., 2014, Impact of additives for enhanced sulfur dioxide removal on re- emissions of mercury in wet flue gas desulfurization, Applied Energy 114, 485–491. Jurng J., Lee T.G., Lee G.W., 2002, Mercury removal from incineration flue gas by organic and inorganic adsorbents, Chemosphere 47, 907–913. Karatza D., Lancia A., Musmarra D., 1998, Fly ash capture of mercuric chloride vapors from exhaust combustion gas, Environ Sci Technol 32, 3999-4004. 2465 Karatza D., Lancia A., Musmarra D., Pepe F., 1996a, Adsorption of metallic mercury on activated carbon, Twenty-Sixth Symposium (International) on Combustion/The Combustion Institute, 2439–2445. Karatza D., Lancia A., Musmarra D., Pepe F., Volpicelli G, 1996b, Kinetics of adsorption of mercuric chloride vapors on sulfur impregnated activated carbon, Combust. Sci and Tech 112, 163-l74. Karatza D., Lancia A., Musmarra D., Pepe F., Volpicelli G., 1996, Removal of mercuric chloride from flue gas by sulfur impregnated activated carbon, Hazard Waste Hazard Mater 13, 95-105. Karatza D., Lancia A., Musmarra D., Zucchini C., 2000, Study of mercury absorption and desorption on sulfur impregnated carbon, Experimental Thermal and Fluid Science 21, 150-155. Karatza D., Prisciandaro M., Lancia A., Musmarra D., 2011, Silver impregnated carbon for adsorption and desorption of elemental mercury vapors, Journal of Environmental Sciences 23(9), 1578–1584. Lancia A., Karatza D., Musmarra D., Pepe F., 1996, Adsorption of mercuric chloride from simulated incinerator exhaust gas by means of SorbalitTM particles, J Chem Eng Jpn 29, 939-946. Mazziotti di Celso G., Karatza D., Lancia A., Musmarra D., Prisciandaro M., 2013, Limestone-gypsum flue gas desulfurization process: modeling of catalyzed bisulfite oxidation, Chemical Engineering Transactions 32, 781-786. Musmarra D., Karatza D., Lancia A., Prisciandaro M., Mazziotti di Celso G., 2013, Adsorption of mercury chloride onto activated carbon on a new pilot scale plant, Chemical Engineering Transactions 32, 547-552. Pacyna E.G., Pacyna J.M., Steenhuisen F., Wilson S., 2006, Global anthropogenic mercury emission inventory for 2000, Atmospheric Environment 40, 4048–63. Pavlish J.H., 2009, Preface to the AQVI special issue of fuel processing technologies entitled: air quality VI: mercury, trace elements, SO3, particulate matter, and greenhouse gases, Fuel Process. Technol. 90, 1327–1332. Pavlish J.H., Sondreal E.A., Mann M.D., Olson E.S., Galbreath K.C., Laudal D.L., et al., 2003, Status review of mercury control options for coal-fired power plants, Fuel Process Technol 82(2–3), 89–165. Prisciandaro M., Mazziotti di Celso G., Vegliò F., 2003, Effect of burning supplementary waste fuels on the pollutant emissions by cement plants: a statistical analysis of process data, Resources, Conservation and Recycling 39, 161–184. Serre S.D., Silcox G.D., 2000, Adsorption of elemental mercury on the residual carbon in coal fly ash, Ind. Eng. Chem. Res. 39, 1723–1730. Uddin M.A., Yamada T., Ochiai R., 2008, Role of SO2 for elemental mercury removal from coal combustion flue gas by activated carbon, Energy Fuels 22, 2284–2289. Yang H., Xu Z., Fan M., Bland A.E., Judkins R.R., 2007, Adsorbents for capturing mercury in coal-fired boiler flue gas, J Hazard Mater 146(1-2), 1–11. Yao Y., Velpari V., Economy J., 2014, Design of sulfur treated activated carbon fibers for gas phase elemental mercury removal, Fuel 116, 560–565. Zheng Y., Jensen A.D., Windelin C., Jensen F., 2012, Review of technologies for mercury removal from flue gas from cement production processes, Progress in Energy and Combustion Science 38, 599-629. 2466