選煤廠介耗分析正文

第15頁(yè)中國(guó)礦業(yè)大學(xué)2011屆本科生畢業(yè)設(shè)計(jì)專(zhuān)業(yè)英文翻譯中國(guó)礦業(yè)大學(xué)畢業(yè)設(shè)計(jì)英語(yǔ)翻譯姓名：學(xué)號(hào)：學(xué)院：應(yīng)用技術(shù)學(xué)院專(zhuān)業(yè)：礦物加工工程翻譯題目：選煤廠介耗分析指導(dǎo)教師：職稱(chēng)：講師英文原文AnanalysisofmediumlossesincoalwashingplantsAbstractAmajoroperatingcostindense-mediumseparationisinreplacementoflostmediumsolids.Thelossofmediumsolids,beingcostly,playsacrucialroleindeterminingtheeconomicsofanypreparationoperation.Coalwasheriesthatemploydense-mediumcyclonesoftenattemptoptimizationoftheprocessesbyvaryingthevortexorthespigotdiameterandthefeedrelativedensity.Whilethesechangeshelpinclosercontroloftheseparationprocess,theyalsoresultinmediumlossesduetochangesinthemediumsplitratio(ratioofthemediumflowrateinoverflowtounderflow).Sincemediumsolidsarelostbyadhesiontoproductsandasmagneticseparatoreffluent,theeffectofthechangeinmediumsplitratioonthedrain-and-rinsescreensand,hence,themagneticseparatorcircuitneedstobestudied.InTataSteel'scoalwasheries,atJharkhandIndia,whichemploysprimaryandsecondarydense-mediumcyclonesinseriestoproducecleancoal,middlingsandrejects,reducingtherelativedensityoffeedmedium,hadaninsignificanteffectonthemediumsplitratio.Ontheotherhand,changingtheconeratio(ratiooftheoverflowdiametertotheunderflowdiameter)changedtherelativedensityandtheflowratesthroughthecycloneoutlets,thusaffectingtheperformanceofthemagnetiterecoverycircuit.Asystematicstudythroughlaboratorytestsandadetailedplantsamplingcampaignhelpedinidentifyingthecausesofmagnetiteloss.Uponimplementationoftherecommendations,themagnetitelossesdecreased,resultinginasavingofapproximatelyUS$27,500perannum.Thestudyalsohelpedinevolvingsomecheckpointsforplantoperatorsforidentifyingmagnetitelosses.Keywords:dense-mediumcyclone;magnetitelosses;drain-and-rinsescreens;magneticseparators1.IntroductionDense-medium(magnetiteinthecaseofcoal),aslurry/suspensionhavingarelativedensityintermediatetothatofvaluablemineralandgangue,isgenerallyusedasthemediumofseparationinmostcoalpreparationplants.Themedium,beingcostly,playsacrucialroleindeterminingtheeconomicsofanypreparationoperation.Dardis(1987)quotesfiguresof20–40%ofdense-mediumplantoperatingcostbeingattributabletomediumlossforplantsengagedinmineralseparationsemployingferrosiliconasthemedium.Thefigureis10–20%inthecaseofmagnetite.Thereareclearlyincentivestoreducethisloss,anditisoftenpossibletodosowithminimalcapitalexpenditurethroughimprovedoperatingproceduresandminorchangestoplantconfiguration.ThispaperdrawsontheexperiencefromonesuchstudycarriedoutbyR&DTataSteel,Jamshedpur,Jharkhand,Indiatoidentifysomeimportantoperatingissuesinmediumlossincoalwashingplantsandthefactorsinfluencingtheloss.2.Causesofmediumlossindense-mediumplantsTherearenormallyonlytwopossibleroutesbywhichmediumcanbelostfromtheplant:?adheredtotheproductsofseparation,afterdrainingandwashingonscreens;and?presentinthefinaleffluentfromthemediumregenerationprocess,usuallymagneticseparators,settlingconesorothersolid–liquidseparationdevices.Thecausesoflossfromthesesourcesareasfollows:?forcesofattractionbetweentheoreandmediumparticles,oreporosity,andinefficientwashing;?magneticseparationandclassificationinefficiencies;?corrosionandabrasionofthemedium,reportedforferro-siliconmedium;?excessivecircuitloadingsduringtheadditionoffreshmedium;?housekeeping(whenthefloorsarebeingcleanedandwashedoff);?plantdowntime(associatedwithhousekeeping);?mediumproperties(size,shape,magneticsusceptibility).Therehasbeenmuchworkdoneovertheyears,usuallybyoperatingplants,toidentifyandquantifythesourcesofmediumlossandtominimizeconsumption.Thetask,however,iscomplicatedbythedifficultyofdetermininganunequivocalmediumbalanceacrosstheplantbysamplingprocessstreams.Itisrarethatabalancethusestablished,forarelativelyshortoperatingduration,reflectsquantitativelytheactualconsumptionrecordedbytheplantovernormalreportingperiodssuchasamonthorayear.2.1.Factorsaffectinglossesthroughdrain-and-rinsescreensNapier-Munnetal.(1995),duringtheirinvestigationsoftheironorewashingplantsatMountNewmanAnincreaseinoperatingrelativedensityalsoledtosignificantincreasesinlosses.Mostoftheincreaseinlosswasattributedtothepoordrainagecharacteristicsofthehigherviscositymedium(Kitteletal.,1987).Asmallincreaseinrelativedensityledtoalargeincreaseinviscosityandthuspoorerdrainagecharacteristics.Thewashingarrangementwasalsofoundtoaffectmediumlossessignificantlythroughdrain-and-rinsescreens.Ofthevariouswashingarrangements,screenswithweirsandavigoroustumblingactionreducedthemagnetitelossesconsiderablycomparedtoslottedspraybarsandscreenswithfloodboxes.2.2.LossesthroughthemagneticseparatorsThereisnoconsensusintheliteratureastothecontributionwhichmagneticseparatorlossesmaketototalmediumlossindense-mediumplants.Dardis(1987),forexample,claimsthatmagneticseparatorsaccountformorethan75%oflosses,whereasMulder(1985)attributesonly18%tothissourcefortheSishenironoredense-mediumcyclones.Kitteletal.(1987)reportedmagneticseparatorlossesbetween2.4%and24%ofthetotalfortheMt.Newmandense-mediumcycloneplant.However,onoccasions,whenveryhighviscositymediawereused,substantialelevationoftheadhesionlosseswasobserved.Adhesiontocoalandthelossesinthemagneticseparatorarethetwomainroutesthroughwhichmagnetitegetslostinacoalwashingplant.Ingeneral,magneticseparatorsseemtocontribute20–40%ofthisloss,thoughthisproportionwillfallwhereadhesionlossesareabnormallyhigh,forexample,withporousores.Magneticseparatorsarethereforeanimportant,though,notnecessarily,adominantsourceofmediumloss.Sincetheirperformancecandeterioratemarkedlyifnotoperatedcorrectlyorproperlymaintained,theydeservecloseattention.AnalysisoflossesinmagneticseparatorscollectedinplantsurveysbyRayner(1994)suggeststhatthiscouldbeduetotheseparatorbeingoverloaded,intermsofeitheritsvolumetriccapacityor,lessoften,itsdrysolidscapacity.Hawker(1971)andSealyandHowell(1977)gaveloadinglimitsintermsofdrysolidsfeedrateofmagneticsandvolumetricflowrateoffeedslurry,whichcouldnotbeexceededwithoutlossofperformance.Dardis(1987)confirmedthattheoperatingvariables,whichaffectmagneticseparatorperformance,includepulpheight,magnetposition(angle),separationanddischargezonegaps,drumspeed,andmagneticstonon-magneticsratio.Lantto(1977),writingfromtheperspectiveofahardrockilmeniteconcentrator,explainedthattherecoveryinamagneticseparatorwasfeedqualitydependent.Healsogaverecommendationsforvariousseparatorparameters.BasedonoperatingexperienceattheIscorironoremines,DeVilliers(1983)observedthatoverloadingofthemagneticseparatorswasthemaincauseofmagneticlosses.HealsogavetheseparatorsettingsusedattheIscorplants.3.InvestigationsatTataSteel'scoalwasheriesTataSteelatJamshedpur,Jharkhand,Indiaownscaptivecoalwasheries,whichsupply60%ofcokingcoalrequirementsforitsintegratedsteelmakingoperations.Inthewasheries, Fig.1.

Schematicmediumrecoverycircuitandthesamplingpoints.Thedense-mediumandcleancoal(middlingsorrejects,asthecasemaybe)islaunderedtosievebendsandonesetofdrain-and-rinsescreens.Thesievebendsandthefirstsectionofeachdrain-and-rinsescreenareusedtodrainmediumfromthecoal;themediumiscollectedinscreenunder-pansandreturnedtotheprimarycyclonesumpviatheprimarycyclonemediumdistributionbox.Thesecondsectionofthescreensisusedtorinseanddrainthecoalfreeofadheringmedium.Thespraywatercontainingthedense-mediumrinsedfromthecoaliscollectedinthesecondsectionofthescreenunder-pansandreturnedtothedilutemediumsumpforsubsequentmagnetiterecovery.ThelevelofmagnetitewaterslurryinthedilutemediumsumpcanbeadjustedusingthePID(Proportional,Integral,Derivative3-termcontroller)loopprovidedforlevelcontrolandthemodulatingsplitteractuator.Whentheslurrylevelsinthesumprises,thesplitteractuatorwoulddiverttheflowawayfromthesystemtomaintainbalance.Theindicationloopalsogenerateshighandlowalarmlevelswithinthecontrolsystem.Thedilutemediumthuscollectedinthedilutemediumsumpispumpedtomagneticseparators,whichproducetherecoveredmagnetiteasover-densemediumandarejecttailingscircuit.Theover-densemediumisreturnedtotheover-densemediumsumpanddistributedtothedense-mediumwashingcircuitsasmakeup.Magneticseparatortailingsareusedasproductrinsingwater.Consideringtheoveralleconomicsofsteel-making,itwasthoughttoreducethecompositecleancoalashatthewasheriesfrom17%to16%startingApril2003.Withaviewtoachieving16%cleancoalash,thefollowingchangesweremadeinthecoalwashingplants:(a)Therelativedensityofmediumintheprimarycyclonecircuitwasreducedfrom1.36to1.3–1.33(b)Thespigotdiameterofthesecondarycycloneswasreducedfrom140

mmto125

mm.Thesechangeswouldhaveaneffectonthemediumsplitratio(ratioofthemediumflowrateinoverflowtounderflow)andhenceaneffectonmagnetiterecovery.3.1.EffectofreductionofprimaryrelativedensityonthemagnetiterecoverycircuitHeandLaskowski(1995)studiedthechangesinmediumsplitratiobychangingthevortexfinderdiameterandspigotdiameterandcycloneinletpressureattwodifferentmediumdensities.Atotalof27differentvortexfindersversusspigotdiametercombinationswerestudied.Thetestswerecarriedoutwithfourdifferentmagnetitecompositions.Thestudiesshowedthatatafixedinletpressure,therelationshipbetweenmediumsplitratioandconeratiowasindependentofmediumproperties.Extendingtheargumenttothechangeinmediumrelativedensityintheprimarycircuit,itwasconcludedthattherewouldbenochangeinmediumsplitratioduetothereductioninrelativedensityfrom1.36to1.3–1.33,andhencenegligibleeffectonthemediumlosses.3.2.EffectofreducingthespigotdiameterinthesecondarycircuitReducingthespigotdiameterofthecycloneswouldindirectlyincreasetheconeratio,i.e.,theratioofthediameterofvortexfindertothespigotthusaffectingthemediumsplit.Theflowrateofmediumthroughtheoverflowwouldincreaseandthatthroughtheunderflowdecrease,thusincreasingtheoverallmediumsplit.Changesintheconeratiowouldresultineitherlower/higherpulprelativedensityandhigher/lowerflowratesthroughthecycloneoutlets.Lowerpulprelativedensitywillhaveanegligibleeffectontheperformanceofthedrain-and-rinsescreens.However,lowerpulprelativedensityinthefeedtothemagneticseparatorwillinhibittheformationofflocs,whichhasbeenidentifiedasthemainprocessstepformagneticseparation.Accordingtothe“conceptualcollectionmechanism”modeldevelopedbyRaynerandNapier-Munn(2000),magneticseparationproceedsthroughtherapidformationofmagneticflocsorstringersassoonasthefeedslurryisexposedtothemagneticfield.Sizeperseofthemagnetiteparticlesisnotinvolvedinthisprocess,althoughmagneticsusceptibilityis.Asubstantialproportionofthemagneticsolidspresentwillbecomepartoftheseflocs.Theresidualmagneticsolidsarescavengedfromtheslurryduringitspassagethroughthecollectionzoneoftheseparator.Thisscavengingoccursbysolitaryparticlesjoiningexistingflocs,andisthereforeafirstorderrateprocessrelativetotheconcentrationofresidualmagneticparticles.Atlowpulprelativedensity,therateofflocculationistooslowforusefulflocstoform,andthuscaptureiseffectivelyasingleparticleprocess.Thiscausesadistinctdisadvantagetosmallparticles,whichmaybepreferentiallylost.Similarlyhigherpulprelativedensitywouldincreasetheviscosityofthemediumcomingoutthroughtheoutletsofthedense-mediumcyclone.Thisinturnwouldreducedrainagethroughthedrainscreensandincreaseadherenceofthemediumtothecoalsamples.Thisincreasedadherenceofmagnetitetocoalwoulddirectlyincreasethemagnetitelossafterrinsing.Withinthecapacityofthescreens,increase/decreaseinmediumflowratewouldnotaffecttheperformanceofthedrain-and-rinsescreens.However,increasedmediumflowratetothemagneticseparatorwouldreducetheresidencetimeandhencetherecoveryofmagnetics.Decreased/increasedmediumflowratetothemagneticseparatorwouldalsoaffectitsperformance.Thepooldepthinthemagneticseparatorneedstobemaintainedatanoptimumforanefficientmagneticseparation.Thiscanbedonebyadjustingthetailingsdischargeinthemagneticseparator.4.ExperimentationDetailedsamplingcampaignswerecarriedoutinthemagnetiterecoverycircuit.Samplesofcleancoal,middlingsandrejects,andoverflow,underflowandfeedtotheprimarymagneticseparatorandonesecondaryseparatorwerecollected.Thesesampleswerecollectedinfifteenincrementsoveraperiodofabout1

minusingstandardsamplers.Eachsamplewascollectedfromallpartsofeachflowstream,andallmaterialinthestreamhadthesameprobabilityofbeingcollected.About15

kgoftailingsand1

kgofconcentratewerecollectedoveraperiodof1

minbythissystem.Bothwetanddrysampleweightsweremeasuredusingappropriatelaboratorytechniques(Rayner,1999),andtherelativedensityoftheslurry(weight/volumeoftheslurry)andthesolidsconcentrationwerecalculated.ThemagneticfractionoftheconcentrateandtailingssamplewasrecoveredusingtheDavisTube.Thedatawerecollectedforamonth,takingcaretoavoidsamplingwhentherewereabnormaloperationsintheplant.Thesedataallowedthecalculationofconcentratesolidscontentandofmagneticsrecoveryforeachtrial.Themajoroperatingvariablesofamediumrecoveryscreenthatinfluenceitsperformanceare(i)theadhesionofthemediumpriortowashing,(ii)quantityofwashwater,(iii)screen-oreconveyingvelocity,(iv)screen-orebeddepth,(v)screeningduration,(vi)oresize,(vii)oreporosityand(viii)propertiesofthemediumpulp.Foranywashingplant,theoresizeremainsconstant,andtherankofcoaltreatedalsoremainsunchanged.Hencethefactorsthatwereconsideredforthelaboratorystudyweretheadhesionofthemediumpriortowashing,thetimeofwashing,screeningvelocity,relativedensityofmediumandthequantityofwashwater.Thelevelsofthevariablesmaintainedweresimilartothatintheplant.Thetestswerecarriedoutwiththecleancoal(consideredascoaltype1)andrejects(consideredascoaltype2)obtainedfromthewashery.Thewashwaterandthecoaltypewerevariedattwolevelsandalltheothervariablesatthreelevels.Alltogether108testswerecarriedout,theresultsofwhichareshowninTable1.Table1.Resultsofsimulationofdrain-and-rinsescreentestscarriedoutinthelabTimeofwashingRelativedensity-1Coaltype1Relativedensity-2Coaltype1Relativedensity-3Coaltype1Relativedensity-1Coaltype2Relativedensity-2Coaltype2Relativedensity-3Coaltype2Washwaterquantity1Screeningvelocity110.704.702.122.717.378.3220.544.191.581.433.6117.4230.0009.461.421.514.903.99Screeningvelocity210.555.081.631.975.374.9321.612.671.301.362.542.0731.091.300.781.331.961.82Screeningvelocity312.313.410.001.721.595.1522.872.500.911.511.471.3431.291.881.301.071.101.53Washwaterquantity2Screeningvelocity110.992.642.901.963.741.9721.062.931.971.141.802.7931.432.701.570.981.131.18Screeningvelocity211.233.431.740.601.151.4621.042.621.740.790.681.4930.912.131.351.080.380.44Screeningvelocity311.4612.601.350.952.101.3521.326.461.991.330.200.7131.533.390.931.390.590.62Figuresshowpercentageofmagnetitelostafterrinsing.Thelevelsofthevariablesarecodedandmentionedas1,2and3.“Washwater1”isthequantityofwashwaterwhichwascodedaslevel1,etc.Asimpletestprocedurewasadoptedtosimulatethedrain-and-rinsescreensectionofthemediumrecoverycircuitinthelaboratory.Mediawiththreedifferentrelativedensitiesweremanuallypreparedusingmagnetite(size:90%,?

0.045

mmand90%magnetics)collectedfromthewashery.Thecoalhadasizerangeof?

13

+0.5

mm,and50

gofsamplewastakenforeachtest.Inthiswork,theexperimentalprocedurewastoinitiallycontacttheorewiththemediumsoastoallowthemediumadhesiontooccur.Theoretobecontactedwasplacedinsideawirebasketandloweredintothere-circulatingmediumforabout30

sandthentakenout.The30-scontacttimeismuchmorethanthenormal4–5

scontacttimeobservedinacyclone.Thecontactedorewasthentakenoutofthewirebasketandplacedonalaboratoryscreenmountedonasieveshakertosimulatethemediumrecoverystepacrossavibratingmediumrecoveryscreen.Thesieveshakerwasoperatedwithoutwashwaterfor15

stoallowtheexcessmediumtodrainfreefromtheore.Asampleofthedrainedorewastakentodeterminethemassofthemediumadheringtotheorepriortothewashingstep.Thesamplewasthentakeninanopenbucketfilledwithwater,thevolumebeingtwicethatofthesample.Thesamplewithwaterwasthenmixedusinganimpellerforabout2

min.Thesamplewasthentakenout,dried,andweighed.Theadheredmediumfromthesamplewasthusrecoveredfromtheorebyscrubbingtheore.Therestoftheoreonthescreenwasfurthershakenforabout5–15

swithwashwatertowashofftheadheredmedium.Particlesthatwerestilladheredtotheoreafterthewashingstepwererecoveredfromtheorebyscrubbing,asmentionedabove.Thetestswerecarriedoutusingstatisticaldesignofexperimentsandtheresultsanalysedusingtheanalysisofvariance(ANOVA)technique.Thevariableswerestudiedatdifferentlevels,andthepercentageofmagnetitelostwascalculatedforeachtest.Thesumofsquares,themeansumofsquares,the‘F’ratioandthe‘p’valueswerecalculated.Thosevariableshavinga‘p’valueofmorethan0.05weretakenassignificantfactors.中文譯文選煤廠介耗分析摘要在重介質(zhì)分選中主要的操作損失是介耗。昂貴的介耗在決定介質(zhì)準(zhǔn)備中起著關(guān)鍵性的作用。選煤人員常常試圖通過(guò)改變渦流或者套管直徑和給料密度來(lái)優(yōu)化選煤過(guò)程。雖然這些變化有利于更好控制分離過(guò)程，但導(dǎo)致介質(zhì)流失是由于對(duì)介質(zhì)分離比率的改變（從溢流至底流的介質(zhì)比率）。由于介質(zhì)會(huì)黏附在產(chǎn)品上并且會(huì)隨磁選機(jī)流出，因此需要改進(jìn)磁選流程。印度jharkhand塔塔鋼鐵公司的選煤人員，使用一段和二段重介質(zhì)旋流器生產(chǎn)精煤，中礦和矸石。降低介質(zhì)的相對(duì)密度在介質(zhì)分離比率中起著一定的作用。另一方面，錐比的改變（溢流直徑與底流直徑之比）改變了相對(duì)密度和旋流器出口的流量，從而影響磁選機(jī)磁選的性能。通過(guò)實(shí)驗(yàn)室試驗(yàn)和機(jī)械設(shè)備采樣運(yùn)動(dòng)的系統(tǒng)研究有助于查明磁鐵礦損失的原因。經(jīng)研究表明，選煤廠介耗降低后每半年可節(jié)省美國(guó)大約27500美元。這項(xiàng)研究還有助于為正在擴(kuò)建中的選煤廠操作人員提供磁鐵礦損失的原因。關(guān)鍵字：重介質(zhì)旋流器；磁鐵礦損失；脫水篩；磁選機(jī)1、引言水煤漿懸浮液中有用的固體和煤矸石的重物（對(duì)煤炭中的磁鐵礦）在大多數(shù)選煤廠被普遍分離。昂貴的介質(zhì)在選煤廠決定經(jīng)濟(jì)成本時(shí)起著關(guān)鍵的作用。達(dá)迪斯（1987年）的報(bào)價(jià)數(shù)字中20-40％的重介質(zhì)選煤廠生成成本是由于選煤廠使用硅介質(zhì)而造成的介耗，對(duì)磁鐵礦則是10–20%。通過(guò)較小的投資改進(jìn)工作流程和改變?cè)O(shè)備型號(hào)能很好地減少這方面的損失。文章鼓勵(lì)以印度jharkhand塔塔鋼鐵公司賈姆謝普爾的經(jīng)驗(yàn)確定選煤廠影響介耗的因素。2、重介質(zhì)選煤廠介耗分析一般情況下，選煤設(shè)備中的介耗有兩種途徑：?排水和清洗工作后依然進(jìn)行產(chǎn)品分離；?介耗通常是因磁選機(jī)沉降錐或固液分離器受損引起。介耗的原因列舉如下：?礦石和介質(zhì)顆粒之間由于礦石空隙而產(chǎn)生的吸附作用；?磁選和分離效率低；?用硅做介質(zhì)造成的介質(zhì)腐蝕和磨損；?在補(bǔ)加新介質(zhì)時(shí)管路太多；?廠房清洗（樓層清理和清洗）；?設(shè)備停機(jī)檢修（與廠房清洗有關(guān)）；?介質(zhì)特性（大小，形狀，磁化率）。多年來(lái)選煤廠做了很多工作以確定介耗的數(shù)量并將介耗降到最低。然而，這種工作由于在礦漿中決定介質(zhì)平衡而存在的困難顯得很復(fù)雜。在相對(duì)較短的運(yùn)行期間這種固定的平衡是很難的，只能通過(guò)分選設(shè)備在一個(gè)月或一年的報(bào)導(dǎo)期反映數(shù)量上的消耗。2.1通過(guò)沖洗篩面等影響因素造成的損失皮爾-文基賢等人（1995年）在摩紐曼和湯姆選煤廠對(duì)鐵礦石的研究發(fā)現(xiàn)黏附損失會(huì)因篩面裝填而增加。負(fù)載篩面顯示超輕負(fù)載篩面的損失（以克/噸/米，篩面寬度表示）有明顯的增長(zhǎng)。操作時(shí)的相對(duì)密度加大也導(dǎo)致介耗明顯的增加。大多數(shù)的介耗增加是由于高粘度介質(zhì)的低水流特性（kittel等人，1987年）。相對(duì)密度小幅度的增加會(huì)導(dǎo)致粘性而且更低水流特性介質(zhì)較大的增長(zhǎng)。通過(guò)選煤流程發(fā)現(xiàn)排水管和沖洗篩面也會(huì)在介耗方面有加大的影響。各種選煤流程表明篩堰和劇烈的篩分與橫向開(kāi)槽給水管和有溢流堰的篩子相比能明顯地減少磁鐵礦的損失。2.2磁選設(shè)備介耗在文獻(xiàn)里關(guān)于選煤廠磁選設(shè)備介耗對(duì)總的介耗的影響沒(méi)有一致的觀點(diǎn)。例如，達(dá)迪斯（1987）認(rèn)為磁選設(shè)備的介耗多達(dá)75%以上，而穆德（1985）認(rèn)為對(duì)于選礦用重介質(zhì)旋流器只有18%的介耗。kittel等人（1987）則稱(chēng)在紐曼重介質(zhì)選煤廠磁選設(shè)備介耗在總的介質(zhì)的2.4%~24%之間。然而當(dāng)使用高粘度的介質(zhì)時(shí)就會(huì)觀察到大量介耗。介質(zhì)黏附到煤塊上和在磁選設(shè)備中的損失是選煤廠磁鐵礦損失的兩個(gè)主要原因。一般情況下磁選設(shè)備的介質(zhì)消耗占20%~40%，但這個(gè)比率會(huì)由于粘度損失異常高而下降，例如孔隙多的礦石。因此磁選設(shè)備雖不是必然但是重要突出的介耗設(shè)備。所以這些設(shè)備的性能會(huì)因不正確的操作或維修而明顯的惡化，因此應(yīng)對(duì)其進(jìn)行密切的關(guān)注。Rayner(1994)通過(guò)分析選煤廠磁選機(jī)中介耗后指出這種現(xiàn)象可能是由于磁選機(jī)所裝物料就其額定容積或?qū)嶋H容積而言超過(guò)了其處理范圍。Hawker(1971)、Sealy和Howell(1977)從固液比和（）給出了磁選機(jī)的裝填極限，這些介質(zhì)損失因素都不能超出范圍。Dardis(1987)指出影響磁選機(jī)分離性能的操作因素有pulpheight、磁鐵方位（角度）、separationanddischargezonegaps,drumspeed以及磁性物與非磁性物的比率。Lantto(1977)從鈦鐵礦收集器得出觀點(diǎn)認(rèn)為磁選機(jī)介質(zhì)回收率與介質(zhì)入料質(zhì)量有關(guān)。他還給出不同分選設(shè)備性能參數(shù)推薦數(shù)據(jù)。DeVilliers(1983)根據(jù)Iscor鐵礦礦井實(shí)際運(yùn)作經(jīng)驗(yàn)發(fā)現(xiàn)磁選機(jī)容積超載是磁介質(zhì)損失的主要原因。他同時(shí)列出了在該礦廠使用的分離設(shè)備。3、塔塔鋼鐵公司選煤廠調(diào)查印度jharkhand賈姆謝普爾塔塔鋼鐵公司擁有提供其煉鋼所需焦煤60%的煉焦選煤廠。在選煤廠ROMcoal經(jīng)過(guò)破碎后以0.5mm粒度界限篩分，大于0.5mm的煤經(jīng)過(guò)重介質(zhì)旋流器分選（初選），小于0.5mm的煤通過(guò)浮選流程分選（精選）。大于0.5mm的煤經(jīng)過(guò)旋流器一段產(chǎn)生相對(duì)密度低的（1.3-1.5）精煤，一段底流作為二段的入料產(chǎn)生相選密度高的中煤和矸石（1.6-1.9）。該廠的磁鐵礦回收循環(huán)是在每個(gè)選煤廠都有的循環(huán)。見(jiàn)圖一圖一磁鐵礦回收循環(huán)及采樣點(diǎn)圖重介質(zhì)和精煤（中煤或矸石）通過(guò)弧形篩和脫介篩脫介。弧形篩和第一組脫介篩用于從煤中脫除介質(zhì)，介質(zhì)在篩下收集后通過(guò)一段旋流器介質(zhì)桶又返回一段旋流器。第二組篩子是用來(lái)沖洗黏附在煤表面的介質(zhì)并將煤脫水。沖洗煤之后含重介質(zhì)的沖水被收集在二段脫介篩篩下并返回稀介質(zhì)桶用于之后的介質(zhì)回收?？梢酝ㄟ^(guò)PID介質(zhì)循環(huán)改變稀介質(zhì)桶中的磁鐵礦礦漿濃度。當(dāng)介質(zhì)桶中磁鐵礦濃度增加時(shí)，會(huì)減少介質(zhì)系統(tǒng)中的介質(zhì)以保持介質(zhì)系統(tǒng)的平衡。循環(huán)特征也導(dǎo)致介質(zhì)控制系統(tǒng)中介質(zhì)濃度的高低變化。稀介質(zhì)桶中的介質(zhì)通過(guò)泵打到磁選機(jī)中與里面的高濃度介質(zhì)混合產(chǎn)生循環(huán)介質(zhì)。高濃度介質(zhì)返回到高濃度介質(zhì)桶中以補(bǔ)償介質(zhì)密度。磁選機(jī)殘?jiān)魑驳V處理?？紤]到煉鋼的經(jīng)濟(jì)效益，該廠曾從2003年4月起將精煤灰分從17%降到16%。為了將精煤灰分降為16%，選煤廠做了如下的改進(jìn)：一段旋流器中介質(zhì)相對(duì)密度從1.36降為1.3-1.33；二段旋流器入料口直徑從140mm降為125mm。這些改變對(duì)介質(zhì)分離比率（從溢流至底流的介質(zhì)比率）有影響因此也將對(duì)介質(zhì)回收產(chǎn)生作用。3.1減少一段介質(zhì)相對(duì)密度對(duì)磁鐵礦回收的影響Laskowski(1995)在兩種不同密度介質(zhì)中通過(guò)改變旋流器溢流管管路直徑以及旋流器入料口壓力研究介質(zhì)分離比率的變化，總共研究了27組不同旋流器入料管直徑，這項(xiàng)測(cè)試用了四種不同的磁鐵礦。研究表明在入料口壓力不變的情況下，介質(zhì)分離比率和旋流器錐比對(duì)介質(zhì)特性是不變的。通過(guò)對(duì)改變一段介質(zhì)相對(duì)密度的爭(zhēng)論后得出結(jié)論認(rèn)為將相對(duì)密度從1.36降至1.3-1.33對(duì)介質(zhì)分離比率沒(méi)有影響，因此可以忽略對(duì)介耗造成的影響。3.2降低二段旋流器入料管直徑的影響降低旋流器入料管直徑會(huì)間接增加錐比，也就是說(shuō)旋流器溢流管直徑與管路直徑之比會(huì)影響介質(zhì)分離。通過(guò)溢流的介質(zhì)流量會(huì)增加而底流減少，從而影響總體介質(zhì)分離。改變錐比會(huì)導(dǎo)致旋流器出料礦漿相對(duì)密度低/高或高/低的變化。密度較低的礦漿對(duì)脫介篩影響很小，可以忽略。但是低密度的礦漿給入磁選機(jī)時(shí)將抑制磁選機(jī)中重要成分絮凝劑的性能。通過(guò)Rayner和Napier-Munn(2000)“conceptualcollectionmechanism”模式發(fā)展，磁選機(jī)快速形成的絮凝劑或者礦漿給料所帶來(lái)的經(jīng)濟(jì)效益很快就在磁選應(yīng)用領(lǐng)域流傳開(kāi)來(lái)。當(dāng)?shù)V漿通過(guò)磁選機(jī)的介質(zhì)收集區(qū)時(shí)剩余的磁性顆粒被排出。當(dāng)單一介質(zhì)顆粒黏附在絮凝劑表面上時(shí)顆粒就會(huì)被排出。礦漿相對(duì)密度低時(shí)，絮凝作用對(duì)絮凝劑與礦漿發(fā)生絮凝形成太慢了，因此絮凝作用就在單一介質(zhì)顆粒中進(jìn)行，這對(duì)易流失的細(xì)小顆粒產(chǎn)生明顯的負(fù)面影響。與此類(lèi)似相對(duì)密度更高的礦漿會(huì)增加重介質(zhì)旋流器出口介質(zhì)的粘度，這種變化將減少脫水篩的排水量而增加煤樣中介質(zhì)的粘性。這將直接增加沖洗磁鐵礦損失。在篩子的處理能力范圍內(nèi)增加或減少介質(zhì)循環(huán)比率不會(huì)影響脫介篩的性能，但是在磁選機(jī)中增加介質(zhì)循環(huán)比率將減少停留時(shí)間因此影響磁性介質(zhì)的回收率。減少或增加介質(zhì)循環(huán)比率也會(huì)影響磁選機(jī)的性能，磁選機(jī)水槽深度應(yīng)保持在最佳液面高度以保證高效的介質(zhì)分離。這可以通過(guò)調(diào)整磁選機(jī)中介質(zhì)排放量來(lái)調(diào)整。4.實(shí)驗(yàn)研究人員在磁鐵礦回收系統(tǒng)中進(jìn)行了詳細(xì)的抽樣實(shí)驗(yàn)，選取了精煤、中煤和矸石，以及溢流、底流、磁選一段入料和二段入料作為樣本。這些樣本通過(guò)一分鐘標(biāo)準(zhǔn)樣本增量的15%來(lái)收集，每個(gè)樣本在所有流體中抽取，流體中所有成分被采集的概率是一樣的。在一分鐘抽樣過(guò)程中大約有15kg尾礦和1kg精礦被收集。干濕樣本的重量用適當(dāng)?shù)膶?shí)驗(yàn)技術(shù)(Rayner,1999)測(cè)量，計(jì)算礦漿的相對(duì)密度（質(zhì)量/體積）和固體濃度。少數(shù)的精礦和尾礦磁性物樣本用DavisTube回收。樣本資料的收集要花一個(gè)月的時(shí)間，應(yīng)注意避免選煤廠有反?，F(xiàn)象時(shí)抽樣。這些資料反應(yīng)精礦固體含量和每個(gè)試驗(yàn)的磁性物回收量。影響脫介篩性能的主要因素有：1、介質(zhì)在潤(rùn)濕之前的粘度；2、洗選用水量；3、篩上物移動(dòng)速度；4、床層厚度；5、篩分時(shí)間；6、礦物粒度；7、礦物的孔隙度；8、礦漿特性。選煤廠的煤粒度是固定的，牌號(hào)也不會(huì)改變。因此實(shí)驗(yàn)室研究的影響因素就是介質(zhì)在潤(rùn)濕之前的粘度、篩分時(shí)間、篩分速度和介質(zhì)的相對(duì)密度以及洗選用水量。實(shí)驗(yàn)所用煤為從洗選設(shè)備中選取的精煤（1號(hào)）和矸石（2號(hào)）。洗煤用水和煤的牌號(hào)分為兩個(gè)等級(jí)，其它因素作為第三等級(jí)。實(shí)驗(yàn)總共做了108組，實(shí)驗(yàn)結(jié)果見(jiàn)表一。表一脫水脫介篩模擬實(shí)驗(yàn)結(jié)果表洗選時(shí)間一號(hào)煤相對(duì)密度1一號(hào)煤相對(duì)密度2一號(hào)煤相對(duì)密度3二號(hào)煤相對(duì)密度1二號(hào)煤相對(duì)密度2二號(hào)煤相對(duì)密度3洗選用水1篩分速度110.704.702.122.717.378.3220.544.191.581.433.6117.4230.0009.461.421.514.903.99篩分速度210.555.081.631.975.374.9321.612.671.301.362.542.0731.091.300.781.331.961.82篩分速度312.313.410.001.721.595.1522.872.500.911.511.471.3431.291.881.301.071.101.53洗選用水2篩分速度110.992.642.901.963.741.9721.062.931.971.141.802.7931.432.701.570.981.131.18篩分速度211.233.431.740.601.151.4621.042.621.740.790.681.4930.912.131.351.080.380.44篩分速度311.4612.601.350.952.101.3521.326.461.991.330.200.7131.533.390.931.390.590.62表中數(shù)據(jù)顯示洗選之后磁鐵礦損失的百分含量。影響因素分為1、2、3等級(jí)，“洗選用水量1”指一級(jí)洗選用水量。在實(shí)驗(yàn)室中可以用簡(jiǎn)單的實(shí)驗(yàn)步驟模仿介質(zhì)在脫介篩中的回收流程。在選煤廠用三種不同相對(duì)密度的介質(zhì)（粒度：90%?0.045

mm，磁性物含量：90%）收集磁鐵礦。每個(gè)實(shí)驗(yàn)用50g粒級(jí)在13-0.5mm之間的煤樣。在這個(gè)實(shí)驗(yàn)中第一步是將礦物與介質(zhì)接觸以使其粘附在一起，被粘附的礦物被帶入篩籃后進(jìn)入介質(zhì)循環(huán)系統(tǒng)30s后排出，這一時(shí)間比在旋流器里的4-5s要多。礦物被排出篩籃后又進(jìn)入振動(dòng)篩篩面通過(guò)脫介篩篩面震動(dòng)模擬介質(zhì)回收過(guò)程，振動(dòng)篩在沒(méi)有噴水的情況下運(yùn)行15s以使多余的介質(zhì)從礦物中排出。通過(guò)排出礦樣的多少推測(cè)出在洗選過(guò)程中介質(zhì)粘附礦物的量。然后將礦樣裝入盛有水的桶中，其體積是礦樣的兩倍。用葉輪攪拌裝置攪拌大約2min，之后取出礦樣，干燥，稱(chēng)重。通過(guò)對(duì)礦物的沖洗就可以回收礦物中粘附的介質(zhì)。篩面上的礦物通過(guò)大約5-15s的進(jìn)一步篩分、噴水以沖出被粘附的介質(zhì)，仍被粘附在礦物中的細(xì)小顆粒用上面所用的沖洗礦物的方法回收。實(shí)驗(yàn)采用統(tǒng)計(jì)方法設(shè)計(jì)實(shí)驗(yàn)，試驗(yàn)結(jié)果用方差分析方法(ANOVA)分析。從不同層次研究實(shí)驗(yàn)影響因素，磁鐵礦損失在每個(gè)實(shí)驗(yàn)中都進(jìn)行了計(jì)算。方差之和、“F”比率和“p”值都能計(jì)算出來(lái)。這些因素中“p”值超過(guò)0.05時(shí)將被看做是有意義的。目錄目錄第一章總論 1一、項(xiàng)目概述 1二、可行性研究報(bào)告編制依據(jù)和