Effect of unburned carbon content in fly ash on the retention of 12 elements out of coal-combustion

Availableonlineatwww.sciencedirect.coJOURNAL OF。 Science DirectENVIRONMENTALSCIENCESJournal of Environmental Sciences 2012, 24(9)1624-1629Effect of unburned carbon content in fly ash on the retention of 12 elements outof coal-combustion iue gasLucie Bartonova', Bohumir Cech, Lucie Ruppenthalova, Vendula MajvelderovaDagmar Juchelkova, Zdenek KlikaFaculty of metallurgy and Material Engineering, VSB-Technical University of Ostrava, T 17. listopadu 15, 708 33Ostrava- Poruba, Czech RepublicReceived 18 November 2011; revised 02 January 2012; accepted 03 January 2012AbstractThe aim of this study was to evaluate whether unburned carbon particles present in fly ash can help in the retention of S, Cl, Br, As, Se,Cu,Ni, Zn,Ga,Ge, Rb, and Pb out of flue gas during the coal combustion at fluidised-bed power station where the coal was combustedwell to be clear which factor governs behaviour of studied elements. Except of S(with significant association with CaO)and Rb and Pb(with major affinity to Al2O3)the statistically significant and positive correlation coefficients were calculated for the relations betweeunburned carbon content and Br(0.959), Cl(0.957), Cu(0.916), Se(0. 898), Ni(0. 866), As(0.861), Zn(0. 742), Ge(0.717), and Ga(0.588)content The results suggest that the unburned carbon is promising material in terms of flue gas cleaning even if contained inhighly calcareous fly ashesKey words: unburned carbon; coal combustion; fluidised bed; trace elements; emissions; retentionDOI:101016Sl001-0742(11)60981-9Introductionthe elements bound in solid particles are efficiently cap-tured in electrostatic precipitator(ESP) or fabric filterTotal worldwide quantity of coal combustion wastes pro- whereas the gaseous fraction of toxic elements easilyduced in large power stations(estimated in the early 1990s) penetrates through the particulate control device directlyexceeds undoubtedly 550 Mt per year(Adams et al., into atmosphere2005; Manz, 1997). Fraction of unburned carbon(UC)in This concerns above all the most volatile elements,coal ashes varies considerably according to ash character such as Hg, As, Cl, Br, or Se. Depending on the coal(bottom ash, fly ash etc )and mainly depending on coal characteristics and the operational conditions during therank and the type of combustion unit. Estimating average combustion, the fractions of these elements released intounburned carbon content in ash as 3%-there is 16.5 atmosphere in gaseous form can reach in some cases asMt of unburned carbon that is being produced annually. much as 80%-90% or even more( Germani and Zoller,Unburned carbon properties depend mainly upon coal rank 1998; Huang et al., 2004; Klein et al., 1975; Klika et alaccepted that unburned carbon is amaterial win enerally 2001; Meij et al, 1986, 1994; Xu et al, 2004; Yi et aland operational conditions. Nevertheless, it isporous structure which leads to an effort to use this waste This is the reason why many researchers directed theirmaterial as a low-cost adsorbent(Baltrus et al., 2001; effort toward the retention of these vaporized elements onBartohova et al. 2009, 2010: Batra et al., 2008; Bunt et some fly ash constituents or using another adsorbent addedal,2009;Hurt et al., 1995; Izquierdo and Rubio, 2008; during the combustion Polluting gas can be cleaned by flyMaroto-Valer et al., 2001, 2002; Xu et al., 2009)ash(harja et al., 2008 ), promising results were obtainedAnother important problem when fossil fuels are used by unburned carbon particles in fly ash( Bartonova etfor electricity generation is the emissions of toxic pol- 2007; Hassett and Eylands, 1999; Hower et al., 1999, 2000lutants. During the combustion, heavy metals and other 2010; Lopez Anton et al., 2007; Sakulpitakphon et altoxic trace elements originally present in coal are redis- 2000)or by中国煤化工 aluminosilicatetributed among bottom ash, Ay ash and emissions. The adsorbentsHuormost problematic is the distribution of vaporized elements et al., 2000;CNMHZhao et al., 2008)between the flue gas particles and gaseous phase because Serre and Silcox (2000), Hower et al.(1999, 2000,Corresponding author. E-mail: lucie. bartonova @vsb. cz2010), Hassett and Eylands (1999), and SakulpitakphonNo 9Effect of unburned carbon content in fly ash on the retention of 12 elements out of coal-combustion flue gas1625(2000)tested fly ashes from large-scale power stations for ESP hoppers) were further separated into particle-sizee retention of Hg from flue gas concluding that higher fractions using dry separation method on the followingunburned carbon content in fly ash is advantageous for sieves: 0.032, 0.036, 0.040, 0.050, 0.053, 0.056, 0.063,the retention of Hg. Unfortunately, works dealing with the 0.071, 0.080, 0.090, 0.100, and 0. 125 mm and the yieldspossible retention of other elements on unburned carbon of all the fractions were recordedparticles are rather rare(Bartonova et al., 2007; Lopez-In all fly ash samples( three bulks and all the parrticle.Anton et al., 2007; Vassilev et al., 2000)ize fractions) the unburned carbon content was deter-While the effect of unburned carbon is usually studied mined on Leco CS-244(the determination in based onusing real power-station fly ashes, the effect of other infrared absorption measurement). Prior to the measure-adsorbents is usually tested using bench-scale or pilot- ment of carbon content, the samples were leached withscale furnaces due to the possibility of changing more dilute HCI (1: 1, v: v) to dissolve carbonates, which couldparameters and conditions than it is possible in large power otherwise have misrepresented the results. In all ashstations. Based on the bench-scale results the positive samples the contents of major, minor and trace elementsinfluence of some adsorbents( Ca-based adsorbents, kaoli- were determined by polarized-beam X-ray fluorescencenite, alumina etc. on the retention of trace elements were spectrometry on SPECTRO XEPOS. Ash content in allreported Sterling and Helble, 2003; Furimsky, 2000; Chen the studied samples was determined gravimetrically inet al., 1999; Zhao et al., 2008)muffle furnace at 815C. Calculations of pair correlationThis work was focused on the evaluation of the effect coefficients(including the evaluation whether they are staof unburned carbon content on distribution of 12 selected tistically significant or not)was performed using Trilobyteelements(S, Cl, Br, As, Se, Cu, Ni, Zn, Ga, Ge, Rb, and Pb) statistical software QC Expert(v. 2.7)in particle-size fractions of fly ashes collected at three ESPhoppers at fluidised-bed coal combustion power station2 Results and discussionThere are at least three reasons why we decided to carryout this study: (1)this work is based on the real power- 2. 1 Analysis of fly ash samplesstation samples(where all real operational conditions inpower station are involved in the fly ash characteristics)Fly ash samples collected under the three hoppers of elec-(2)all these samples contain naturally also rather high Cao trostatic precipitator were dry-sieved providing altogethercontent(10%-25%), which enables us to take also this 26 particle-size fractions. All these fractions are listed in(maybe competitive)factor into account; (3)to the best Table 1. All the results for the three bulk fly ashes(3of our knowledge-the effect of unburned carbon (in real hoppers of ESP)and 26 particle-size fractions(preparedand at the same time rather calcareous fly ashes)on the from them) are also summarized in Table 1redistribution of as much as 12 elements have not beenAbove all, this study was focused to the evaluation ofSaudithe possible effect of UC content on the retention of IThus, the main aim of this work was to evaluate whether elements out of the flue gas: S, Cl, Br, As, Se, Cu, Ni,unburned carbon in fly ashes can(or can not) help with Ga, Ge, Pb, Zn, and Rb. Considering the results of otherthe retention of many toxic elements out of the coalresearchers(Furimsky, 2000; Chen et al 1999 Senior etcombustion flue gas even in highly calcareous ashesal., 2000; Sterling and Helble, 2003; Zhao et al, 2008),wecan conclude an effect of Ca-bearing additives or alumi1 Materials and methodsnosilicate adsorbents on the retention of studied elementsduring bench-scale combustion tests, the effect of CaO andThe samples were collected at Energetika TfinecAl2 O3 content in the studied samples was evaluated as wellfluidised-bed power station where lignite and lime- to be clear that the distribution of elements is given by thestone was combusted at 850C( the desulphurization ofemissions was achieved by means of dry desulphtDistribution of unburned carbon within the particle-sizetion method ) The four-sectioned electrostatic precipitator fraction of all the three fly ash samples is demonstrated(working at 142-145.C) is installed in this power station in Fig. la, while the same for CaO and Al2O is given infor the retention of solid particles out of the flue gasFigDuring the whole combustion test, at regular time inter-Columns plotted in Fig. la suggest significantly highervals the partial samples of the fly ashes from the first three UC content in fly ash samples collected under the 2nd andhoppers of electrostatic precipitator (ESP) were collected. 3rd hopper of ESp, average UC content in fly ash fromSince majority of solid particles is captured in the first the lst section is about 1%, the remaining two sectionsthree rows of the precipitator, the fly ash from the fourth produced fly ash with 4%-5% UC. Distribution of Caorow was not studied due to insufficient amount. Fly ashwas collected under the three hoppers separately, whichfractions) varies not so significantly as in case of UC.enables us to study three separate fly ash samples within 1 2.2 Correlatcombustion test. Each ash sample was then mixed properlyelementH中国煤化工sof12 studiedCNMH GaO, and Al20and using quartering method about 5-kg samples of all thecontentashes were set apart for the analysis in the laboratoryThe three fly ash samples(collected under the three Pair correlation coefficients r between the contents ofJournal of Environmental Sciences 2012, 24(9)1624-1629/ Lucie Bartonova et al.Table 1 Chemical analysis of bulk and particle-size fractions(mm) of fly ash collected at the lst, the 2nd and the 3rd hoppers(including the yield ofthe fraction, ash content and unburned carbon(UC)content)BULK I<00320032-0036-0040-0050-00530056-0063-0071-0.08000900.100>0.12500500.053005600630.1000.125125480.34198518111861178317.7818201905191721761992226023522557280053.049043046013420252402225224.7423.9923.2721.5922.7515.8321.1215.229631204005005Fe2O3(%)s(%)2.50258242201.73Cl(ppm)29126127025Ni(ppm)884882828Se(ppm)93.73.1.1(ppm)628559247578567.1BUK2<003200320036-0040-00500063>0080BULK3<00400040-0050-0.0560063->008000400537828.77758100.00925948593639066938993899267928792894.191960195619.7319.151955197020322045204720.11202540554203P205(%)0730700660.78078K20(%)1651678.82174216.230.06Fe2O3(%)2.392.512402371251o118713212113712383602665Br(ppm)4547063.5614368.768412 studied elements and UC, Cao, and Al2O3 contentsTable 2 Mutual (pair)correlation coefficients R for the relationsare given in Table 2. All these calculated correlationbetween the studied elements and unburned carbon (UC), CaO, andAlO, contentscoefficients were also subjected to evaluation whether theyare statistically significant or not(significance level usedCaOwas a=0.05). For total number of evaluated data of 29 s0.1070.7550.76(3 bulk samples and 26 particle-size fractions) the critical CI0.153value of correlation coefficient is 0.368(with increasing Ni0.358number of measured data this value somewhat decreases).Zn0.742For easier visualization of studied interrelations, the Ga0.5880.6550.321strength of mutual correlations of the 12 elements and Ge0.7170.4240.06UC and Cao content is demonstrated in Fig. 2, where As00123970.52the elements were divided into groups according to the Br0.113trength of their relation to UC and Cao0.600.693Statistically significant and positive values of correlatiePb0.1950.8340578oefficient concerning relations with UC were calculated *Bold-typedIHicallv evaluated as"significantCu, Zn, a, Ge, As, Se, and bi中国煤化工9 elements of 12 elements studied( they are all situated are higherCNMH Gcients obtained forin the right third of the diagram). The strongest positive Ge, Zn, arSUucwuat luwei (out still statisticallyrelation with unburned carbon was obtained for the most significant)volatile elements: Br, Cl, Cu, Se, Ni, and As(all R values Sulphur showed the only positive significant relationEffect of unburned carbon content in fly ash on the retention of 12 elements out of coal-combustion flue1627n且Rnn且,图,,o∽oIner爱喜gParticle-size fraction(mm)口caO日ALOEEEH§舀§旨§§18品g§8Particle-size fraction(mm)Fig. 1 Distribution of unburned carbon UC (a), Cao and Al O3 contents(b)within the particle-size fractions of fly ash collected at the Ist, 2nd and the3rd hoppers of electrostatic precipitator.0.6- Significant positiveSignificant positivecorrelation with CaoR3(A2O)=-0.76correlation with both0.5F and significant negativeone with uCSignificant positivecorrelation with CaO03Rs(AL2O3)=-0.520.1Significant negativeSignificant correlationcorrelation with UCneither with UCSignificant positive Ascorrelation with UC0.1nor with CaO8-0.4F Significant negaticorrelation with botSignificant negativecorrelation with CaO(AL2O3)=+0.14Rn(Al2O3)=+0.58Ra(Al2O3)=+0,32Significant positive correlation with UCRgb(A1O)=+0.91Pbnd significant negative one with Cao080.706-0.5-0.40.3-020.10.00.1020304050.60.70.80.91.01.1中国煤化工Fig. 2 Correlation coefficients for the relations of 12 elements and unburnedYHentsCNMHGwith Cao(r=0.755), it is the only element situated in the of vaporized SO2/SO3 with Ca-bearing minerals(calciuupper third of this diagram. This illustrates the interaction moxie, portlandite, calcite etc. ) The relation between S1628Jourmal of Environmental Sciences 2012, 24(9)1624-1629/Lucie Bartonova et aland uC was not of greater significanceof this fluidised-bed power station where also 1%-5% ofIn the case of Rb and Pb it was the only Al2O3 unburned carbon was present in fly ash the dominant factorcontent, for which positive and significant correlation governing the distribution of Cl, Ni, Cu, Zn, Ga, Ge, Ascoefficient was calculated. Some caution is needed in the Se, and Br was the unburned carbon content (while theinterpretation of this result, it should be reminded that effect of Cao content was not of greater significance)it is the retention of vaporized element out of flue gas The results obtained in this work undoubtedly suggest thatthat is studied in this work. On the assumption that the in this power station the condensation/adsorption of cl,elements are vaporized and then condensation/adsorption Ni, Cu, Zn, Ga, Ge, As, Se, and Br on unburned carbonof UC indicates some retention. However, not all of the with Ca-bearing minera f gfollows, their increasing contents with increasing content particles was preferred to the interaction of these elementselements studied in this work are typically volatile. Thetypical example of non-volatile element in this work is 3 ConclusionsRin os liches (w hice is the vend reason of its rat her lowThe main aim of this study was to evaluate whether thevolatility). Therefore its strong relation with Al2 O3 content unburned carbon particles present in fly ash can help within fly ash may be a consequence of its occurrence in the retention of Cl. Br. As, Se, Cu, Ni, Zn, Pb, rb, saluminosilicates in the coal combusted. Moreover, if very Ga, and Ge out of flue gas during the coal combustion athigh correlation coefficient between Rb and AlO3 had fluidised-bed power station where the coal was combustedbeen brought about by its condensation/adsorption out of with limestone. The main advantage of this work is thatthe flue gas, absolutely no interaction with UC under such the effect of unburned carbon was evaluated in real fly ashconditions would have been rather unlikelysamples containing also quite high Cao contentOn the other hand, rather complicated is the situation It was found out that except of S(with significantof Pb, because the volatility of this element varies a association with CaO), Rb and Pb(with major affinity tolot depending on its association in coal, including sul- Al2O3), the statistically significant and positive correlationphides, carbonates, but also aluminosilicates(Finkelman, coefficients were calculated for the relations between un-1994; Raask, 1985; Spears and Booth, 2002; Swaine and burned carbon content and Cl, Ni, Cu, Zn, Ga, Ge, As, SeGoodarzi, 1995). Therefore, in case of Pb both explana- and Br content. The effect of unburned carbon in fly ashtions are possible, either adsorption of vaporized form onto was significant and dominated even over the influence ofaluminosilicates or its occurrence in aluminosilicate matrix CaO content. This observation could be brought about byof coal taking Pb into aluminosilicate matrix of ashunfavourable time and temperature profile in the power sta-Except for S, Rb, and Pb the remaining 9 elements(Cl, tion for the interaction of these elements with CaO; underu, Zn, Ga, Ge, As, Sc, and Br) showed dominantassociation with UC in ash(despite rather high Cao elements onto UC, which was preferred behaviourcontent in these samples); this observation was obtainedeven in case of acid-forming clements, such as Cl, Br, As AcknowledgmentsSeIn this context it should be mentioned that the re- This paper was created in the project Nosults supporting the efficient retention of As (or other CZ. 1.05/2.1.00/01.0040"Regional Materials Sciervolatile elements) presented in the literature(Furimsky, and Technology Centre "within the frame of the operation2000; Chen et al., 1999; Sterling and Helble, 2003; Zhao programme"Research and Development for Innovationset al., 2008)were usually obtained on bench-scale of financed by the Structural Funds and from the state budgetpilot-scale combustion units where Ca-bearing adsorbents of the Czech Republic. The authors would also like towere tested without the occurrence of unburned carbon express their acknowledgement to OP Va vPi by projectgrains. Moreover, the combustion/adsorption conditions No ENET CZ.1.05/2.1.00/03.0069during these tests may have been somewhat different asin our fluidised-bed power station. The most common and Referencesthe most significant differences can be the combusttemperature, flue gas composition and the temperature Baltrus J P, Wells A w, Fauth D J, Diehl J R, White C M,profile during flue gas cooling because the time interval2001. Characterization of carbon concentrates from coavailable for the interaction of vaporized elements and thecombustion fly ash. Energy and Fuels, 15(2): 455-462.adsorbent may be the critical factor determining whether Bartohova L, Klika Z, Spears a D, 2007. Characterization of(even thermodynamically favored)reaction will actually berealized or notcombustion in CFBs. Fuel, 86(3): 455-463In other words, even if the reactions of some acid. Bartonova L, Kofistkova M, Klika Z, Kolat P, Simha-Martynlf elemental volatility in a fluidised bedforming elements(Cl, Br, As, or Se) with Ca-bearing中国煤化工 bon study. Chemicaldsorbent are thermodynamically favoured(which wasand Palso verified by bench-scale of pilot-scale combustion bartonovaCNMHGZ. Cech B 2010tests(Furimsky, 2000; Chen et al., 1999: Sterling andStudy of unburned carbon and volatility of elements duringHelble, 2003: Zhao et al., 2008), under the conditionsco-combustion of coal and waste alternative fuel. chemicalNo 9Effect of unburned carbon content in fly ash on the retention of 12 elements out of coal-combustion flueand Process Engineering, 31(4): 725-739ManzO e, 1997. Worldwide production of coal ash and utiliza-Batra V S. Urbonaite S. Svensson G. 2008. Characterization oftion in concrete and other products. Fuel, 76(8 spec iss.unburned carbon in bagasse fly ash FueL, 87(13-14): 2972-691-6962976.Maroto- Valer MM. Taulbee D N Hower J C. 2001. Characteriza-Bunt JR, Wagner N J, Waanders F B, 2009. Carbon particletion of differing forms of unburned carbon present in fly ashtype characterization of the carbon behaviour impacting oreparated by density gradient centrifugation. Fuel, 80(6)a commercial-scale Sasol-Lurgi FBDB gasifier Fuel, 88(5)5-800771-779Maroto-Valer MM, Andresen J M, Lu Z, Zhang Y, Jones AFinkelman R B, 1994. Modes of occurrence of potentiallyet al. 2002. Assessment of the commercial utilization ofhazardous elements in coal: levels of confidence, Fuelactivated carbons produced from high carbon fly ashesProcessing Technology, 39(1-3): 21-34In: Proceedings of the Technical Sessions Presented byFurimsky E, 2000. Characterization of trace element emissionsthe Conference on Unburned Carbon on Utility Fly Ashfrom coal combustion by equilibrium calculations. FuelPittsburgh, May 14, 2002.Processing Technology, 63(1): 29-44.Meij R, Janssen L H J M, van der Kooij J, 1986. Air pollutantGermani M S, Zoller WH, 1998. Vapor-phase concentrationsemissions from coal-fired power stations. Kema Scientificof arsenic, selenium, bromine, iodine, and mercury in theand Technical Reports, 4(6): 51-69stack of a coal-fired power plant. Environmental Science MeijR, 1994. Trace element behaviour in coal-fired power plantInd Technology, 22(9): 145-159uel Processing Technology, 39(1-3): 199-21Harja M, Barbuta M, Rusu L, Apostolescu N, 2008. Utilization of Raask E, 1985. The mode of occurrence and concentration ofcoal fly ash from power plants. I. Ash characterization. Entrace elements in coal. Progress in Energy and Combustionironmental Engineering and Management Journal, 7(3)Science, 11(2): 97-118289-293Sakulpitakphon T, Hower J C, Trimble A S, Schram W H,Hassett D J, Eylands K E, 1999. Mercury capture on coalThomas G A, 2000. Mercury capture by fly ash: study ofcombustion fly ash. Fuel, 78(2): 243-248.the combustion of a high-mercury coal at a utility boiler.Hower J C, Trimble A S, Eble C F, Palmer C A, Kolker A,Energy and Fuels, 14(3): 727-7331999. Characterization of Ay ash from low-sulphur and Senior C L, Helble J J, Sarofim A F, 2000. Emissions ofelements with particle size. Energy Sources, 21(6): ohigh-sulphur coal sources: partitioning of carbon and tracmercury, trace elements, and fine particles from stationarycombustion sources. Fuel Processing Technology, 65-66525263-288Hower J C, Maroto-Valer M, Taulbee D N, Sakulpitakphon T, Serre S, Silcox G, 2000. Adsorption of elemental mercury on the2000. Mercury capture by distinct fly ash carbon formsresidual carbon in coal fly ash. Industrial and EngineeringEnergy and Fuels, 14(1): 224-226Chemistry Research, 39(6): 1723-1730Hower J C, Senior C L, Suuberg E M, Hurt R H, Wilcox J Sterling R O, Helble J J, 2003. Reaction of arsenic vapourL, Olson E $, 2010. Mercury capture by native fly ashspecies with fly ash compounds: kinetics and speciation ofcarbons in coal-fired power plants. Progress in Energy andthe reaction with calcium silicates. Chemosphere, 51(10)Combustion Science, 36(4): 510-529Huang Y, Jin B, Zhong Z, Xiao R, Tang Z, Ren H, 2004. Trace Spears D A, Booth C A, 2002. The composition of size-elements(Mn, Cr, Pb, Se, Zn, Cd and Hg)in emissions fromfractionated pulverised coal and the trace element associa pulverized coal boiler Fuel Processing Technology, 86(1)ations.Flel,81(5):683-690.Swaine D J, Goodarzi F, 1995. In: Environmental Aspects ofHurt R H, Davis K A, Yang N Y C, Headley T J, Mitchell G D,Trace Elements in Coal. Kluwer Academic Publishers. The1995. Residual carbon from pulverized-coal-fired boilers: 2.Nederlands. 168Morphology and physicochemical properties, Fuel, 74(9): Vassilev S V, Eskenazy G M, Vassileva C G, 2000. Contents,1297-1306.modes of occurrence and behaviour of chlorine and bromineChen J C, Wea M Y, Ou wY, 1999. Capture of heavy metalsin combustion wastes from coal-fired power station. Fuel,by sorbents in incineration flue gas. Science of the Total79(8):923-937Environment, 228(1): 67-77.Ku M, Yan R, Zheng C, Qiao J, Han J, Sheng C, 2004. StatusIzquierdo M T, Rubio B, 2008. Carbon-enriched coal fly ash as aof trace element emission in a coal combustion process: aprecursor of activated carbons for SO2 removal. Journal ofreview. Fuel Processing Technology, 85(2-3): 215-237Hazardous materials, 155(1-2): 199-205Xu S, Zhou Z, Gao X, Yu G, Gong X, 2009. The gasificationKlein D H, Andren A W, Carter J A, Emery J F, Feldman C,reactivity of unburned carbon present in gasification slagFulkerson W et al., 1975. Pathways of thirty-seven tracefrom entrained-flow gasifier. Fuel Processing Technology,elements through coal-fired power plant. Environmental909):1062-1070Science and Technology, 9(10): 973-979Yi H, Hao J, Duan L, Tang X, Ning P, Li X, 2008. Fine particleKlika Z, Bartonova L, Spears A D, 2001. Effect of boiler outputand trace element emissions from an anthracite coal-firedon trace element partitioning during coal combustion in twopower plant equipped with a bag-house in China. Fuel,fluidised-bed power stations. Fuel, 80(7): 907-917.87(10-11):2050-2057Lopez-Anton M A, Diaz-Somoano M, Abad-Valle P, Matrinez- Zhao Y, Zhang J, Huang W, Wang Z, Li Y, Song D et al., 2008Tarazona MR, 2007. Mercury and selenium retention in flyArsenic emission during combustion of high arsenic coalsashes: Influence of unburned particle content. Fuel, 86(14fromspec iss. ) 2064-2070H中国煤化工 Energy ConversionCNMHG