Activation of Rejected Fly Ash Using Flue Gas Desulphurization (FGD) Sludge

Ⅴol.18、o.4Journal of Wuhan University of Technology - Mater. Sci. EdDec.2003Activation of Rejected Fly Ash Using Flue GasDesulphurization(FGD) Sludge06QIAO Xiu-chen POoN Chisun" LIN Zong-shouI The Hong Kong Polytechnic University Wuhan University of TechnologyReceived: April 17, 2003: Accepted Sept 23, 2003)Abstract: Lme-grade fly ash( rejected ly ash, rFA), a significnt portion of the pulverized fuel ash( PFA) Produced from coal-fired pouer plants and rejected fro the ash classifying press, remains walled due toits high carlon conten and large particle size(>45 -m). But it is thought that the rejected ash may howe potenial uses in chemie/ stabilisation/ solidification(S/S)Processes chich need relatively lower strengths and a lowerheminal reactivity. Flue Gas Desulphurisation( FCD)sludge is a bmy-puoduct of air polhution contrl equipncnu incoal fired power plants thase chemic eowroNition is mainly gypswn. As there is no effectire usage af loth of theseto materials, i is f interest to research on the possibe actiation of rFA using FCD. This paper presents experimental resulis of a stuly on the properties of rFA activited by the FGD in rFA-ccmerd pusles. Different percentagesof FCD were added irto the mix to stndy the effects of the FGD on the reaction uf the rfA wended cement pastesThe resalts showe that FGD takes effect as an activator only at late curing ages. Ailing Ca((H)2 enhances the ef-fert of F() on activing the hydration of rFA. Also, 10% FCD by weight of rFA is the optimal addition in theFA-cemetu pastes. The results of the comressive strength measurements correlate well with the porosity remitsKey words: rejected fly ash; PFA; strength derelopment actinator: fue gus destdjhIntroductionerate the dissolution of the glass phase of the fly ash particles Poon et al found that anhydrite was more effeclivethan gypsum in increasing the early-age sirength of the ce-Pulverized fuel ash( PFA)is a by-product generated ment/fly ash syslem. Results of Ma showed that thefimm buming coal during the generation of cleetrieity. Thepore sizes of untreated fly ash pastes were similar to thatfiner fraction of PFA (<45 m) produced by passing the of CasO, 2H20 aclivalerl lly ash/cement pastes. Ilye./used in building construction".2. But the reject of the reported that the addition of 0%o anhydrite by weightclassifying pmcess (ie, coarse fraction)is not suitable could double the strength of a cement/pfa paste. All of thefor use as a construction material, due to its high carboabove studies were conducted using the fine(classifiedcontent and large particle size(> 45un) In Hong Kong, portion of the fly ash. Few studies have been conducted onthis malerial is currently being dumped in large ash lahe activation and potential applicalions of the coarse(re-ons. But it is thought that the rejected ash may have po- jected)ashtential uses in chemical stabilization/solidification (S/Srocesses which require a relatively lower strengthed as a by-product by air pollution control equipmentstruction malerial have been widely reported. These studiesaphur dioxide and generales a sludge as a by-productnvolved alkali activation and sulfate activation ofThe chemieal composition is mainly gypsum. This FGD)of hydration of the ly ash particles. The former aims to process is adopted in many coal-fire power plants includ-le Lanma Power Plant operated by the Hong Kongvated alkaline envimnment to accelerate the hydration rea-Electric Co Itd. In Hong Kong, there is no effective uselion 3-7. The latter is based on the ability of sulfates to of the FGD sludge produced at presentreact with the aluminum phase in the ash particle to pro-The objective of this paper is lo sludy the effects ofduce ettringite(AFi)that contributes to the early strength FCD on the activation of rFA blended cement pastes.Theof the cementitious material 8-12I. Xu and Sarkar&reportresults of strength and microstructure developments of theed that the use of gypsum appeared to be useful to accelactivated fly ash-cement pusles are presented2 ExperimentalQIAO Xiu-chen(乔秀臣): Born in1973; lecturer; Master de.gree;Institute of Material Science and Technology, Wuhan Universit中国煤化工Funded by the Research Grants Council of Hong KongHCNMHGIhe Flue Gas desulProject No PolyU 5056/00F)were bantu iron two loeal coal firedVol 18 No 4QIAO Xiu- then et al: Activation of Rejected Fly Ash Using Flue Gasver plants. Fig. I shows their X-ray diffraction(XRD) commercially available ordinary Portland cenent(OPC)pattems. It can be observed that besides the main gypsun equivalent to ASTM Type and hydrated lime(Capeaks, there are some CacO, and Ca( OII), contents in (OI1)2)were used as binders. The chemical and physicalFGD sludge. The rFA coresponds to ASTM Class F properties of the rFA, FGD and the eement are given inand its particle size distribution is shown in Fig. 2. A Table I and Table 2Q: QuartzM: MulliteCH: Ca(OHh-FAi<007500750.150.30.61.182361000200030.0040.005000600ig. 1 X-ray diffraction patterms of the rFA and FD)Fig 2 Particle size distrbution of the rfATable I Chemical Compositions of Materials Used in ExperimentFe, Or -PFA47.238.4224.548.28).398.0060.06320121.3Table 2 Physical Properties of Materials Used in119.02.2 Preparation of specimensThirty-five mixes were prepared with a water-to-solidratio(w/s)of 0. 35 according to Table 3. The relativeCH25100.025.025proportion of rFA to cement was kept at 100: 25 in all thMixes.The Ca( OH), content in the Fgroup wasFGD0100.010.025ased fron 10% to 30% by weight of rFA. In the Ggroup the FGD content was varied from 10% to 30% byweight of rFA. Based on the proportion of each mix in the10-0.010—3.0-5100.0Fgroup, the FGD content varied from 10% to 30% byCHIO [HI0-F)eight of rFA in CHIO, CIl15, CH20, CH25 and CH30LHIG-IID5All the pastes( shown in Table 3 )were first mixedfor 10 minutes in a mechanical mixer at room tempeture. Water was added after the first minute of mixing. Themixes were then casl into plastic vials(A40 x 80 mm)Cl0and compacted with a plastic od until all entrapped airas removed. ' Ihe prepared samples were placed into a fogtank at 25 C. The samples were rcmoved from the plastivials after 7, 28 and 90 days of curing an subjected toH-H2100.025,025the following3)025.02. 3 Compressive strength testAt the given ages, three specimens per mix were sub-yh中国煤化工33jectDenicompression machine. ' The ends of these specimens wereCNMHGens after the conground to ensure that the end surfaces are flat and parallel prre preserved for other tests. To stop thebefore lesting. The strength results shown are the average of hydration reactions, the fractured pieces of the pastes werethe three specimens with variations of not more than J0%. soaked in acetone for a tolal of 14 days, with the arcloneJournal of Wuhan University of Technology- Mater. Sci. EdDhx.2003changed after the first 7 days. Then they were dried at60℃for48 hrs in a vac3 Results2.4 X-ray diffraction analysisThe preserved samples obtained above were ground 3. 1 Compressive strength developmentand used for XRD analvsis. A Bruker d8 advanced scanIt can be found from Table 4 that the compressivener was employed and run at a 20 of 0. (4 per stepstrength decreases with the: increase of Ca( OI1),conlent2.5 Mercury intrusion porosimetry(MIP)at 7 days. 'This may be due to the reaction of rFa onlyThe samples for MIP were those ungrounded frac began after 28 days and the lower overall cement conltured pieces obtained after the conpression test. They ent when Ca(OH), was added. At 28 days, the strength ofwere measured by a Pore Sizer 9320 mecury intrusion po- mix CH20 shows a highest strength value although it isdosimeter with a maximum mercury intrusion pressure of still lower than the control (i e, mix withoul Ca( OH)210 MPa. A cylindrical pore geometry and a contact angle dition ). Al 90 days, all the strength values ofof 140 were assumal-". The mercury intruded pore di- (except for mix CH30)are higher than theameter was calculated by dp=-4yc0s A/P(where y= 4)0.483 Nm, the surface lension of mercuryTable 4 The Compressive Strengths of Diferent Samplesoppressive StrengthenSive StrengthCEllO6.825.10CHIS- CIMo5.96.CHIS3.857.3(:H15-255H2022.64CIII5-FGD3O5.7CH2SCIM0-FCDIO112D-FGDISCH2D-FGmOCH2D-FGin5CH25-FrDIO6CH|O-卜CD20225,14.70CH0-FGDLSCIlIO. FCD305.5CH30-FGI2OCTUS FCDIO2.97.2CHG0-FG25CHIS-FCDIs6.617.90CH30-FGDthe specimens with CH contents, the mixes containirstrength values of G group decrease with the increase in 10% FGD by weight of rFA show highest strength valuesFGD content al all euring ages. The strength values of the which are also higher than those with only Ca(OH),ormixes with FGD addition are lower than the control at 7 FCD additand 28 days. But at 90 days, the strength of mix FGD10 of 3. 2 X-ray diffraction analysisis higher than the controlFig 3 shows that the addition of Cu(oh ),aloneThe effects of FGD on the compressive strength of (mix CH20) has little effect on the hydration of rFA.Butthe mixes with Ca(OH)2 are shown in Table 4. Similar to the addition of FGD alone(mixes FGD10 and FGD25the results of the G group, the strength values in general accelerates the reaction of both cement and the rFA. Thdecrease with the increase in FGD content. At 90 days, for gypsum peaks almost disappear in mix FCD10( Fig 3)te G: ypsM: MulliteCH Ca(OH), 2: QuartyCI20-7CH2H+G1330-7人、ED7MwtCI20-FGD25-7AFGDIO-7AAhl, CH20-FGD20-7vy+和小…,2,山 Control-MAAL Cll20-FGD15-751015202530中国煤化工+4 CH20-FCD10-7CNMHGFig 3 X-ray diffraction pattems of FA specimens withonly Ca(OH)2 or FGD)at 7 daysLoth Cu(OH), and FCD ut 7 daysol.18N.4QLAO Xiu-chen et al: Activation of Rejected Fly Ash Using Flue GasCH: Ca(OH) Q: QuartCa(OH), Q Quartz M: MulliteGMEMullitoAFt G MCHG QCH20-FGD30-28Nuw/N FCI25-28AW/NN FGDIO-28HAMNWCH20-FGD20-28n小小 wr+.FCD20-28偏小ww认小 whw Control-28H20FGD10-282-Thetaonly Ca( O11), or FGD at 28 duvs turnsFig 6 X-my diffraction patterns of rfA specimens withboth Ca( OH), and FGD)at 28 daysr-FA: CH: FGD: Cement三0.3000一F310020025)F3(100:20:0:25)030:(H2(00230.2000120-FCD102010250.2000G(1000:1025)CH20-FG25100:20:25:250.1500G1(1000:1025)G41000:25:25)+ Cantrol(100: 0: 0:亠 Control(10:0.000100.100Diameter/micrometersFig 7 Effexts of FGD on pore size: distribu ion( (a)28 days, (b)al 90 daysTable 5 Average Pore Diameters and Porosity of SpecimensCa(oH),Average pore diameter/pRosity/( %, r/v)PGDt90 laysCnta0.0213,69FCDIO0.0262540.32CIL2O-FGDIOCH20-H254.O558It can be noticed from Fig 4 that the addition of increasing amounts of FGD weakens the main peak of the r- 4 DiscussionFAs quartz and increases the gypsurn peaks. The XRDpAllens show there is no obvious fomation of AFt in allThe results of this study indicate that the addition ofthe mixes at 7 days. However, the AFt peak can be ob- FGD can have significant elects on the development ofserved at 28 days( shown in Fig. 5 and Fig. 6). The peak the compressive strength, the formation of hydration prodintcnsity of AFt in the mix FC25 is stronger than that in ucls and the pore size distribution of the r-FA-cerment sys-mix FCDIO(Fig 5). For the specimens with Ca(OH)2 tem, particularly at the latter curing agesand FCD, the formation of AFt can orly be found in mixes 4. At 7 daysCH20-FGI20 and CH20-FGIn25( Fig 6). This tmay meanXu reported that at early ages, due to the differthat an appropriate ratio of Ca( OH)2 to FGD is required cnce in solubility of Ca(OH)2 (4. 68 x 10)and CaSAfor the formalion of AFt(7. 10 x 10-), a lower Ca concentration for gypsum3.3 Pore size distributionprecipitation is resulted Gypsum would not be: an ellexliveThe influence of FGD on the pore size is shown in activator until it is first dissolved in solution, which onlyl'able 5. The pore size distribution pattems are shown in occurs when Ca( OH)2 formation has been slowed downFig. 7. it can be noticed from Table 5 that the porosities of Fom Fig 3, it can be found that no obvious fonnation ofthe specimens decrease as the curing ages increase For Ca(OH), in the control sample can be detected at 7the specimens with FGD addition, the porosities incrcase So中国煤化工 cannot activatewith the increase in FGD) content at all euring ages. At 90 rFAys, compared with the spec imens without FCD addit- of GCNMHGIrend withil,10% FGD contenl by weight of rFA decreases the crease in FGD content. Due to the low reactivityporosity of spcimenbcfore 7 davs sl, the strength of samples decreases withJoumal of Wuhan Universily uf lec hnology-Mater. S: i FAhe increase in FGD content even when Ca(OH)2 is added Research Grants Council of Hong Kong( Project Nlo the samples( shown in Table 4)5056/00F). The authors also would like to thank4.2 At 28 daysChina Light and Power Co. Ltd and Hong Kong Electricg 8 shows that the addition of FGD enhances the Co. Ltd respectively for providing the: PFA and FCI)saul-formation of AFt at 28 days. Moreover, the: formation of plesAFt inereases with the inerease in Fgd contentstrengths of the samples with FGD decrease with the in- Referencescrcase in FGD content (shown in Table 4). Thto not all the FGD added to the samples can contribute tows langley, GG Carette, v M Milholra Stnelural Coterelthe hydration reaction. There remains a signifieantincorporating Hiph Volumes of ASIM Class F Fly Ash. ACTof unused FGD with little contribution to the strength en- 2 Horiuchi s, Kawaguchi M, Yasuhara k, E.ective Use of Fly Ashhancement. This is consistent with the results of the porusSlurry as Fill Matcrial. Journal of Hazardous Materials, 76ity study which shows the porosity increases with the in-2000:301-337crease in FGD content(shown in Table 5)3 Cuijun Shi, Robert L, Day. Acceleration of the Reactivity of Fly4. 3 At 90 daysAsh hy Chemical Activation (em Conor Res, 25, 1995 15-21From 'Fable 4, it can be noticed that the strength val-4 Qian Jueshi shi Cajun, Wang, Zh. Activation of Blened (e-ues of the samples increase significantly at 90 days. Moreover, the strength development of the samples with Ca 5 Caijun Shi, Robert 1. Day. Acceleralion of the Reaclivily of Fly(OI1), is higher than those without il. This means that theAsh ly Chemical Activation. Cem. Concr. Res, 25, 1995: t5presence of Ca( OH)2 benefits the strength development. It21be because the addition of Ca( OH)2 can provid6 A Katz, Microscopie Study of Alkali-Activater Fly Ash. CemCa"to form the hydration products. the results in Table 7 AL A Fray, JM Bijen, Y M de Haan. The Reaction of Fly-ashConcr,Res,,28,1998:197-x084 also show that the strength values of the samples with10% FGD by weight of rFA are much higher than that ofin Correte A Critical-examinalion. (m. Concr. Res, 19, 1989235-246the control sample. This is consistent with the results of8 Xu Aimin, Shondeep I sarkar. Micnulructural Stidy of Gypillmhe porosity study (table 5)which shows that the porosiActivaled Fly Ash Hydration in Cement Paste. Cem. Concties of the samples with 10% FGD are lower than that ofRr.,21,1991:1137-1147the control sample. It means that 10% FGD by weight of 9 CSPoon, SC Kou, I lam, Z s lin Activation of Fly tsh/(-rfA is the optimum addition in the samplesshowsment Lsing(alcium Sulfate Anhydrate( CaSO4 ).Cen. Cncethat the addition of FGD is uscful to the strength developRes.,31,2001:873-88Iment of the rFA- cement pastes at late euring agel0 Weiping Ma, Chunling Liu, Paul W Brown, Sridhar KomarmeniPore Structurcs of Fly Ashes Activated by Ca( OH)> and Ca-5 ConclusionsS0:, 2H,O Cem. Corr, Res,, 25, 1995: 417-425II A F L. Hyde. Some Preliminary Test Data on Cernent-pfa-an-h ydrite: Mixes. Magazine of Concrete Research, 36, 1984: 174a) FGD is useful to the strength development of rFAcement pastes and FDG begins to take elfect as an aetiva- 12 Weiping Ma and Paul W Brown. Hydration Reactions of Flytor omly al lateAsh with Ca(OH), and CaS04. 2I, 0. Co, ConeT. Res, 27b)10% FGD by weight of rFA is the optimum addit-1997:1237-1248ion in the rFA- cement pastes13 CS Poon, XC Qiao, Z S Lin Pozzolanic Properties of Rejectc)The presence of Ca( OH)2 enhances the activatingFly Ash in Blended Cement Pastes. Com. Concr, Res, 32003, accepted for publicationeffect of FGD14 FM lea Cement and Conerete Chemistry. Third edition, LondonArnold. 1970AcknowledgementThe work described in the paper is funded by the中国煤化工CNMHG