Click here to load reader

Decanter Centrifuge Handbook Alan Records 2001

  • View

  • Download

Embed Size (px)


Decanter Centrifuge Handbook Alan Records 2001

Text of Decanter Centrifuge Handbook Alan Records 2001

DecanterCentrifuge Handbook 1st Edition ThisPageIntentionallyLeftBlankDecanter Centrifuge Handbook 1stEdition Alan Records Ken Sutherland EL SEVI ER ADVANCED TECHNOLOGY UK USA JAPAN Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB. UK Elsevier Science Inc. 665 Avenue of the Americas, New York, NY 10010,USA Elsevier Science J apan, Tsunashima Building Annex, 3-20-12 Yushima, Bunkyo-ku, Tokyo 113, J apan Copyright Q 2001 Elsevier Science Ltd. All rights reserved. No part ofthis publication may be reproduced, storedinaretrievalsystemortransmittedinanyformorby anymeans:electronic, electrostatic, magnetictape,mechanical, photocopying, recording or otherwise, without permission in writing from the publishers. First edition 2001 Library of Congress Cataloging-in-Publication Data Records, Alan Decanter centrifuge handbook / Alan Records, Includes index. ISBN 1-8 5 6 17-369-0 (hardcover) 1. Centrifuges-Handbooks,manuals, etc. 2. Centrifugation- Ken Suther1and.-1st ed.p. cm. Handbooks, manuals, etc. I.Sutherland, Ken. 11. Title. QD54.C4 R43 2000 660'.2842-d~2100-049 524 British Library Cataloguing in Publication Data A catalogue record for this title is available from the British Library. ISBN 185617369 0 No responsibility is assumed by the Publisher for any injury and/or damage topersonsorpropertyasamatterofproducts liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Published by Elsevier Advanced Technology, The Boulevard, Langford Lane, Kidlington, Oxford OX5 l GB, UK Tel: +44(0) 1865 843000 Fax: +44(0) 1865 843971 Typeset by Variorum Publishing Ltd, Rugby Transferred to digital printing 2005 Printed and bound by Antony Rowe Ltd, Eastboume CONTENTS Preface and Acknowledgementsxiii Chapter 1Introduction 1.1The Decanter Centrifuge 1.1.1The basic decanter 1.1.2Separation principle 1.1.3Decanter applications The History of the Decanter 1.2.1Origins 1.2.2Machine and application development 1.3Decanter Manufacturers 1.4Present Trends 1.5References 1.2 Chapter 2Decanter Design 2.1Basic Construction 2.2Basic Components 2.2.1Orientation 2.2.2Flow 2.2.3Materials of construction 2.2.4Bowl hub weirs hub bearing hub and bearings discharge 2.2.5Beach 2.2.6Conveyor 2 2 3 5 6 6 8 10 13 14 17 19 19 19 21 21 22 22 23 24 25 26 28 28 29 29 31 31 vi 2.2.7 2.2.8 2.2.9 2.2.10 2.2.11 2.2.12 Gearbox Frame Casing Floc/rinse zone Wear protection Conveyor bearings and seals Bearing supports Feed tube Vibrationisolators Casing baffles Cake discharge Centrate discharge Casing seals Vents Sub-frame Main drive Back-drive 33 33 34 36 37 38 38 39 40 41 41 42 42 42 43 43 45 2.3Variations to Main Components47 2.3.1 2.3.2 2.3.3 2.3.4 2.3.5 2.3.6 2.3.7 2.3.8 Orientation47 vs. horizontal47 decanter seals and bearings 49 decanter casing seal51 Flow51 Materials of construction52 Bowl variants54 hub54 weirs55 bearing58 Beach59 hub61 discharge61 liner64 Conveyor64 hub66 zone69 2.3.6,sWear protection71 and seals73 Gearbox73 Frame76 supports76 tube76 2.3.9 2.3.10 2.3.11 2.3.12 isolators Casing discharge discharge seals Sub-frame Main drive Back-drive 2.4Special Features 2.4.1 2.4.2 2.4.3 2.4.4 2.4.5 2.4.6 2.4.7 2.4.8 2.4.9 2.4.10 Basic construction decanter decanter countercurrent extractor for temperature and cantilevered bowl hubless conveyor decanter dual beach decanter Centripetal pump Skimmer pipe Centrate weir design dam dam Noise suppression Bowl baffles baffledisc disc baffle baffle Clarification enhancement flow axial flow Conveyor rake Conveyor tiles Conveyor pitch pitch decanter pressure extremes vii 77 77 77 78 79 79 80 80 80 82 86 86 86 86 89 90 90 90 90 92 93 95 96 96 96 97 97 99 99 100 101 102 103 104 104 105 105 106 107 108 109 109 viii pitch Counterbalance and scraper flights2.4.1 1 2.4.12Feedzone 2.4.13The reslurry collector 2.4.14CIP 2.4.1 5The Rotodiff 2.4.16Power regeneration 2.4.1 7 2.4.18Floating conveyor 2.4.19Decanter controls Dual main drive motor 2.5References Chapter 3Applications 3.1Application Classes 3.2Application Analysis 3.3Waste Sludge Processing 3.3.1Industrial wastes 3.3.2Water treatment sludges 3.3.3Municipal sewage treatment 3.4Energy Materials Production 3.5Processed Fuels 3.6Minerals Extraction and Processing 3.7Food and Food By-products 3.7.1 3.7.2Fish processing 3.7.3Fruit andvegetable products 3.7.4Other food processing Meat and meat products processing 3.8Beverages 3.9The Chemicals Industry 3.9.1Bulk inorganic chemicals 3.9.2Bulk organic chemicals 3.9.3Fine and household chemicals 3.9.4Pharmaceutical and medicinal chemicals 3.10Other Applications Chapter 4Decanter Theory 4.1Basic Theories 4.1.1Acceleration force 4.1.2Differential 4.1.3Conveyor torque 4.1.4Process performance calculations 4.2Particle Size Distribution 4.3Clarification 4.3.1Sigma theory 110 110 112 113 114 114 115 116 116 116 118 122 125 127 127 129 129 132 134 135 136 136 137 138 140 141 142 143 143 144 144 146 149 149 150 151 151 154 159 159 4.4 4.5 4.6 4.7 4.8 4.9 4.10 4.1 1 4.12 4.13 4.14 4.15 4.3.2Sigma enhancement 4.3.3Flocculant requirement Classification Three-Phase Separation Thickening Conveying 4.7.1TheBeta theory 4.7.2Conveying on the beach 4.7.3Dry solids conveying Conveyor Torque Dewatering and Washing 4.9.1Solids dewatering 4.9.2Washing 4.9.3Solids compaction Dry Solids Operation Fluid Dynamics 4.1 1.1Reynolds number 4.11.2Moving layer 4.1 1.3Cresting 4.1 1.4Feed zone acceleration Power Consumption 4.12.1Main motor sizing 4.12.2Main motor acceleration Mechanical Design 4.1 3.1Maximum bowl speed 4.1 3.2Critical speeds 4.13.3Liquid instability problems 4.13.4Length/diameter ratio 4.13.5Bearing life 4.13.6Gearboxlife 4.13.7Feedtube Nomenclature References Chapter 5Flocculation 5.1The Principle of Flocculation 5.2Polymer Solution Make-up 5.2.1Dissolving solid polymers 5.2.2Diluting dispersions 5.2.3Final flocculant solution characteristics 5.3Polymer Choice 5.4Pretreatment 5.5Admitting Flocculant to the Decanter IX 165 166 167 168 170 173 175 175 176 177 179 180 180 181 185 186 192 192 194 194 195 196 197 198 200 200 202 203 204 204 206 206 208 213 217 220 220 221 222 225 229 230 X 5.6Flocculant Suppliers 5.7Low-Toxicity Polymers 5.8Applications 5.9Performance 5.10References Chapter 6TestWork and Data 6.1Test Equipment 6.2Test Procedures 6.3TestLog 6.4SomeTest Data 6.4.1Spent grain 6.4.2Agricultural products 6.4.3Lime sludge classification 6.4.4Clay classification 6.4.5Waste activated sludge thickening 6.4.6Digested sludge thickening 6.4.7Lactose washing 6.4.8Coal tailings dewatering 6.4.9Dry solids (DS) dewatering Chapter 7Calculations and Scaling 7.1Basic Calculations 7.2Three-Phase Calculations 7.3Classification Calculations 7.4Washing 7.5The Probability Scale 7.6 7.7 7.8Main Motor Sizing 7.9DS Scaling Scale-Up of Centrate Clarity Limiting Applications Simple Dewatering and Torque Scale-Up Chapter 8Instrumentation and Control 8.1Decanter Plant Modules 8.2Instrumentation 8.2.1Flow meters 8.2.2Solids concentration meters 8.2.3Level probes 8.2.4Speed probes 8.2.5Temperature probes 8.2.6Torque measurement 8.2.7Timers 8.2.8Counters 8.2.9Electrical meters 233 235 236 237 241 245 248 2 52 255 255 258 259 261 263 265 267 269 269 284 288 291 294 298 300 302 306 308 317 319 319 319 320 321 32 1 321 321 322 322 XI 8.2.10Bearing monitors 8.3.1On/off devices 8.3.2Variable output devices 8.3Controlled Equipment 8.4Controllers 8.5Integrated Controller 8.6CIP 8.7References Chapter 9TheDecanter Market 9.1Market Characteristics 9.2Market Trends 9.3Market Size Estimates 9.3.1Overall decanter market size 9.3.2Regional market estimates 9.3.3Application market estimates 9.3.4Suppliers' market shares Chapter10Suppliers'Data Chapter11Glossary ofTerms 322 323 323 324 325 328 3 30 331 3 34 335 336 336 337 337 338 3 3 9363 Appendix 3 7 9Index 413 ThisPageIntentionallyLeftBlankPreface andAcknowledgements Byvirtueofitstitle,whichinvolvestheword"handbook", thisbookis intended,aboveallelse,tobeuseful.Itsaimsincludetheexpl anat i onof the nat ureandmethodsof operationof thedecant ercentrifuge,andadescription of the kind of performancet hat might be expectedfrom adecanter.Thedecant ercentrifugeisadeviceforcont i nuousl yseparat i ngparticulate solidsfromasuspendingliquidorliquidsbysedi ment at i onanddecanting. As such,itispartofthegeneralrangeofsedimenting,filteringandother mechanicalequipmentusedfor separationprocesses.A distinguishedrangeof booksexistst hat describesthiscompletespectrumofequipment, andthe processesbywhi chtheyoperate.Apreviousbookcoversthewholerangeof centrifuges,bothsedimenting(likethedecanter)andfiltering,butthisisthe first booktodealsolely withthesolid-bowl,scroll-dischargecentrifuge,which is the decanter. Thebookisaimedatallthoseforwhomthedecant ermaybepartof theirstudies,oftheirresearch, oroftheirworkinglife.Itisintendedtobe of valueinundergraduat ecoursesonfiltrationandseparation, butit willalso offerthepractisingengineerinend-usercompaniesmuchthatisofdirect valueto thedaily jobof designing,specifyingoroperatingthissophisticatedly engineered,butveryuseful,pieceofprocessingequipment. Thishandbook willfinduseinresearchestablishmentsandequi pment manufact urers'engineeringdepartments, asitgivesguidanceonbasicdesignandoperating features,someinregularuseandsomeonlyrecentlyintroducedtothe market. Thisessentiallypracticaltextneverthelesscoverstheunderlyingtheory ofcentrifugalsedimentationseparationsinsomedetail,whichfurther extendsitsusefulnesstotheresearchordesignengi neerlookingfornew ideas. Thear r angement ofthehandbookfollowsalogicalpattern:ageneral introduction,followedbytechnicaldescriptionsofequi pment featuresand theindustrialusesof thedecanter. Thencomesthet heoryof thedecant er' s design,anddetaileddescriptionsof operationalandtestprocedures.Thebook finisheswithsomemarket i ngdata,anddescriptionsof theequi pment ranges of the mainmanufact urers.xivPrefaceand Acknowledgements Theaut hors(bothCharteredChemicalEngineers)haveaweal t hof experiencein thedecanterbusiness: 9AlanRecordsretiredfromaseniorequipmentapplicationand developmentrolewithAlfaLaval,afteralmostafulllifetime' sjob involvedwithdecanters,coveringresearch,design,commissioning, operationandservice,ina wide rangeof industrialapplications;and 9KenSutherland, foratimeTechnicalManagerforSharpies,haslater beenheavilyinvolvedwiththemarket i ngaspectsofseparation equipment, includingcentrifuges. Theput t i ngtogetherofabookofthisnat urerequiresthehelpandco- operationofmanyindividualsandorganisations.Thecontributions, help, advice,workandkindpermissionsofthosementionedbelowaremost gratefully acknowledged. LennyShapiroandJanCederqvistcontributedtothemechani calinformation, while Bert Guilleassistedwiththeelectricalcontent.The process datawereobtainedasaresultof painstakingworkinthefield,ofteninfarless t hanasalubriousenvi ronment , bynumerousfieldengineers,ourformer colleagues,andinparticularJohnJoyce,BetinaPedersen,andKeithSmith. Apologiesare extendedtoall thosenot mentioned. DenisLockecontributedtotheworkonmanyoftheillustrations, professionallyexecutedby Mike Nicklinson. GrahamDawson,withthehelpof someof hisformercolleagues,advisedon thesectiononflocculanttechnology.KeithKernahanadvisedonthedetails of theViscothermequipment. TheTritonElectronicCompanyco-operatedinprovidingphot ographsand detailsof theirCST equipment. Thedecant ercentrifugemarketisahighlycompetitiveone,andthus manufact urersare,underst andabl y, reticentinprovidingspecificdataand informationontheirproducts.Withoutsuchdataandinformation,however, thisbookwouldbereducedinvalue.Theaut horsarethereforeespecially gratefulfor thedatasuppliedby thecompaniesAlfa Laval.Baker Process(Bird MachineandBirdHumboldt),Broadbent.Centriquip,Centrisys,Flottweg. Gennaretti, Guinard,Hiller,Hutchison-Hayes,Noxon,Pennwal t India, Pieralisi,Siebtechnik,andWestfalia/Niro.Permissiontoreproducesketches anddrawi ngshasbeenobtainedfromAlfaLaval,BirdMachine,Bird Humboldt,Broadbent,Centriquip,Centrisys,Cyclo,Flottweg,Noxon, Siebtechnik,Tomal,ViscothermandWestfalia Separator. Finally,gratitudeisexpressedtoBentMadsenandhiscolleaguesfor checkingtheearlymanuscripts. ThebookowesitsorigintoNickCorner- Walker,t henDirectorof EngineeringwithAlfaLaval,towhomtheaut hors areindebtedfortheinspiration,forhispersonalsupport,andforputtingthe resourcesofamajormanufact urerofdecantersbehindtheventure. The Preface and Acknowledgementsxv aut horsareveryhappytoacknowledget hat debthere,butalsoto acknowledgetheinputfromtheothercompanieswhoseideasand illustrationshave been used at theappropriate partsof the text. Tothese,andallof theotherworkersinvolvedwiththedecanterfor the60 years of its effective operating history, theaut horsexpress their thanks.AlanRecords Ken Sutherland ThisPageIntentionallyLeftBlankCHAPTER1 Introduction The decantercentrifugehasbecomeamajor processing tool ina wide rangeof liquid/solidseparationapplications.Thishandbookaimstobeat horough introductiontothedesign,performanceandapplicationofthedecanter.It aimsalsotobeausefulguideforthecentrifugeengineer,bothinequipment manufact uri ngcompaniesandintheend-usercompanies,andtheir associated contractorsandconsultancies. Thehandbook' sfirstchapterintroducesthereadertothedecanter,toits history andto themanufact uri ngsector withinwhichit is made.The contents ofthischapterareintentionallybrief,withmajorexpansionofthetopics covered in later chaptersof thebook. 1.1TheDecanterCentrifuge Thesolid-bowlscroll-dischargecentrifugewnowalmostuniversallyknown asthedecantercentrifugehas,indeed,becometheworkhorseofawide rangeof liquid/solidseparationactivities.Itsapplicationtothedewateringof wastesludgeshasmadeitamostvaluabletoolincombatingenvi ronment alpollution.Thishasmadethedecanterawell-knownandwidelyappreciated piece of equipment. 1.1.1Thebasic decanter Althougha complicatedpiece of machinery, thedecantercentrifugeembodies asimpleprinciple,thatofthescrewconveyor.Inbasicterms,thedecanter comprisesasolidcylindricalbowl,rotatingathighspeed.Insidethebowlisa scroll(screwconveyor)rotatingataslightlydifferentspeed.Thedifferential speedbetweenbowlandscrollprovidestheconveyingmotiontocollectand remove thesolids, whichaccumulateat the bowl wall. Aslurryof liquidandsuspendedsolidsisfedalongthecentreline.tosome fixed positionwithinthebowl,andis acceleratedout wardsto jointhepondof liquidheldonthebowlwallbythecentrifugalforce.Thissameforcethen causesthesuspendedsolidstosettle,andaccumul at eatthebowlwall.The clarifiedliquidthenflowsalongthebowl,toleaveatoneendof it,oversome kind of weir design,whichsets thelevel of theliquidsurface in thebowl. Theotherendofthebowlisslopedinwards,towardsthecentre,thus providingabeach,upwhichthesolidsareconveyed,tobedischargedfrom thebowl,atthetopof thebeach.Whilstthesolidsareconveyedupthebeach, some,hopefully most,of the entrainedliquiddrainsback intothepond,to join the liquid flow towards the far end. Thescrollusuallyiscarriedonahollowaxialhub,t hr oughwhichthe slurryfeedtubepassestothefeedzone.Thediameter,thenumber, andthe pitchof theconveyor flights arechosento mat chtheneedsof theslurrybeing treatedasarethedepthof thepond,thelengthof thebowl,theconveyor differential speed,and the angle of slope of thebeach. Mostdecantersoperatewiththeiraxishorizontal,inwhichcasethey usuallyaremount edinsubstantialbearingsateachendof thebowl.Vertical Introduction3 LiquidsFeedSolids C~,m8C~veyorBowlFl i #tDtsctmrl~.tZoneDischtr~ Feed Tube Figure1.1.The main operating parts of a decanter centrifuge. operationispossible,inwhichcasethebowliscarriedonlyononesetof bearings,atthetop.Ifthedecanterisshort,thencantileveredhorizontal operationis also possible, with bearingsat one endonly. The rotating bowl is enclosed ina casing,whichis divided to ensuret hat the dischargedliquid(the"centrate")andsolids cannot remixafter separation. The basic decanteris completedwitha drive motor,usually electrical,anda gearbox,whichcontrols the differentialspeed of the conveyor. Aspectsofthephysicalformsofthedecanterinitsdifferentversionsare described in Chapter2. 1.1.2 Separationprinciple Thedecanteroperatesmainlybysedimentation,aprocesscausingthe separationof suspendedsolids by virtueof theirhigherdensitythantheliquid inwhichtheyaresuspended.Ifthedensitydifferenceishigh,thengravity may provide sufficient driving force for theseparationto occurinareasonable timeasisthecasewithlarge-tankclarifiersandclariflocculators,orwith lamellaandinclined-plateseparators.If thedifferenceindensityissmall,or theparticlesizeisverysmall,thengravityseparationwouldtaketoolong, andtheseparationforcemustbeaugment edbytheimpositionof centrifugal forces many times thatof gravity alone. The centrifugal force may be imposed by virtueof the flow of the slurry,as in ahydrocyclone,orbymeansofmechanicallydrivenrotation,asinthe sedimenting centrifuge. 4TheDecanterCentrifuge Thereare several types of solid bowl sedimenting centrifuge,including: 9the tubularbowl centrifuge,mainly usedfor liquid/liquidseparation,for whichuseanysuspendedsolids would requirecessationof operationfor theirremoval(thetubularbowlcentrifugeisalsousedforverydifficult solid/liquidseparations,wherethereisalowconcent rat i onofsolids, which cannotbe flocculated): 9theimperforatebasketcentrifuge,whichisoperatedbatch-wiseforthe removal of collected solids; 9thedisc-stackcentrifuge,originallydevelopedforliquid/liquidsep- aration(creamfrommilk),butwhichhasbeenimprovedtoachieve continualsolids removal(although,in most cases,not fully continuous),by a variety of devices at the outer periphery of the bowl; and 9the decanter. Theprimebeneficialcharacteristicofthedecanterinthisspectrumof sedimentationequipmentisitsabilitytoremoveseparatedsolidsfromthe separationzoneonafullycontinuousbasis.Itcanoperate,unattended, for weeks,if not mont hs, at a time. By comparison,therefore,with: 9gravitysedimentationthedecantercanachieveseparationst hatwouldbeimpossiblylengthy(orjustimpossible)inaclarifierorlamella separator,andit produces drier solids; 9hydr oc yc l one s - thedecanterhasamuchhigherliquidcapacity,can handlemuchhigherslurryconcentrations, andproducesmuchdrier solids: 9tubularbowlcentrifugesthedecanteroffershighercapacities,the ability to handleconcentratedslurries,andcont i nuous operation: 9imperforatebasketcentrifugesthedecanteroperatescontinuously,canhandlemuchhighersolidsconcentrations, andproducesmuch drier solids:and 9disc-stackc e nt r i f uge s - thedecanteristrulycont i nuousinoperation, canhandlemuchhighersolidconcentrationsinthefeedslurry (althoughitcannotusuallymatchthehighcentrifugalforcesofthe disc-stackdesigns,andsodoesnothavethesameclarification performance),andproduces drier solids. Inadditiontotheseothertypesofsedimentationcentrifuge,thedecanter competeseffectivelywithseveraltypesof solidsrecoveryf i l t e r - suchasthe plate-and-flamefilterpress,andthevarioustypesofbandpress,wi t houtrequiring the use of filter aids. Thetheoryof theseparationanddewateringbehaviourof thedecanteris describedin Chapter4. (I t mustbercmcmbercdthat thereare manyother typesofindustrial c:enIril'iigr, hul: t.tiese achieve seprati oribyriieansoffiltrationrather than sedimenl:;ii.ion --:Ill.hoiighthesrreen-howldwarilerr:omhiriest.he two sepa r ii t.ion rn ~ha n isms .) 1.1.3 Decanter applications Thc dccantcr cciitrifugc can beused for most types of liquidjstrlid separ;ition, and its ability to handle a wide rarigeof feed slurry L.r,rir:entrol.iorrsodds to its general versatility. 11 C ~ I I be 11sed Tort . 1 ~ clussificntian of' solids in Liquidsuspc.nslon. where a single required tielween IWC) sixes ofsolid particle (or, less often. bctweeii solids ofdiffering density). I1i s i i verygcind device for this purpose, and its early history inclnded developme,nt for thc kaoliii (china clay) iiidustry. The can be used for the clnrjficntiori of a liquid. it can be operated so aslo giveahi ghdegree ofclarification,although itisnotusuallyusedto clocify o slurry that contains only a small d i d s in suspensiun, I t (::inalso be iwd in thc recovrr~j d a valuable scilid irom i tssuspcnsioii in a liquid, :ind li!Llowingsuch rcrovery it is capable of\.wr.shing t.he recovcrcd solid free ofthe original mother liquor, andof drliqimrin[j( d wn t e r i i i g ) the wrls1it.d solids to a hi gh dcgree of dryri Whcrc thc slurry isawasteneeding treatment priorto saledispos;iI, thc decanter again can dewat.ersuch slurries to a high luvcl of ilryriess. Finally the decariler c:inbe opcratcd so as to act as ili hi r kmv - , producing a clcar liquidand i-lmore concentratcdslurryeitherin a manufacturing proccss, or i t 1 ws s t ctreatment. This wide mnge of'potcntialuses. coupled with its continuous operalion,its ability tn acceptawide range offeed concentrations. and its arailahility in a wide raiigcolfredcapacities, t.9.3m3/h[ A I 20.0025.00 Figure 6. J J.Graph -Cake Dryness v Polymer Dose -DS. 276SomeTest Data scale-upof data,thedatafordifferentsizesof decanterneedtobecoincident onthegraph.ThisisnotsoforthetwosizesplottedinFigure6.34.Thisis becausethe slope of thelineandtheco-ordinateinterceptare dependentupon the maxi mum dryness achievable,x~,whichis a functionof the depthof pond inthebowl.Inthe152mmdiameterbowlthepondisveryshallow.A better correlationis shownin Figure6.35for threedifferent bowl designs,involving different beachangles and baffle types,and two different sizes. ThedataseriesinthegraphinFigure6.35areseparated,foreasier definition,inFigures6. 36- 6. 39. Itcanbeseenfromtheset hat allbutthe smaller machinelines are coincident.The slight difference of thesmaller one is probably due to its slightly reducedpond depth. 30.0 28.0 26.0 24.0 22.0 Ig -20.0 018.0 0 I Io16.0 14.0 12.0 10.0 0.00 it 152mmBowlOia. ! n~ ~ - ~ ] !: 425mm Bowl Dia. j - -I- "tt ! )J --. -. . . . . .0.05O. 10O. 150.200.25 Q4/g-Volh "I Figure 6. 34. Graph -Cake Dryness v Feed Rate/g-Volume - DS. TestWork and Dat a277 35.0"' ~~176. . . . . . I"-.I.',J~. - " 25.0. . . . . . . . . . . .~20. 0....t me575mmBowlDJa TypeAI! 0 u575mmBowlDJa TypeB)J i ii 15.0I,9575mmBowlDia Type C i ii l I lo.oI. . . . . Q~g-Volh "4 Figure6. 35.Graph-CakeDr yne s s vFeedRa t e / g- Vol ume -DS. 35.0 30.0 25.0 m= e~2o. 0 o 15.0 10.0 0.00 -t I ~-~575mmB~~Dia TypeA1 Ch/g-Volh "4 Figure6. 36. Graph-CakeDr yne s s vFeedRa t e / g- Vol ume -I)S. 278Some Test Data 35.0. . . .30.0 25.0 m = a20.0 m 15.0 10.0 0.00 ) I !!m 575mmBowlOia TypeB J I 0.05O. 10O. 150.200.25 Q4g-Volh "t Figt~re 6.37.Graph- Cake Dryness v Feed Rat e/ g- Vol ume-DS. 35.0 30.0 ~e25.0 m = E, 020.0 m u 15.0 10.0 0.00 1I ,I 1J ) 1 ) I i I i{" 5zsm~~,~,D~. l,,] Qg/g-Volh 'n Figure 6.38.Graph-CakeDryness v Feed Rat e/ g-Vol ume-DS.Test Work and Data279 35.0 30.0 25.0 m = o20.0 0 m 15.0 10.0 0.00 i!I" 0.05 " ' - " - tJJt tII ~425mmBowlDiaTypeA1 J JJJ Qt / g-Vol h "t Figure 6.39.(;raph- CakeDrynessvFeedRat e/ g- Vol ume -I)S. ThisPageIntentionallyLeftBlankCHAPTER7 CalculationsandScaling Eachtestruninanytestpr ogr ammerequiresanumber ofbasic calculationsdependinguponthesourceofthedata.Theformulaerequi red aregiveninChapter4.Whi chcal cul at i onsarenecessarydependuponwhati nst rument at i onis installedonthetestfacility.For instance, if cent rat erat eis measuredthenthefeedrateneedstobecalculated,andifthefeedrat eis measuredthencent rat eratehastobecalculated. Boththeseratesareneeded to calculatesolids recovery, anot her of thecalculationsneeded. Someoftheearlymodelsofdecant erwereequippedonlytoindicate gearboxpinionspeed,rat hert hanspecificallyconveyordifferential.This meant thatdifferentialspeedhadtobecalculated, andcalculatedusingthe nomi nal bowl speedwi t hout abowlspeedmeasurement , or wi t hjustaone-off measurement . Withthiscal cul at i onithasto beborneinmi ndt hat bowl speed canvary by1 O0-2()0rpmwi t hchangeofload, andbelt slip,if any.Thiscould leadtoamajorpercentageerrorondifferentialcalculationif thedifferentialis low. Earlydecantersusingeddycur r ent brakeswerenotabletoindicate conveyor torquecont i nuousl y. A readi ngof brakespeed,andanot her ofbrake current, hadtoberecordedandusedwhenreferringtoabrakecal i brat i on chart, toobtainthebraketorque. Conveyortorquecouldbeobtainedby multiplyingthisfigurebythegearboxratio.Atypicaleddycurrent brake calibrationchart is shownin Figure7.1. Whenflocculantsareusedthepolymer dosagelevel hasto be calculated.On specialapplicationsuni quecal cul at i onswillneedtobeperformed.Inthree- phaseworkthemassbal anceneededtoworkoutcent rat eratesismore complex.Oilrecoveriesandlosseswillneedtobeassessed.Withthreemass balances,solids,oilandwat er, somei mbal anceistobeexpecteddueto experimentalerror,andcareisneededtoensuret hat theseerrorsarenot allowed to affect thereliability of the result. In classificationwork massbalancesmaybeneededonahostof size ranges.Whenplottingcumul at i vedat asetsagai nst size,thesizeusedneedstobeat 282Calculations and Scaling E Z o o"p-o J< (I; =_ m 140.0[ 120. 0.iI- l 1.75 Amps -___ i 1.50 Amps ,1.25 Amps I /,1~! I/. . . . . . i. . . .il,F,!~0, 75Amps,Ijo0 o. o,.!I.. 100.0 80.0, 60.0 40,0 20.0 050010001500200025003000 Brake SpeedRPM 3500 Figure 7.1.A typical Eddy CurrentBrake calibration. Calculations and Scaling283 oneendofthesizerange, theenddependi nguponwhet her thedat aare cumul at i ve undersizeor oversize. Testdatamayindicatet hat thesizeof testmachi neisadequat efor thedut y envisaged.Alternatively, themachi netestedwouldbetoobig,whi chisnotusual. Moreusual l ythetestdat ahavetobescaledtoalargersizeof decant er.Whenthedataneedtobescaledtoanot her decant ersize,ot hercal cul at i ons mayneedto be performedfor eachrun, suchasconveyor t orque/ vol ume, feed rat e/ Si gma, wetsolidsr at e/ conveyor differentialspeed,aswellastheSigma valueitselfifthedat ainvolvechangesofbowlspeed.Aswillbeseen,these i nt ermedi at ecalculationshelpwi t h the scale-up. Whencont empl at i ngthescalingofdata,onehastoconsiderwhat isthe limitationtotheperformanceonthetestmachine, becausedifferent limitationsrequiredifferentscalingt echni ques. Moreoverifthescale-up factorislarge,t henthescalingmayi nt roduceanot herlimitation,if thescale factorfor thesecondlimitation is smaller t hanthe first. Themai ndecant erperformancelimitationsareas follows: 9cent rat eclarity; 9cakedr yness- non- DS:9cake dr ynes s -DS: 9scrollingvolumetriccapacity; 9scrollingtorque; 9mainmot or power. 7.1BasicCalculations Forasetofexampl ecalculations, whi chhavetobeconduct edoneachtest run, the dat afrom r un26of Table A. 6in theAppendi xwill be used.Data" Feed rate19.6m3/ hBowl speed3150rpm Feed solids2.6%w/ wPinionspeed650rpm Flocculantrat e0. 98m3/ hGearbox ratio125 Pol ymer0. 13%w/ wBowl di amet er425mm concent rat i on Cent rat esolids1350ppmClarifying l engt h800mm Cake solids10.7%w/ wPonddi amet er257mm Itwillbeassumedt hat thedensitiesof feed,pol ymersolution, andcent rat e areunity. Thus, from equat i on(4. 13), c e nt r a t e rat e" ( 10. 7- 2. 68) ( 10. 7- 0. 13)Qt=19.6+0. 98 (6.9-0. 135)(10. 7-0. 135) 19. 6x8. 020. 98x10. 57 =+ 10. 56510. 565 =14. 88+O.98 Ol~"15.9m 3/h From equat i on(4. 14), r ecover y:(15.9x0.135) =100(1-0. 0409)R~95. 9% From equat i on(4. 15)pol yme r dosage: 0. 98x0. 13 PD-- 19.6x2. 68 PD- 2. 43kg/t.db x1000 Cnlrirlntians nvd ScnIing2 8 5 where dh indicates dry basisand wb would indicate wetbasis. Thus, inthis case the polymer dose is quoted in kilograins of active polymer (dry) per ton of dry solids in the feed. From eqrialion (4.9),conveyor differential: Ilie data in Table A.6 are fur a thickening application. For thickeriirig. iiis l hrn equation (4.59),psi; sornelirnes useful LO calculate psi, the thickening factor. 2 0 19.6 x2. 68 x10.7 *= Notethe mixture ofunits (rpm divided hy tn3$l.For thc purist. the answermirybe indtiplicd by2 x xMJ, tc>givt: uriit,sofm- ,Psi is uscd fur cornparison purposes. and therefore. 50lorig a s[he same units arc uscd throughout, t.he choice of units is immatcriel. Thc Sigmavalue fur llie decaqt esnormallywouldbeobtainedhornthe decantcrmanufacturer.hutTordemonstrationpurposwAv; Il ucwqllbe estlmatcd hcrc. Frntii cquaiion ( 4, #) . centrifuge g-level: > 21. 25 = 33 0. .x - 481 1:ro111equation ( 4. 32) . Sigma: 114576 = 27900 x-~400 x0. 503 = z7v00 x569.45 c = 1.59 x1ocm 286Basic Calculations Thus,0/ ~"19.6 - 1. 59x107 Q/ E=12.3mm/h xl O00xl O00x10 Forthepurposesofdemonst rat i onapiniont orqueof10Nmwillbe assumed. In t hat thiswasa t hi ckeni ngapplication,where t herewould be little interestint orquemeasur ement , thepiniontorque, inallprobability,woul d havebeenmuchlower.However,thisfigureisatthelower endof what woul d be experiencedin a dewat eri ngapplication. From equat i on(4. 10), conveyortorque: T=125x10=1250Nm T =1.25kNm Cl ari f yi ngvol ume:V- - 71-5272 400(42- 25) x 3 =71990cm V- 71. 991 800 10 Thus,T~ V: 1.251000 T/ V=x 71. 991000 T/ V- 1.74N/cm 2 x1 ()0 Q/g-Vol: 19.6x1000 Q/g-Vol=2357x71. 99 Q/ g-Vol -11. 55x10 -2h -1 Thus,thenecessaryfactorshavebeencalculatedshouldascale-upbe requiredfromthisonesetof data.Naturally, inarealsi t uat i oncalculations wouldnotbeconduct edbothforthickeningandDS.Theyaredonebywayof examplehere.Witha10.7%cake,itisunlikelytobeaDSapplication. However, wereittohavebeenaDSapplication,andthedrynesswas consideredadequat eand, for instance, doublethecapacitywasbeingsought,t henadecant erwi t htwice theg-volume would be required.Theg-level woul d probablybechosensimilar,thusthebowlvol umewoul dneedtobedouble, andthereforetomai nt ai nthesamedrynessagearboxoftwicethet orque Calculations and Scaling2 8 7 wouldberequired.Morenormal l ythesecalculationsareconductedforeach run, andperformancelevelsplottedagainstthem,sucht hat theopt i mum performancecanbe chosenandscale-upis madefrom there. 7.2Three-PhaseCalculations Similarcal cul at i onsneedtobeconduct edonanymoni t or edr unint hree- phasework. A sampl esetof cal cul at i onsis givenhere. Data: Feedrat e4m 3/hBowldi amet er 42 5 mm Feedsolids39%w/ wOil di schargedi amet er 2 70mm Feedoil24%w/ wWat er di schargedi amet er 274mm Feedwat er 3 7%w/ wCake di schargedi amet er 264mm Rinsewat er rat e1 mS/ hBowlspeed3150r pm Cakesolids50%w/ wClarifyingl engt h750mm Cakeoil6 % w/ w Cakewat er 44%w/ w Oil rat e0. 8m3/ hEffluent solids4. 4%w/ w Oil wat er 1%w/ wEffluentoil0. 4%w/ w Oil solids1% w/ wEffluent wat er 95. 2%w/ w Oil98%w/ wOil densi t y0. 85 Firstatotalmassbal ance, followedbyasolidsmassbal ance, isconduct ed.Oneofthetwounknowns , solidsrat eoreffluentrat e, isel i mi nat edby subst i t ut i ngfromoneequat i onintotheot her, andt het wounknowns are cal cul at ed. Thus, fromanequat i onsi mi l artoequat i on(4. 11),thetotalmass bal anceis: FeedRat e+Wat er Rat e=CakeRat e+Ef f l uent Rat e+Oi l Rat e 4+1=CakeRat e+Ef f l uent Rat e+0.8x0. 85 whence 4. 32=CakeRate+Ef f l uent Rat e Thesolidsmassbal anceis: 40.39+0-0.5xCakeRate+0. 044EffluentRate+0.80.85x0.01 1.56=0.5xCakeRate+0. 044EffluentRate+0. 0068 Calculationsand Scaling289 whence 1. 5532-0.5xCak e Rat e +0. 044xEf f l ue nt Rat eBy substitution" Caker at e -3. 0027t/ h Ef f l uent rat e-1. 3173t/ h ,-,3.ot / h ,,~1.3t / h Thisassumest hat thedensitiesof effluent,feed,andcakeareallunity.This is notquitetrue,butwi t hi ntheexperi ment al errorassociatedwi t hthistypeof work,thisis acceptable. Re c ove r yof s ol i ds3. 0027x0.5 =x100~96. 2% 4x0.39 Re c ove r yofoi l0. 80x0.98 4x0. 24 x100,~81. 7% Oill os s i nc ake3. 0027x().06 4x0. 24 x100~, 18. 8% Oill os s i nwat e r1. 3173x0. 004 4x().24 x1 O0-~, 0.~/o Notet hat theoilrecoveryandoillossesdonotadduppreciselyto10()%.This is duetoexperi ment al error, andsometimescanbe muchlarger. Int hree-phasework,itcanbeusefultodet ermi netheapproxi mat eposition of thee-line(equilibriumline).Usingequat i on(4.61): 0" 85x~2[~ 2 g' 7- ( 2~0) 2] - -l"Oxa~2[~2f f ' 7- ( _~) 2]whichsimplifiesto" 0. 85( r ~- 1352 )=1. O( r ~- 1372 ) O.15r~=18769- 15491. 25 V/ ~3277.75 r,.- - O. 15,~,148mm=296mmDi a Thus,e - l i ne di amet er" =296mm 290Three-PhaseCalculations Thisgivesadept hofoilof( 296- 270) / 2-13mmoverawat er dept hof ( 425- 296) / 2-64. 5mm.Assumi ngitisj ust aseasytosepar at et heoilfromt hewat er asitisto separ at et hewat er fromt heoil,whi chisnot necessari l yso,itiswor t h cal cul at i ngt heQ/ Efor eachphase.For t helightphase, from equat i on(4. 32), Sigma: E7rx750x3302( 2962-2702) - - - X 981x10202l n( 296/ 270)14716 =26155. 9x 400x0. 0919 :1. 05107cm 2 Thus, light phase Q/E: 0. 8 0/ z - 1. 05x10- Q/ E-0. 76mm/h xl O00xl O00x10 For t heheavyphase(flow rat e-1. 2018) , Si gma" E=Tr x750x3302x( 4252- 2962)981x102021n( 425/ 296)93 O09 =26155. 9x 400x0. 3617 =26155. 9x642. 86 -) E=1. 68x107cm- Thus, heavy phase O/ E:1. 2018 Q/ E- 1. 68x1071000 1000 10 O/ E-0. 72mm/ h Wi t hQ/ Eval uesbei ngsosi mi l arfort het wophases, itwoul dseemt hat t he differentialpondset t i ngisopt i mum, bar r i nganycrest i ngeffectsorback- pr essur eeffectsfromanydi schar gedevice.Ifitwerenecessar y, say,to i mpr ovet hequal i t yoft heoilphaseatt heexpenseof wat er effluentqual i t y,t hent hewat er di schar gedi amet er woul dneedtobei ncreasedveryslightly, to i ncreaset hedept hof t heoil level int hepond.7.3 Classification Calculations _-- Asa11exampleofthccalculationsneededforadecantertestrun On a cIassitic;jt.ion duty. the data uscd for Figures 6.1 2 - 6.1.7 will he er~ipluyed.'The dat.a are given i n. arid adjacent to, Tablc 7. 1.The parl.icle size analyses frdkilTablc7. 1 arc plott,ed i n Figerr: h.1 2.Prom thisgraphfrequencydist.rihutions arecalculated.'I'he dataabovcarccc- tabulated, ciilculi~tingthe percentage it1each sizeinterval, anddividing that percentagebyt hesizei nkrval . 'I'Iiesefigures.thuscalculatcd,givet he rclativcfrequcncyforeachsizeinterval.Thefreqrirric:ies c:eril.raf.c! distributioe arc thcn multiplied by(1-solidsrecovery ;IS iiItect.ion),to make the frequcncics in thc ccntratc distribution correspond t o those in the feed. by virtue o fthe particles lost in the cake. 'J'hesefigures are tabulated in Table 7.2, Table 7.1.PAr l i i : l c s i x;rn;rlyscs ParticlrCum. 9:,r 1 1 1 1 r . '%I sizcundcrsiecwi dcr si ac (pn)feedcentrale 1 I).O07. 40 9 ,!I 7 9. 096. 9Y S. Y h 8.09f1.199.9 1 7 . 0 95. 1)99.88 6.093. h 9 9 .7 h 5.UY1 . 3 9Y . 53 4.088. 099.00 3 . 0 X2.h97.60 1.o7 2 .o9 3 . 0 01.5h 3 . ci H7.00 I,I)50.07 3 . 0 00. 842.6~ 1 3 . 0 00. 0'3'1.649.00 0.422.030.00 0 . 3 15. 21X.hO 0.28.08. 20 0.12.51. 20 _..-..,., ,, .,, , .,-, -,-- .. .~. .. .. . ...Fw drate Howl dlamr:tr:r iliirifylng lerlgth Cakc discharge diamcter Pond diameter Krrwl spct:d Sol i dsrrtwvt.ry I ) i k r c n Ii alSGof feed Feed solids Cakc solids Cc r ~t r . a~c u o l i d s292Classification Calculations Table7.2. Frequencydistributions MeanSizeFeedFeedCentrateCentrate s i zeinterval% in(%/~tm)% in(%/~tm) (~tm)(pm)intervalinterval Centrate (%/~tmx(1-R)) 9.501. 8.501. 7.501. 6.501. 5.501. 4.501. 3.501. 2.501.010.610.64.604.602.70 1.750.58.416.86.0012.007.03 1.25O. 513.627.214.0028.0016.41 0.900.27.437.010.0050.0029.30 0.700. 0.500.211.658.019.0095.0055.67 0.350.16.868.011.40114.0066.80 Byt aki ngt hedifferencebet weent het wof r equency(%/lam)col umns, t he size f r equencydi st r i but i onfor t hecakeis obt ai ned. Thi sis s howninTabl e7. 3.Thet hr eef r equencydi st r i but i onsarepl ot t edont hegr aphinFi gure6. 13.Thecut poi nt isobt ai nedast hepart i cl esizeatwhi cht hecakeandcent r at e di st r i but i onsi nt ersect . As,att hi spoi nt , t hef r equenci esfort het wo di st r i but i onsareequal andt ot al t hat for t hefeed,it followst hat t hi sf r equency ishal f t hat int hefeed,whi chis t hedefi ni t i onof cut poi nt .Thecut poi nt foreachflowrat eissi mi l arl yobt ai ned, afterwhi cht hegr aph of flowr at eagai nst cut poi nt maybeplotted, asinFi gure6. 16.Fr omt hef r equencydi st r i but i ons, t hes epar at i onal efficiencyforeachand anysizema ybeobt ai nedbyt aki ngt herat i ooft hefrequenci es, caketofeed, andmul t i pl yi ngby100. Thiscanbedoneforeachfeedr at etested. The efficiencyplotfor10m 3/his s howninFi gure6. 14.Fi gure6. 15givest her ecover yforeachfeedrat e, whi chisneededto cal cul at et hecent r at ef r equencycur ve, andt hus t hecakef r equencycurve.~ un u~ ~ ~ C 0 uq u~ 0 ~ ~ C C ~ 0 0 ,_,. N ~. i N ,..,. ,-t 0" ,_,~ 0 ,..,. 7.4 Washing ThedatadepictedinFigure6. 23willbet akenasthedat asourcefora demonst rat i oncal cul at i on. Theassociateddat aareas follows: Feed rate Feed i mpuri t ylevel Feed suspendedsolids Cake moi st urecont entWashrate 5500kg/ h 8. 125%w/ w 37. 5%w/ w 25%w/ w 7.5% Figure4.11(theri nsi ngwithdiffusiondi agram)isreproducedasFigure 7.2,withspecific figures,thederi vat i onsfor whi charegivenbelow. Was hrat e:0W--- 7.5 100 x5500- 412. 5kg/h Assumi ngfull recoveryof solids.Dr yCa ke Rat e" 37.5 QsXs =x5500=2062. 5kg/h.db 100 Q, =5500kg/h Q. =412. 5~ c,=y. =0 Figure 7.2. Rinsing with diffusion-Mass flows. Q.~2750k~ c2=? Calculations andScaling295 Wetcakerate: 100 ( 100- 25)x2062. 5-2750kg/ h. wb Impuri t yl evel offeed: I f - 8. 125%. wb 1 O0 =8. 125x 37.5 !1-21. 67%. db Basedont heliquorinthefeed,i mpuri t yconcent rat i on:I()0 cl-8. 125x=13. 0% ( 100- 37. 5) Thislasti mpur i t ylevel.13 %.willbet hesamefort hecakeafterdecant i ngthe excessliquor, wi t hout rinse.However, basedont hesolids,cakei mpuri t y l evel ( wi t hout rinse): ( 275( ) - 2()62. 5) I s - 21. 67x ( 5500-2062. S)I~--4.33%. db Comparethisfigurewitht heordi nat ei nt ercept ont hegr aphinFigure6.2 3.It willbeseenfromthisgr apht hat thisfigureagreesext r emel ywellwi t hthe pract i cal resul t .Cakemoi s t ur e di s chargerate" O.~ps(1-xs)-2750- 2062. 5-687. 5kg/ h Ri nserate: 7.5 0, , , - l ooX~500 O, , , - 412. 5kg/ h Thus, wi t hperfectrinsing, cakei mpuri t yl evel wi t hri nse: ( 687. 5- 412. 5)I s =x 4. 33=1. 73%. db 687. 5 296Washing Thisisthelowesti mpuri t ylevelpossiblewith7.5%rinse.Inpract i ce(Figure 6.23),forthesmallestdecant ertested,thelevelis2.3%.Thus, ri ns i ng efficiency" _-( 4. 33- 2. 3) x100 ( 4. 33- 1. 73)=78% Todet ermi netheeffectt hat anypossiblepar amet er changemayhave, itis necessarytodet ermi netheeffectivemasst ransfercoefficientforthese conditions. Impuri t yc onc e nt r at i oni nmoi s t ur e ofcake" 2062. 5 c2=x2. 3=6. 9% 687. 5 Thencetheconcent rat i oninthecent rat eliquorcanbeobt ai nedusi ng equat i on(4.77). Cent rat el i quori mpuri t yl evel : ( 13-6. 9) x687. 5-412. 5( c3-O) 687. 5 c3=6.1x 412. 5 =10. 17% Thel ogari t hmi cc onc e nt r at i ondi f f erence:Ac =( 6. 9- 0) - ( 13- 10. 17)l n[ ( 6. 9-0) / ( 13-I O. 17) ]Ac - 4. 57% From equat i on(4.88)" 687. 5( 13- 6.9) hDA,,-- IO0x4. 57 =9. 18kg/ h. %conc.diff Thequestiont hat nowcouldbeposediswhet her , byi ncreasi ngconveyor differentialonthesmallerdecant er, thecakei mpuri t ylevelcouldbelowered tomat cht hat achievedonthelargermodel,asshowninFigure6. 23;i.e.t he i mpuri t ylevelneedstobeloweredfrom2.3%to2.0%,stillwi t h7.5%rinse rate. Increasi ngthedifferentialfrom60rpmto,say,75rpm, increasest he superficialvelocity ofri nseover thecake by75/ 60=1.25. Thisincreaseinsiiprrfirialvelocitywillincrrijsrthemasstransfer coefficient by thc Same proportion (see quati oi l(4.85 ) ) . Thus. revised h d , := 1.25 x9.18= 1 I . 5~ ~ / h . % ~ O t l C . d ~ /Theimpuritylcvclsinthecakeciinnownnlybebackcalculatedby iteration. 'I'ablc 7.4 gives figures obtained in the it.eraf,i()ri,fromhid^a resdt can he iriterpolakd. llsing l'able7.4, it will bc seen thata diflerential of75 rpm (1r)ok al ong the row for which hoAc is nearestto 1 1 . 5 ) only rcduccs thc c ak e impurii.ylevelfrtim2 . 3to 2.1 7 .barely half thc reduction rcquired. To achieve the2.03: tignre, t.he ccinveyor di l kenl i alwouldneed lo be doubled. assuming that. it. were prar:t.icahle. 'lable Z4.iterative clilculations for cake impuritylevels CakeCakeMi ( kg/ h )Cer it ml t: moistureimpurity i%db)liquor iinpurity(%)impurity('%,1 5.21.71i3.h 11 3 . 0 05 , 5 1.8.55 1 . 5 0 12.50 h . 0 2. 0048.1411.h; 6.52.1; 44.071O.H 1 7.02. 3341. 1510.00 Z. i O17.H 70 . 1 7 n. o2.6714. 368. 33... __ .. - . . . .-, _ ".- -f . 3..-.. -...... . . . . . .0.011'.Y ' , ON'4. 79 3. 0915. 55 1. Y 5I 1 . J I4. 72H.74 5.4')(3.89 8.534. 03 .............. - 7.5TheProbabilityScale Sometimes it is necessarytoplot onalog-probabilitygraphbutt heprobability scaleisnotavailableassuchinaspreadsheet . Theprobabilityscalecanbe calculatedusi ngther i ght - handsideof equat i on(4. 19 t. Theprobabilityscale, therefore,willbeproport i onal toer f - l ( 2Cx- 1) , whereCxisapercent age figure for whi chthescale is required. Themat hemat i cal t erm, erf(x),isat abul at edintegral, whi chmaybe obtainedfromanygoodmat hemat i cal bookoft abul at edt r anscendent alfunctions. Forthespreadsheet alook-uptablewillhavetobecreat edofCxagai nster f - ~( 2C~- l ) , andasimpleformul ai nt roducedtoi nt erpol at elinearly betweenvalues. ThisishowthegraphsinFigures6. 12and6. 14were created.Thelook-uptableusedis shownin Table7.5. Theprobabilityscaleisfrequentlyusedindecant erworkanditisusefulto knowhowto createsuchascale whenaready-madeoneis notavailable. Calculationsand Scaling299 Tabl e7.5.Look-uptableforCx againsterf-l (2Cx-1) Percent Cx 0 0. 01 0. 05 0.1 0.2 0. 5 1 2 5 10 20 30 40 50 6O 7O 80 9O 95 98 99 99. 5 99. 8 99.9 99. 95 99.99 1 ()0 Er f - l ( 2Cx- 1) 0. 00 O.37 0. 68 0. 82 0. 97 1. 18 1. 36 1. 55 1. 84 2. 05 2. 33 2. 63 2. 82 3.00 3. 18 3. 37 3. 68 3.95 4. 16 4. 45 4. 65 4. 82 5 .()4 5. 19 5.32 5.63 6.()() 7.6Scale-Up Applications ofCentrateClarityLimiting Scalingupbetweentwodecantersizesisgenerallybestdonewhenthereis geometricalsimilaritybetweenthem. Thismeansthesamebeachangle,the sameconveyorpitchangle,andthesameconveyorandfeedzonedesigns. Whenthereare differencesthenthescale-upmaynotbe reliable. Acentrateclaritylimitingapplicationischaract eri sedbyafalloffin centratequality,whenfeedrateisincreased,i ndependent ofconveyor differential,onceponddepthhasbeenoptimised.Spentwashdewat eri ng,discussedin theprevious chapter,is one suchexample. Notetheprovisoconcerni ngdifferential.ReferringtoFigure6.3,the conveyordifferentialneededtobeatleast18rpmforthecent rat etobe unaffected,and raisedevenhigherat thehighercapacities. To scaleanyof thecapacitiestestedtoanot hersizeof decanter, orthesame decanterwitha differentbowl speed,theratio of Sigmavaluessimply wouldbe used. Thebestdrynessachievableonthetestmachi neistakenfromFigure6.4, usingthemi ni mumdifferentialnecessarytoachievethebestrecovery, atthe ratechosen,shownonthegraphinFigure6.3.Transl at i ngtheopt i mum differentialtothelargermachi neisusuallynotdonebycalculations, butby trialanderroradj ust ment ofdifferential,whencommi ssi oni ngthelarger decanter.However,if necessary,calculationsof cakescrollingrateandcake residencetimemaybemadetoensuret hat thenecessarydifferentialrange requiredonthelarger machi neis available. Theclayandlimeclassificationapplications,coveredinChapt er6,arealso examples of datawhi chwould be scaledby Sigma ratios. Thelimeclassificationisinterestingint hat itinvolvestwomaterialsof suspendedsolidswithdifferentdensities.Inthisapplicationitwasrequiredto produceacake withless t han60% magnesi umhydroxide, to prevent slagging duringcalcining.Foreconomicreasons, itwasnecessarytorecoveratleast 85%of thecalciumhydroxide.FromFigure6.11itwillbeseenthat, onthe testmachine, anycapacitybet ween12.5and20mS/hwoul dachieve theobjective.Thiscapacityrangewouldbescaledproport i onal l ytoSigma. Calculations and Scaling301 Thelowestcapacitywouldachieve90%recoveryof calciumhydroxide,and the highest85 %. 7.7 Simple Dewatering and Torque Scale-Up While themajority of applicationswill be scaled by Sigma,thereareoccasions whenhighconveyor torquesareexperiencedduringtest work.Thenit willbe necessarytoscalethetorquesexperienced,totheproposedlargermachine, notonly toensurethatthiswill notbelimitingbutalsoto estimatethepower requiredfor the drive motor.Some"drysolids"applications(coveredinSection7.9)willhavetheir capacitieslimitedbygearboxtorquerating.Whenscalingtheseapplications tolargerdecantersizes,itisi mport ant thatthetorqueratingavailableis greater t hanthe scaled-up torque. ThegraphinFigure6.6givestorquedataforthespentwash dewateringapplication.Thesedataneedtobeconvertedtoaformsuitable forscaling. Fromequation(4.70)itwillbeseenthattorqueisproportionaltofeed rateandinverselyproportionaltodifferential.Therefore,thedataof Figure 6.6shouldbere-plottedastorqueagainstfeedrat e/ conveyordifferential speed.ThisisdoneinFigure7.3.Hereitwillbeseenthatmostof thedata, forthelowesttwodifferentialstested,formastraightline1.1kNm/ (m3/ h/rpmdill). Asanexample,thesetorquedatawillbescaledtoadecanterof737mm bowl diameter.Data: Test decanterLarge decanter Bowl speedrpm3150? Bowl diametermm42 573 7 Pond diametermm261480 Cake dischargediametermm264483 Clarifying lengthmm12 O022 60 Conveyor pitchmm12 754 Wetted areaof bowlm 22.26.7 Beachaream 20.51.4 Conveyor differentialrpm13? Feed ratem3/h16? 1.80 1.609 1.40- E z1. 20 .ar Q J =1 "1. 00- I -0. 80 >" 0. 60 0.40 0. 209I I 0.00, 0.000.20 Calculations and Scaling303 Il j .. /t 9Diff.13.2RPM ==Diff.18. 2RPM i &Diff.23. 2RPM . . - - . - - - - - - __~] ~ J 0. 400. 600. 801. 001. 201. 40 FeedRat e/ Di f f er ent i al Q/ N Figure 7.3. Conveltor Torque vs. Feed Rate:D(fJerential Ratio. Firstlyt hebowl speedfort hel ar ger ma c hi ne needstobecal cul at ed,as s umi ngt hat itwillneedtohavet hes ameg-l evel ast hetestmachi ne. Fort he testmachi ne, bowl speed: S-330r ad/ s e cg-level(equation(4.8)): 3302425 g' 981x20 g,.=2357 Fort hel ar ger machi ne, bowl s pe e d:2357x981x20 S=737 =250r ads / s e cS.~2400rpm 304SimpleDewatering and Torque Scale-Up Thefeedratewillbescaledby Sigma,whi cht husneedstobecal cul at edfor eachmachi ne, usingequat i on(4.32).For the test machi ne, Sigma: E7r x1200x3302F,42s,[~,_yO__)2_, 261, ]i_2_ff)2 ----X 981x10ln(425~ k2611 41849. 5x281. 26 0. 4876 E=2.41x107cm 2 For thelarger model,Sigma: E37r x2260x2502[ ( ~) 2- (~~17621 - - X 981x10l n( ~)45234. 5x781. 9 0. 4288 E=8.25x107cm 2 Feed rate: 8. 25 Of-2.41 x16=54. 8m3/h Differentialisthenextparamet erval uetobecalculated. Intheabsenceof anyotherinformation, initiallyatleast,thedifferentialfor thelargerdecant er wouldbefixedtohavethesamecakethickness,inthebowl,asthetest machine. Cakethicknesswillbeproport i onal tofeedrate,andinversely proportionaltoconveyorpitchanddifferential.Thus, forthel argermachi ne,conveyordifferential: 54.8127425 N- xx13 16254737 N~13rpm which, conveni ent l y butcoincidentally,is the sameasthesmaller machi ne.Torqueisscaledupusi ngequation(4. 70). Conveni ent l yagain, t hereisno drybeach,andsothereisnoneedtousethefactork4.However, theheel torqueneedstobeestimatedusingequat i on(4. 71). Theheelt orqueforthe smaller decant eris theordi nat einterceptonthegraphinFigure7.3,whi chis approxi mat el y 0.3kNm.Thus, for thelarger decanter, heel torque: 6.72357737 To=2.223574250.3 To~1.6kNm Calculations and Scaling305 Tor queont het est machi ne, at t hespecified feed rat e, is1. 6 5 kNm.Ont hel ar ger machi ne, torqueaboveheel torque: ( T- To)=54. 8x425 16737 ( T- To)=7.5kNm 1.4 x0. 5x( 1. 65-0. 3) Totalconveyi ngtorque" T- 7.5+1. 6=9.1kNm 7.8MainMotorSizing Thedataforthespentwashdewat eri ng, usedinthelastsection,willalsobe usedto estimate themot orsize requi redfor thelarge decant erexampl e.Equations(4. 130)-(4. 134)areusedtocalculatethecomponent sofprocess power,exceptthepowerconsumedtoovercomewindageandfriction.This component wouldbeobtainedfromthemanufact urerofthedecanter, orby measurementinthefield. Themanuf act ur er suppliesthefiguresof4. 4and3 3.6kW,wi ndageand friction,forthetwosizesofdecant er, operat i ngatthespeedsspecified.The figureforthetestdecant erisnotneededforcal cul at i ons, butisgivenfor compari sonpurposes. Thecent rat eandcakedi schargeatapproxi mat el ythesamedi amet er and thereforeitisnotnecessarytoworkoutthecomponent sofpowerforthese twostreamsseparately. Werethedi schargedi amet erstobemarkedl y different,t hentherateforeachst reamwouldneedtobecal cul at ed, and equat i on(4. 131) wouldt henneedto be usedfor eachst ream.Forthelargerdecanter, assumi ngprocessdensitiesareclosetouni t y,pr oces s a c c e l e r a t i onpowe r , subst i t ut i ngtheappropri at evaluescal cul at ed intoequat i on(4. 131)(w =250, Qr =54.8, rd=480/2/10): Pp=548x2502( 4800) 2 . x2xl=54.8x10 ~ W 1000 3600xl OOx100 PI' -54.8kW From equat i on(4. 133 ), c onve yi ngpower :27r Ps =13x9.1x1000x6--0 =12.4x103Nm/s Ps =12. 4kW For thebraki ngpower,it is necessarytoknowthegearboxratio,whi chwill be t akenas98(from themanuf act ur er ) .Calculations and Scaling307 From equat i on(4.9),pi ni onspeed:Sp=2400- ( 13x98) Sv=1126r pm andfrom equat i on(4.10),pi ni ont orque" Tp=9.1x1000/ 98 Tp=92. 9Nm From equat i on(4. 134), br a ki ngpower " PB--92. 9x1126x =11x103Nm/ sP~- - 11kW 27l" 60 Fromequat i on(4. 130), thedecant errequiresatthebowlpulley,t ot alpower " PT- 3 3.6+54.8+12.4+11.0 PT--l l l . 8kW Theprecisemotorspecificationwilldependuponanumber of ot herfactors suchasthetypeofdriveandst art er, andhowmuchcont i ngency, or expansi on, forwhichtheuserwant stocater. Thelossesbetweenthemot or andthecentrifuge, suchasbeltfriction,andfluidcoupl i nglossesif applicable, allneedtobetakenintoaccount . However, themot orwillbenosmal l ert han 125kW,thenextst andardsizeabovethepowersofarcalculated. If amuch largersizeof mot oris chosen, thepowerfactorforthemot orefficiencywillbe reduced,imposingagreat er penaltyonthecostof electricity. 7.9DSScaling TableA.9intheAppendixgivesasetof14dat apointsforanunspecified effluent,whi chwill be usedto demonst rat escalingof DS data. Therequi rement istospecifyasizeof decant ert hat canprocessthesame sludgeat50m3/ hto give acakeof 30% drynessor better withthemi ni mumof pol ymerusage,andgoodcent rat e. Thesalientdat aforthetestmachi neand t wolarger machi neswor t hy of consi derat i onaregiven below. DecanterTestNo.1No.2 Bowl diametermm42 5575737 Clarifying l engt hmm80020001550 Bowl volume186385460 Nom.scrolling ratet ph/ r pm0. 281.002. 00 Max.conv. t orquekNm2.71620 Gearbox ratio125267254 Bowl speedrpm315029002400 Cake dischargedia.mm264326483 Ponddia.mm252306463 Thepondusedfor thetestmachi newas6mmaboveneutral. For thelarger machi nes10mmabove neut ral hasbeenchosenas t hat is knownto beagood worki ng level inpractice. Firsttheg-volumeneedstobecalculatedforeachmodel.Noticethat,in theg-VolcalculationsforDSwork, itistheglevelatthepondsurfacet hat is usedrat hert hangatthebowlwall,whi chismorecommonl yusedin centrifugework. Usingequat i ons(4. 105)and(4. 106), forthetestmachi ne,g- vol ume:(2 7r x3150) 2 g-Vol =86x60 g-Vol =120. 2m s 252 20x1000x981 Calculations and Scaling309 For largermachi neno.1, g-volume: --( 2r r x2900) 2x g-Vol385x60 g-Vol =553. 8rn 3 306 20x1000x981 andfor thelargermachi neno.2, g-volume" (2 7r x2 400' ~ 2463 g- Vol - 460x6~JX20x1000x981 g-Vol-685. 7m 3 Fromthesecalculationsitwillbeseent hat theg-levelatthepondsurface oneachmachi neis,respectively,1400, 1440, and1490, whichareallvery similar,indicatingthatscaleupshouldbest rai ght forward, providingthe geomet ry of thetest machi neandthetwolargermachi nesaresimilar. Thedatafrom thetest decant erareplottedingraphsin Figures7.4-7.7.The figurescalculatedaboveareusedwiththesegraphstoassessthelikely performanceofthelargermachinesproposedfortheduty. ForFigure7.4scrollingratesareplottedagainstdifferential.Thisindicates thatthetestscrollingrateis0. 28t ph/ r pm, whi chiswhat isexpected. Therefore,noadj ust ment is neededfor thisparamet erfor thelarger machines.2.50 2.00 ,Q .=1.50 Q. Q @ .x1.00 q ro 0.50 0.00 0.0 II II 19 J I/ II I t ' ZI I J I I ,I O: ;cries1] I ConveyorDi fferenti al RPM Figure7.4.Cake Ratevs.Differential. 310DSScaling InFigure7.5testdat aof cakedrynessareplottedagai nst t or que/ vol ume. It showst hat at or que/ vol umefigureof 2.0N/ cm 2 will besufficientto producea cake of 30%.Thus, for no.1 machi ne, c onve yor t or que " T =2.0x385/ 100=7.7k Nm andfor no.2machi ne, c onve yor t or que " T =2.0460/ 100-9.2k Nm Thesetorquesarewell wi t hi nt hecapabilityof thesize of machi nesselected. Itshouldbenot edt hat ifdriercakeswillbeobt ai nedinthefut urewi t h developmentof equi pment , t echni quesorchemicals, theconveyort orquewill increase.Reservegearboxt orquecapaci t ywillpermitsuchi mpr ovement and also result inlongergearboxlife. Themaxi mumcapaci t ypossiblemaybegaugedfromthegraphinFigure 7.6.Herecake drynesstest dat aareplottedagai nst feedr at e/ g- vol ume. Above thelinedrawnont hegraph, cent rat esareliabletobedirty,aswerethe cent rat esforthepointswi t hopensymbols.Belowthelinegoodperformance canbe expected.For therequi red50m3/ honmachi neno.1, O/ g- vol ume:34.0 32.0 30.0 ~e m : 28. 0 ~28.o 24.0 22.0 I t i 20.0 0.00 I , fI l I 36.0 J f l1 0.501.001.502.002.503.00 ConveyorTorque/VolumeNlcm = Figure 7.5. Dryness vs. Conveyor Torque:Volume Ration 3.50 Calculations and Scaling311 g-Vol O_ g-Vol =50/ 553. 8t7 -1 =9. 0x10 -2h -1 andfor machi neno.2,O/ g- vol ume"g-Vol 0 g-Vol =50/ 685. 7h -1 =7. 3x10 -2h -1 FromFi gure7.6,itwillbeseent hat fora30%caket heset wopoi nt sare comfort abl ybel owt heline.Fr omthisgr aphalso,itwillbeseent hat 50m3/ h onmachi neno. 1cor r espondstoabout 11m3/ hont hetestmachi neandfor no.2,about 8. 5m3/hont hetestmachi ne. Thisest i mat eisfromknowi ngt hatt heabscissaval uesfor t hepoi nt sare3,5,7,10m ~/h,etc. Thepol ymer doser equi r ement needstobeest i mat ednow. Thegr aphin Fi gure7.7plotscakedr ynessagai nst pol ymer dose,fort wotestr at esof7 and15m~/ h. Itisnecessarytoi nt er pol at eal ongt he30%linefor11and8.5 40.0 I1 38.0 36.0 34.0 32.0 lid lid 930.0 r Q 928. 0 I u26. 0 24.0 22.0- 20.0 0.0 II 1 I9 I l, 204.06.0 .' >.. / /i Poo=" Centrate -"~ 8. 010.012. 014. 016. 0 Q~Jg-Voll O~h"~ Figure 7.6. Cake Dryness vs Feed Rate:g-Vol Ratio. 9 18.020.0 312DS Scaling m3/ h. Ideallymoredatawoul dbedesirable,butfromTableA.9itisknown t hat 31%cakewasachievedat10m3/ hatapolymerdoseof 9.5kg/t.Thusa r oughestimateforthedosageondecantersno.1andno.2woul dbe, respectively,9and6kg/tdb.Afurthersmalltestmi ght beinitiatedtoobt ai n more precise estimates.As is often thecase,a choice needs to be madebet ween capitaland revenueexpendi t ure. A larger andthusa more expensive machi ne will consumeless polymer t hanthesmaller machine.Thepond dept hrelative to thebowl diameter is largerfor no.Idecant erand smallerfor no.2.Thiswoul dsuggestthereis someperformanceinhandwi t h theestimateforno.1andsuggestsomecautionwi t htheestimateforno.2. Thiswouldbringtheperformancelevelsforbothcloser,andt husthechoice woul d probably be biased t owardsthesmaller of thetwo. Itremai nstocheckt hat theconveyordifferentialsneededforthetwo scaled-up machi nesare wi t hi nthe worki ng rangeof thegearboxes specified. We t c ake rat e:OsPs=50x3/ 30x1.0tph Qsps=5tph No.Idi f f erent i al " N- 5/ 1. 0- 5rpm 38.0 36.0 34.0 32.0 i30.0 t Ir ~"28.0 a a26.0 (..1 24.0 22.0 20.0 0.002.00 1 V .S 7 4.00 f . . . . tf 6.008.0010.00 PolymerDosekg/t db I 12.00 JeT'm31h! II15m3/h .1 14.0016.00 Figure 7.7. Cake Dryness vs. Polymer Dose. No. 1di f f e r e nt i al range:2900 =0to=0to10. 9r pm 267 Calculations and Scaling313 No. 2di f f erent i al :N- 5/ 2. 0=2.5r pm No. 2di f f e r e nt i al range:2400 =0to=0to9.4r pm 254 Thisconcl udestwoverysatisfactoryscale-ups.Thescale-upshaveshown t hat therequi reddrynesscanbeachi evedwithbothofthet wol arger machi nesatthedesiredcapacity. Thecal cul at i onshaveshownwhatconveyor t orquesandwhat differentialswoul dbe neededoneachmachi ne.ThisPageIntentionallyLeftBlankCHAPTER8 InstrumentationandControl Thefirst product i ondecant erswerevi rt ual l ydevoidof i nst r ument at i onand control,apart fromthemai nmot orst art er. Today, i nst r ument at i onand controlsaremany[ l ],andcanbe quitesophisticated. Thepresentt endencyis forfullaut omat i on, tomi ni mi setheneedforhumani nt ervent i on, andreduce labourcosts.Improvedsafetyst andardshaveencour agedthedevel opment of someuseful,andreliable,i nst rument s. Thedevel opment ofsmall,affordable controllersthemselveshasenabledthei nt roduct i onofsomemuchneeded processi nst rument s. Hi t hert oanexpensiveprocessi nst r ument couldnot be justifiedto be usedmerely asamonitor.Whenadecant erisaut omat ed, aut omat i onofalotoftheassoci at ed equi pment isalsonecessary, t oget herwi t hinterlocking. Fori nst ance, it wouldbei nadequat etohaveadecant eroperat i ngaut omat i cal l yunat t ended iffailureofthecakeoff-takesyst emcouldoccurwi t hout communi cat i onof thefact to thedecant ercontrolsystem. Figure8.1depictsani nst r ument andflowdi agr amforadecant erpl antusingflocculant,withal t ernat i vecakedi schargeflowsfort hi ckeni ngand dewat eri ng. It is notpossibletocover everyevent ual i t ywi t honedi agram, butthis onecoversthemaj ori t yof usual si t uat i ons. Thei nst r ument at i onshownis notnecessarilyal waysused,butist hat whi chthepl ant engi neerwoul d consideruseful,wereitpossible.Theequi pment t hat couldbecontrolled aut omat i cal l y, oriscontrolledinst andardplants,ismarked. Eachoft hese possiblei nst rument swillbediscussedint urn, afterout l i ni ngthevari ous modulesof adecant erplant.! [ ~ 6 fellI"+| [| +| ~ljy~j|I +II 1~--~(~lThickening DewateHng . , . . . ,Kt? CControl Input EElectrical Amps/Watts LLevel nCount PPressure QFlow Rate sSpeed tTime TTorque 9Solids Concentration O!Temperature Fi#ure 8.1.An instrumentand.flow alia#ram for a decanter plant. 8.1DecanterPlantModules Afullyequippeddecant ercent ri fugepl ant willnormal l yhaveseveraldistinct moduleswi t hi nit: 9theflocculantsystem; 9theprocessslurryfeed system; 9thedecant er itself; 9thecent rat eoff-take system;and 9thecakedi schargesystem.Theflocculantsystem for adecant erplant, part i cul arl yfor thelargerplants,isusual l ysuppliedasaseparat eentity,wi t hitsowncontrolsystem.Someof these controlsyst ems canbe quitesophisticated, withdosingcontrolledfroma feedsolidsconcent r at i onsensor. Nevertheless,thereisnoreasonwhythis controlsyst em could not be coupledinto the mai n control system.Thereareanumber oftypesofpol ymermake-upsystem.Theone representedinFigure8.1ist heusual dualt ankbat chmake- upsystemfor solid-gradepolymers. Itbasicallycomprisesapowderhopperwithascrew feeder,di schargi ngintoastirredvessel.Thevolumeof wat er iscontrolledby level probesinthisvessel. Thecont ent sarestirredforafixedtime,toallowt he pol ymertodissolveandagetoitsfullpotency. Aftertherequiredagei ngtime, itisaut omat i cal l yt ransferredtothepol ymersupplyvesselwhenact uat edby alow-levelsignalfromthissecondtank. Thepolymerpumpis controlledfrom thedecant ercontrolsystem. Thefeedwill be suppliedfromthemai nplant.This couldsimply beateeinto apipelineof t heplant,ormoreusual l yfromastoraget ank. Avariablespeed pump, usual l yaprogressivecavitytype,feedstheprocessslurrytot he centrifuge.Therateis fixed manual l yor by aplantcontroller. Thedecant ersystem itself hardl yneedsfurt herdescription.Themai nmot or andback-drivemot orsaret hemai ncontrolinputs.Largerdecant ersmay haveaseparat eoill ubri cat i onsystemforthemai nbearings, inwhi choil flows,t emper at ur esandpressuresare moni t ored.Thecent rat eoff-takesyst emisgenerallyalargepipetodrain,ortoa receivervessel.Occasionallythedecant erwillbefittedwi t hacent ri pet al or 318Decanter Plant Modules skimmerpump,whenapressuriseddischargewilloccur,whichmayhaveto bereleasedbelowtheliquidproductsurfaceinthereceiver,topreventor reducefoaming.Inthree-phasedecantersasecondlightphasedischargewill be present,the flow of whichwill also need measuring.Dewateredcakeisoftendischargedontoabeltconveyor, straightintoa hopper,orperhapsintoascrewconveyororelevator.Wheredecanters employnegativepondoperation,pondsdeepert hanthecakedischargelevel, unwant edliquiddischargefromthecakeoutletcanoccurduringstart-up. Thiscanproduceanunpl easant messonbelts,causi ngt hemtoslip,andwill cont ami nat etheproduct.Thisissometimespreventedusingnotchedweir plates,orspecialstart-upandshutdownprocedures.Alternatively,devices are fittedunderthecakedischargeto feed thewetcakebackto thefeed vessel. Thesedevicescouldbe,alternatively,aflapdiverter,orahopperthatis automaticallymovedunderthedischargeatstart-up.Theunwant edliquor dischargeisthenpumpedbacktothefeedtank.Afurtheralternativeisto anglethebeltconveyor,suchthatliquidflowsbackdownthebeltintoa hopper,whilesolidsconveyupwardsonthebelt.Withallthesedevicessome flush may be requiredafter the wash-out has ceased. Thickenedcake dischargecansimply be intoa hopperwhichis emptiedby a pumpactuatedbylevelprobesinthehopper.However,moderntechnology oftenrequiresthedischargeto bemonitoredfor solidscontent,if notrate.For thisasmallstirredbuffert ankis used.A samplefromthistankis pumpedand recycledcontinuously to provideacont i nuoussample.Thestirredbuffer tank issizedtosmoothoutmajorfluctuationswhichcanoccurinthedecanter discharge,due to hold-up in the casing. 8.2Instrumentation Thissectionisseparat edintothevariouscategoriesofi nst r ument s, suchas flow meters,solidsconcent r at i onmetersandtimers.Lastly,controllerswill be covered. 8.2.1Flowmeters Flowmetersforaqueousslurriesarereliable,accurat eandseldomrequi re adj ust ment afterinitialcalibration. Moreovertheyareamenabl efor connect i ontoPLCs,comput ersandcontrollers. Themostcommonmodels usedondecant erplantsareeddycurrent andul t rasoni ctype.Flowmet ersare usedonthefeed lineafter thefeed pumpandsimilarly onthepol ymer line. Flowmeasur ement is employedontheoill ubri cat i onlines,butisusual l yof therotameter, orvariableorificetype.Thismeanst hat t heyareusedfor indicationonly,andarenotreadilycoupledintothecontrolsystem,unless simply as al armfeatures. Ideally,a flow met ershouldbefittedonthethickenedcakerecycleline.This isbecausethesolidsmoni t orworksontheprinciplet hat thecakesolids concent rat i onisafunctionofviscosity,whi chint urnismoni t oredasa pressuredropwhenflowing.Thustheflowratealsoaffectsthepressuredrop, andthereforemust bekeptconst ant . However, oftenitisfoundt hat pl antusersrelyontheconst ant rat efromamet eri ngpump, maki ngperiodic adj ust ment sto flow or calibrationshouldthepumpwear.Thetotalflowofthickenedcakeisusuallyobtainedbycal cul at i on, buta checkcanbemadebymeasur i ngthetimei nt erval sbet weendi schargeof the sumptank.Thiswouldbehowtheoilflowismeasuredfromat hree-phase decant er.Ifdilutionwat er isused,thisisgenerallymeasuredwi t har ot amet ervariableorificemeter. Howeverif thisflowhastobei nt egrat edintoacontrol system thenanelectronicmet hod, as usedfor thefeed, will benecessary.8.2.2Solidsconcentrationmeters Thesemoni t orst endtobet hemostexpensivei nst rument s, butenabl ethe mostsophisticatedtypeof processcontrol.Wi t hout t hem"l i ve"meas ur ement320Inst rument at i on of,forexample,solidsrecovery,polymerdosage,cakeandcent rat erates, andproduct qualitywouldnotbepossible.Somel aborat oryanalysestakea fewhourstoperform,bywhi chtimetheplantcouldbewayoff thecontrol desired. Afewcompani esofferdevicest hat cancont i nuousl ymeasuresolids cont ent of feed flows.Variousprincipleshavebeenused,i ncl udi ngthecoriolis effectandtheuseofaradioactivesource.Themet hodusingaradioactive sourcehasprovedreliable,butthereisresistancetousingitwherea wat ercourseisinvolved,andmoreovertherearest ri ngent regul at i onswi t h regardtothedisposalof thei nst rument onceitisattheendof itsusefullife. Neverthelessthesuppliersnat ural l yofferacomprehensi veservice.Light reflection or t ransmi ssi onis anot her met hodt hat is in use. Centratesolidsconcent rat i onmeasur ement isani mpor t ant par amet er for decant ercontrol.Severalsuchi nst rument sareavailabletomeasureinthis range. However, oneproblem presentedby cent rat efrom adecanter, onmany applications,isthecopiousproduct i onof bubblesorfoamintheflow.These bubblesareread,bymanyi nst rument s, assolids,t husprevent i ngtheuseof suchdevices.De-aerationofasampleflowofthecent rat ehasmetwi t ha modi cumofsuccess.Somedecant ermanuf act ur er sdevelopingtheirown i nst r ument [2]haveobtainedmore success. Cont i nuousmeasur ement of solids,or moi st urecont ent , of dewat eredcake, asfarasis known, hasnotbeenpractisedondecant ersyet.However, infrared devices,whi chcanmeasuremoisturecont ent of productsonconveyorbelts, whenpositionedabout 30 cm above thebelt.havebeenreported.Thesolidscont ent of theflocculantsolutionwoul dbeausefulparamet erin anyintegratedcontrolsystem.However,thesolidsaredissolved,andthe concent rat i onsinvolvedareafractionof1%,wi t hanaccuracyrequi rementprobablydownto0. 001%. Moreoverviscosityofsolutionscoversawide range, andis dependent uponanumber of ot herfactors,nottheleastof whi ch ist emperat ure. Therefore,itisnotpracticabletomeasurepolymer concent rat i ondirectly.However,thereisnoreasonwhythepowderfeeder couldnotbecalibratedfor thepowderinuse,anditson-timemeasured. With liquidpolymer make-up, theon-timeof acalibratedrawliquidpol ymerpump woul d be measured.8.2.3 Level probes Thereisnogreatneedtomeasurelevelsintheplant,butmerelytohavean indicationof whet her at ankorhopperisempty, full,orinbetween. Thisis achievedbyconductivity, orsonic,probes.Theyarefittedtothetwopolymer tanks,toinitiateanewbatchmake-up, andtoact uat etransferbeforethe pol ymersupplyt ankempties.Probescouldbeemployedinthepolymer powder hopper toguardagainstr unni ngoutduri ngoperation. Smaller plants will not usepowder probes,and rely onasystem usingseveral days' supply. Instrumentation and Control321 8.2.4 Speed probes It is part i cul arl y necessary to measur e the speed of r ot at i onof the decant er bowl andthegearboxpinionshaft.Occasionallyat achomet er willbebuiltintot he braki ngdevice.Moregenerally, bowlandpinionspeedsaremeasur edby proxi mi t y probes,act i ngonapr ot uber anceorcast el l at i on, onaspigot,hubor shaft. Thespeedsof thefeedandpol ymerpump, andalsothecakesamplepump,areusefult houghnotabsolutelynecessarytomeasur e, ascompari ngthis speedwi t hacal i brat i onspeedwill indicatetheonset of wear.Measuri ngthespeedofthepol ymerscrewfeederhasal readybeen ment i oned.Anot herusefulspeedmoni t orwoul dbeont hesolidsconveyordri ven shaft.Allt hat isneededhereisanindicationt hat theshafthasstopped,for i nst anceifthebeltshouldbreak.Inanon- at t endedplantitisessentialto knowiftheoff-takesystemceasestofunction, sot hat thefeedmaybe arrested. Itiswort hnot i ngt hat themaj ori t yofdownt i meofadecant erpl ant is causedby failuresinancillaryequi pment , rat hert hanthe decant eritself. 8.2.5 Temperatureprobes Thet emper at ur eof thel ubri cat i ngoilfromthebeari ngsisusual l ymeasur ed witht hermocoupl es. Thet emper at ur eof thefeedisonlymeasuredif thisisan operat i ngparamet er. Thet emper at ur eofmot or wi ndi ngsareusual l y moni t oredbyt hermi st ors, connect edtoasafetycut -out syst eminthemot or controlgear.Obt ai ni ngadirectreadi ngof mot orwi ndi ngt emper at ur ewoul d be unusual .8.2.6 Torquemeasurement Conveyortorquetodayisanessentialpartofdecant ercontrol. However,directreadi ngofconveyortorqueisverydifficulttoachieve.Evendirect readingof piniontorqueisdifficult,butcouldbedoneusingst rai ngaugeson thepinionshaft.However, themostusual met hodis touseacal i brat i onof t he braki ngdevice.Thecontroldeviceforthebrakewillgivearead-out , on request,of thebraki ngtorque.8.2.7 Timers Timersarei nt egral partsof someof thecontrolsystems. Theyareusedint he starterof themai nmotor, toswitchfrom startodeltaoperation. Theyareused incontrolsystems,forthesequent i al st art -upandshut - downofanci l l ary equi pment . Timersareusedfortheageingof thepolymer, andtheon-t i meof 322Inst rument at i on thefeeder.At i merwouldbeusedtomeasur ethefill timeof thecakesump, to checkcake rate. However, al t houghrun-t i memeterscanbefittedtomost motors, thisis usual l y, if at all,onlyonthe mai nmotor. 8.2.8Counters Count ersareusedtocount bat chesofpol ymermadeup,tokeepanoverall checkonusage. Cumul at i veflowisoftenfoundinelectronicflowmeters, to keep account of totalflows t hr oughthe plant.8.2.9Electricalmeters Thecur r ent tothemai nmot orisoftenmoni t oredtoprevent overloading. It alsogivesanindicationofthepowerbeingconsumed, al t houghabetter device for thisis thewat t met er.8.2.10Bearingmonitors Int erest isnowbeingplacedini nst rument sthatmoni t or theheal t hof beari ngsinoperation. Pr emat ur efailurecanbepredictedbeforeexpensive damageoccurs.These i nst r ument sare notyet inwide use. 8.3ControlledEquipment Thecont rol strategyforadecant erpl ant oftenwillhingeonexperience,theuser' srequi rement s, what isavailableandtheextentofcont rolrequired.Oneofthemaindecisionstomakeisregardi ngtheflocculant control.Theoptionforflocculantcontroliswhet her tohaveafeed-forward control,requi ri ngafeedsolidsmeter,orwhet her tohavefeedbackcont rolusingacent rat esolidsmonitor. Withfeed-forwardcontrol,theflocculant rateismodul at edaccordingtothelevelofsolidsinthefeed.Theratioof flocculanttofeedsolidsmayhavetobet ri mmedoccasionallyshouldthe qualityofthefeedvary.Withfeedbackcontrolthecontrolperformanceis i ndependent offeedquality.Neverthelesssomecent rat emoni t orscanbe badlyaffectedbyaerat i onandfoamwhi chcanoccurwithsomepolymers andfeeds.Theext ent ofthesophisticationofthecontrolwilldependupon howmuchoftheplantisrequiredtobei ncorporat edintothedecant er system.Thegoodfunctioningoffeedtanklevels,off-takepumpsand conveyorsallmayneedtobebrought intothestrategywithappropri at e interlockcontrols. Todeviseacontrolstrategyforadecant erplant,itisnecessarytoknow what devicesareavailabletothecontroller.Thesemaybeon/offdevices,or devices whi chcanbe variedin out put by thecontroller. 8.3.1On/offdevices Thesewillincludethestirrersinthepol ymersystemandthickenedcake sump.Alsoincludedwillbecompletemodulesystems,suchasthepol ymer system,theoill ubri cat i onsystem,andperhapsthecakeoff-takesystem.The decantermai nmot orisalsoacontrolledon/offdevice,al t houghavari abl e speed mai nmot or canbe employed. Thepumpsact uat edbythelevelprobesonthepolymersystem,andthe sumpdischarge,arealsoon/offdevices,asarethebeltconveyor, thecake diverter,andthepol ymer screw feeder. Inacompletelymanual plant,eventhefeedandpolymerpumpscouldbe on/off,andmerely controlled on or offby safety interlocks. 324ControlledEquipment 8.3.2Variableoutputdevices Thesearemai nl ythefeedandpol ymerpumps, andthedecant er braket orque orspeed.However, inspecialcases,theact ual bowlspeedcouldbeapart of a controlstrategy. Theponddept hitself,usi ngt heinflatabledam, couldbeused ina t hi ckeni ngcontrolst rat egy.Thepol ymerfeedercouldbeusedinacont rol system,if widerangesof feed concent r at i onwereto beant i ci pat ed. As farasis known, thishasnotyet been used. 8.4Controllers Modernelectronict echnol ogyhasprovidedi ndust rywithawidechoiceof small,user-friendly,cost-effectivecontrollers, withproport i onal integraland derivative(PID)controlaction.Thesecanbeusedindividuallyonthei nputflows,orintegratedintoamast ercontroller. Itwoul dnotbeunusual tohave oneeachonthefeedandpolymer pumps. Asignalfromtheflow met erwoul d be suppliedto thecontroller,whi chwoul dadjustthespeedof thepumpto give aflow agreeingwiththesetpointenteredbytheoperator. Thesetpointcould be set alternativelyby a mast ercontroller.A separatecontrolleris suppliedwitheachpolymer make-upsystem.When energised,itwill,accordingtoinputtedsetpoints,controltheon-timesof make-up wat erandpolymer feeder.It will controltheageingtime inthemake- up tank,when the stirrer is switched on and when switched off. Transfer of aged polymerwillonlybeallowedwhenthesupplyt ankisbelowacertainlevel, whenthetransferpumpisenergised,andafterwardsde-energised.Thisisa simple but very effective system. Therearesome vari at i ons from manuf act ur erto manufact urer. Thepolymer is very hygroscopicanddifficult to dissolve,and ifnothandl edproperly,cancreateani nordi nat emess.Onemanuf act ur ersuppliesanairblowertotransferthedrypowderintoacyclonewet t i ng chamber,to minimise the onset of glue-like deposits in the lines. Thepolymercontrolsystemcanbeaugment edwi t hafeedsolidsmeter,to give"feed-forward"control,fixingthekg/tdbpol ymerusagetoanoperat or set point. Themai nmot orcontrollerisaseparatecontroller, anddependsuponthe typeofinstallationandmotor. Themot or couldbeAC,DCori nvert ertype. Rarely,itcouldbeahydraul i cmotor. ThestartercouldbeDOL(direct-on- line),part i cul arl yifafluidcouplingisfitted,itcouldbeasoft-starti nvert er system,oraDCsystem.Withani nvert ersystemthcback-drive,alsoan invertertype,couldbeconnect edt hr oughtheDCbustoallowpower regenerat i on. Thestarteritself couldbeact uat edbyaseparatemast ersystem. Undoubtedly therewill be interlockswi t hthestarter,to causeit to de-energise with cert ai nscenarios. Allthecontrollersarei mport ant , but themosti mport ant controllerforthe processistheonecontrollingthegearboxpinionshaftbrake.ThisPLC 326Controllers (programmabl elogiccomputer)willberequi redtocontrolthebrake,ei t her togiveasetconveyordifferential,orasetout put torque.Whilstthisduty, as specified,seemssimple,theoveralldut yexpectedmakesit,i nt ernal l y, quite complex.Incertainci rcumst ancesit is requiredtocontrolspeedsclosetozero andeventoreversespeed.Itisexpectedtobesuitablefor thecompleter ange ofamanuf act ur er ' sdecanters,andyetexpectedtocontroleachwi t hi nsafe limits.Moreoveritneedstobeappreciatedt hat reducingout put t orque allowsalowerdifferential,whi chincreasestorque!Thustoallowahi gher conveyortorque, thecontrollereffectivelyhastoreduceitsout put t orque.Nevertheless,excellentcontrolshavebeenestablishedonseveralt housands of installations.A good brakecontroller will be requiredto indicate: 9Bowl speed: 9Conveyor differentialspeed: 9Brakeor conveyor torque; 9Tor quehi gh/ l owalarm; 9Differentialhigh/lowalarm: 9Status(start-up or runni ng);9Mode of control(torque/differential): 9Set point. Accessisneededtotheoperat i ngparamet ers, withanencrypt edcodeto preventunaut hori sedtampering. One suchcont rol l er is shownin Figure8.2. Onlyafterusingsuchani nst rument cantheextentof theneedsforsucha device be appreciated. Theoperat i ng paramet ersmay include: 9Entry code; 9Modes permitted: 9Upper andlower alarmlimits; 9Set points: 9Set poi nt l i mi t s;9PID settings: 9Secondary PID settingsfor two-stage control; 9Senseof alarms(normally on or off): 9Gearbox ratio; 9Pulsesper revolution for probes; 9Pulley ratios for speed recalculations; 9Control ramp rate; 9Calibration of external signals; 9Paramet ersfor t ransmi ssi onof data; 9Paramet ersfor comput er communi cat i on;9Braket or que/ cur r ent calibrationreference. Instrumentation and Control327 Thisisaverybriefsynopsisofwhat couldbe60ormor esepar at e par amet er [email protected] 0000 [email protected] 0000 ~.:AlfaLaval Figure 8.2.AnAlfa Laval AutomaticBackdrive Controller (,4 BC). 8.5IntegratedController Withseparatecontrollersalreadyprovidedtocontrolseparatefunctionsof theplant,itisanobviousnextsteptointegratethemintoonemaster controller,ortosupplyamastercontrollertosupervisetheindividual controllers.This is beingdemandedincreasingly.Some largeplantsdemanda centralremotecontrolroom,withmimicdiagrams,allcontrolledbyone central,large industrialcomputer. Somedecantermanufact urersalreadyhavetheirownintegrated controller,allwithvarying degreesof sophistication.Some of thedutiesof an integratedcontroller are described below. Allthesignalsavailable,showninthediagraminFigure8.1,needtobe continuouslyfedtothecontrollerandconvertedtodigitalfigures.Itshould be possible to display any of these figures on request. Thefiguresthenneedtobeprocessed,accordingtotherelationshipsin Chapter4,to provide figures of: 9Solids recovery; 9Polymer dosage: 9Torque/volume; 9Feed rate/g-volume; 9Centrate rate; 9Cake rate; 9Cake rate/differential; 9Powerusage on the decanterandthe total. These should all be displayable. Acostdisplayshouldbepossible,onceapplicationitemisedcostdataare inputted.Thedatarequiredfor aneffluentwouldincludethecostof powerat varioustimesoftheday,costofeffluentdisposal,polymercost,andcake disposalcost.Othercostst hat maybeincludedwouldbe,forexample, amortisationof capital.The processor would thenwork outtheplantrunni ng costs for display, or periodic print out. Thecontrollerprocessorwouldhavein-builtcontrolalgorithmsforthe plantmanagertoselect.Controlcouldbetominimiseoverallcost,maximise Instrumentation and Control329 dryness,minimiserevenuecosts,ormaxi mi set hr oughput . Itcouldalsobeon the basisof keepingthefeed t ankdownto a cert ai nlevel.Prioritieswoul d need to besetfor thevari ousperformancefactors,suchassolidsrecovery, dryness,costandt hroughput . Maxi mumandmi ni mumlevelsforeachwoul dneedto be set. Thecontrollerwouldbesetonacont i nuouslooptoconduct the calculations, performcontroladj ust ment s, displayandifnecessarypri ntresults,andactasanannunci at or for al armsandmai nt enanceschedules. Thecontrolmet hodcouldbeasimple"hillclimbing"t echni quewher e smalladj ust ment sofonevariableatatimearemade, andperformance checked.Theadj ust ment cont i nuessolongasperformanceimprovesanda stepbackismadeonceadet eri orat i onisdetected.Thenext variableist hen adjustedinthesameway.Adj ust ment stepscouldt henbereducedonceall variableshave beenused.Theprocess is t henrepeated. Anal t ernat i vecontrolmethod, whi chisanewl ydevelopedtechnique, uses a t echni quecalledfuzzy logic[ 3 ]. 8.6CIP Theequi pment usedfor theCIP feat ureis describedinSection2. 4. 14. A small PLC,oranadj unct tooneof thest andar dcontrollers, isrequi redtosupervise theCIPoperation. Anoperat orgivingthest art commandtothePLCor equi val ent willinitiatetheCIPsequenceof events. Thefeedwillbestopped. Thenthemai nmot orandback-drivesystem will bede-energised, andallowed tocoastdowntot herequiredCIPspeed,whent hemai ndrivedonkeymot or willbeenergisedandtakeovertorot at ethebowlatsuchaspeedasto generat eslightlylesst hanl g(about 70%ofl g). Themoresophi st i cat ed systems will alsohavea donkeymot orto rot at et hegearboxpinion.Timerswillcontrolthedurat i onof thelow-speedr unni ngandot hertimers willopenvalvestoadmi t cl eani ngfluidintothebowlandintot hespraybars onthecasing.Somesystemswillperiodicallyreversetheback-dri vedonkey mot ortoreversetheconveyordifferential.Thisfeaturemust beusedwi t h caut i on, asreversingthescrollingcouldj amsolidsbet weent hefrontbowl hubandtheconveyor, andul t i mat el ybendconveyorflights.Theprogramin thePLCwilldictatethedurat i onoftheCIP,thedurat i onof eachphase,and howmanytimestheconveyor, andifnecessarythebowl,arereversed.The programwillalsodictatewhen, andforhowlong,thecl eani ngtluidsare applied. TheCIPfeatureisavaluableassetinfoodandphar maceut i cal processing. Theabilitytokeepthedecant ercleanandhygienic, wi t hout theneedto di smant l eit,hasenabl edtheuseof decant ersinprocesseshi t her t oimpossible. Decanterscanber unformanymont hswi t hout di smant l i ng, wi t hacceptable st andardsof cleanliness. Wi t hsuitabledesignsofdecant er, theCIPprocedurecanbeused,where necessary, for sterilisation,insteadof,or with,chemi cal cl eani ng.8.7References lWLeung,P Wardell,L Hales.(Baker HughesInc.)Methodandapparatusfor controllingandmonitoringcontinuousfeedcentrifuge.US Patent5948271, 1 December 1995 2J G Joyce(AlfaLaval)Turbiditymeasurement.USPatent5453832,6March 1991 3CyonAltrock,BKrause.Fuzzylogicapplicationnote:optimizationofa water treatmentsystem, ThisPageIntentionallyLeftBlankCHAPTER9 The Decanter Market In atotal worldmarketfor liquid/solid separat.iriri eqiiiprnent. of' about $6 billioris(coveririg;illappliciitions,domesticandinstitutionalaswellashdUSt r i al ),the decanter hascome tobean importantcornponeat. with;t marketshareof ahoui10'K (11'that figurc. This chapter looksbrieflyat. lhe market hi- decanters, tiow i lis made up, and how it is expeckd I,{)develop. 9.1MarketCharacteristics Thedecant ercentrifugeisani mpor t ant processingtool,butisbynomeans cheap,sothedecisiontoinvestinanewdecant erisonet hat hastobet aken withcare.Themar ket ischaract eri sedbythepresenceinitofafewlarge suppliers,withmanyyearsofexperi enceandwi t hwider angesoftypesof decant eravailable.Thereist henagroupof smaller,general suppliers,plusa handful of nichemar ket suppliers(mostlytooliveoilproduct i onandsimilar applications).Thereiscert ai nl yenoughexperienceavailableinthemar ketplace,toenableanypot ent i al pur chaser to obtainsat i sfact oryquot at i onsfora newmachi nefromanumber of competitivesuppliers. Thepurchaseof anewdecant erisverystronglyinfluencedbythei nt ended processduty, andal most allsuchpurchasesaremadeonlyaftercareful analysisby thesupplierof therequiredperformance, and. possibly,aftersome kindoftrialwitht hecust omer' sprocessliquor.Trialsmayinvolvet he installationofat empor ar ytestdecant er, andanci l l aryplant, asastaticor mobilerig.Suchatestmaybeforanextensiveperiod,tocoverallthelikely vari at i onsinprocessslurrycharact eri st i cs. Thetestrigcouldbeafullsize, pilot scale,or l aborat oryinstallation. Experienceof apart i cul arapplicationby thesuppliermakest heselectionprocessonewhi chcanbeapproachedwi t h confidence,andthepotentialpur chaser woulddowelltoenqui reastot he level of rel evant experienceavailable. Themaj ormanuf act ur er shavesalesorsubsidiarycompanyofficesinmost,ifnotall,thelargerindustrialcountries, andlocaltomanyofthelarge decant ermarkets. Theseofficesareusual l ystaffedwi t hverycompet ent sales engineers, abletoconvert thesupplier' sweal t hofexperienceintoa prel i mi naryquot at i onquiteeasily.Suchastartmust usual l yt henbe followed by thetrial processal readyment i oned.9.2MarketTrends Asisshowninsomedetailinthishandbook, thedecant erisanext remel y versatileprocessingdevice,byvirtueof themanydifferentitemsof itsmake- up t hat canbe changedto suit theprocessneeds.Inthisway,thedecant erhas beenabletomeetawiderangeofprocesschallengesoverthepasthalf- century.Themai ntrendinthemarket placecant husbeexpectedtobeast eady i mprovement indetaileddesign,toenablethedecant ertomeetfurt hersuch challenges.Thesei mprovement swillspreadtoallthemainsuppliers,sot hatthechoice ofdecant ersourcewill stillremai nwide. Themaj orapplicationgrowt hwill cont i nueto beintheprocessingof wast e slurries,andthis dut yrequiresasinexpensiveamachi neas possible,al t hough coupledwithquiteadvancedspecifications,inordertoachievehighdryness figuresinthedischargedsolids. ThetrendsidentitiedinChapters1and2willimpactonthemarket , butthe majormarket i nginputcont i nuestobetolettheworldoftheprocess industriesknowwhat ausefulthingthedecant eris.andhowitcansolveso manyliquid/solid processingproblems. 9.3MarketSizeEstimates Theest i mat i onoft hesizeofthedecant ermar ket isbesetbytheusualproblemsfacedbyanyat t empt toenumer at eamarket : definitionof scopeand avoi danceofdoublecount i ngbeingtwoofthemostdifficult.Currency vari at i onscanhaveamarkedeffectonsizeest i mat i on, especiallywher e historicaldat aare beingext rapol at ed.Marketsizeest i mat i onsmaybeapproachedfromtwodirections:t op-down andbot t om-up.Thedownwar dsappr oachstartswi t hnat i onal ori nt ernat i onal dat a, for tradeandproduction, andusestheseto derivecomponent sof themarket . This met hodismadedifficultbythelackofcommonidentityamongcategoriesof data,andbytheomission, certainlyfromnat i onal product i ondata, ofmostsmallcompani es. Theexistenceof asinglesupplierinanat i onal mar ket may also be asufficientreasontoomit thefiguresfrompublishedstatistics. Theupwar dsapproachstartswithindividualcomponent sof thepart i cul ar market , andaggregat est hemtoarriveatanoverallfigure.Thesecomponent s maybethesalesintopart i cul arend-uses, orthesalesbyi ndi vi dualmanufact urers. Itisinthismet hodt hat theproblemsofscopeanddouble count i ngaremostlikelytooccur.Manycompanies, fori nst ance, donot differentiatebet weenmachi ne- onl ysales,andallof theanci l l aryworkt hat is doneto makeupa finalsalescont ract , or of thesize of after-saleswork.Theresultof thei nadequaci esof eachof thesetwoapproachesist hat bot h havetobeusedt oget her, toderiveanest i mat et hat appearstosatisfyboth.Theremaythenbe,asisthecasewi t hthedecant er, ot herpublishedmar ketsizeestimates, whi chcanbeusedtocorroborat etheresultsofthedirect analyses. Thesepubl i sheddatararel yagreeveryclosely,once(andif)a commonbasiscanbeestablished, but t heydogiveouterlimits toacal cul at ed figure,andpresentsomeconfidenceasto theresultsof thework.9.3.1Overalldecantermarketsize Bymeans, then, oft hemet hodsjustoutlined, atotalworldmar ket fort he decant ercentrifugehasbeenderivedof $625millionsfor2000, atami d-year valueof theUS dollar.Thisfigurehasaprobableaccuracyof +10%. itrelates thepriccpaidbythat customer: i t coversthesaleorwholedecanl ersonl y,:indr i o t;ilrs work; i t coversf.he salea1.t hei.irne ofthe wholemachine of;inysi;jnd;ird supplyof sp:ireparts, but not ofany spares for thatmachine saidlater: i dit ex~l i i desill1xiditional equipmentsoldwith thc dccantcr thatis not necessary fur the safe and cficient operatioil or the riiacliirw. 4 This salcs valuc corresponds to a tigurc i rithe regioiiof1500 10.3000 Ibr Ihe number of decantercelltrifuges t o he sold in 20(')0. I tisexpectedthatthedecaliterrnarkei,whi chhiishcengrowingquite strongly in size sirice the eiid ofthe recession ol thc early1 Y Y Os, wi l lcontinue thi s growt h p;ittrrri,Over the next tive pears. indccd. t l ~ c markct i st-xprt.tet1 t u growat brlwyerii 4 iind4.5%pcraiiiiuiii(i.c. cnmbrtabl yiricixrt?ss ol' thc expect ed iiicreasr in gross dnint?stic product tigurcs). 3 j . l ' i L foral lwater aridw;ir;te watertreatment. industrial as M C I I a srri i in Ici pal : (1.3%lor fuel malt'rral extraction atid processing; 19.1'XIfor food arid brvwi gc proccssing: 41 3.4' %,for minerah arid hulk inorganic chcrnicala: I 1 0.4'K for h l h organi c chcinicals and petrochemicals: I 8.i'%,fur finechuinirals and pharmaceutical?; iind a7.[1'%for other applications. 338Market Size Estimates 9.3.4Suppliers'marketshares Thesuppliersofdecant erstotheworldmar ket arement i onedinChapt er1, whi chincludesalistofmost ofthemanuf act ur er sknowntobeproduci ng decant ersin2000. Ifal l owanceismadeforthecompani esnotlistedin Chapt er1(believedtobeallsmallones),t henthemaj orholdingsofmar ketshareare: 936.8%by Alfa LavalandTomoe; 916.8%by Baker Process(BirdMachi neandBird Humboldt); 99. 6%byFl ot t weg;98.8%by Westfalia;and 96.4%by Pieralisi. Sharesofbet ween1.5and3.0~areheldbyBroadbent(withTanabe),Guinard, andSiebtechnik. andofbet ween0.5and1.5~byAmenduni ,Centriquip,Centrisys.Hiller,Hut chi son-Hayes. Noxon,andPennwal t India. Thisleavesamar ket shareof6.1%heldbytheother, unspecifiedcompani es.Itcanbeseent hat thefivelargestcompani esholdalmostfour-fifthsofthe totalmarket .CHAPTER10 Suppliers' Data This chapter lists the mai n rnari~1I 'a~:~i~rt:rsofd(:r:arllers, together with details of thcir company structiirc and of their ranges ofdecanter centrifiige. This is lint an cxhausti vr list. but it includes data alrcady in the public domain, issued bymanufacturersintheirbrochures.and augmentedbydatasuppliedby some ofthc companies Ibr thc purposcs of'this book. 'L'hc coveragc hcre is intcndcd to be that ofallof the mainmanufacturers. pl us as ~iiaiij'othcrs a scould bc located. 1Pci t each ri i an~i fa~turerare given the si i l i cnl r i 1c l . s i i l >(>11i i1.sIieatiq~iiirters irrirlut ticrilrlrlresses,plusits mariufat:tutirig rurlpci. arid othcr iriforrn;iliond u s t :t o the rtwdtir. l ktai l s ofits drcarilcr rriodels art! I hrri ti ~hul u~etl . 'I'hisiriforniu~ivri isprvvirledl uenitblethereadertodevelopsome parnmctcrs ofchoicc when a ncw purchase ofzidecanter is to be undertaker]. The dat aglvcnshouldnot,howcwr, bcuscdTordesigniiiidspec-ilicaliori purposes.butniorcforinitialstudiesa s to whatcoi ~l dliep s s i t ~ l r . or frir corn par is(>nstud i cs. F i11a 1rcc onim e 11d a t i o 11 ss 11 o u Id;I 1w ;I y s hr!so11 g h tfrom thc prcftrrcd suppliers, Thc ent.rics arc i n :Ilph~ibetic:il nrrlcr, and no attempt has been rnadc by the authorstuberritriclivcinanycntry.IJ ndertheheading"company uwnership", rriutitionis rn:itic!of m:ijorclwncrr;lIip by another soriipariy. or 01 thc cxistonce ofmaj or cqiiity holders. Otherwisc. ownership is assuo~edt.o hr private o rpublic: sharc ownership. according to t h t b type oTc.orrlp:iny. khtricsundcr"othcrmainbiisiacsscs"refert,ooI,}ierrjon-dscanter aotivitirs ofthc riameti c.ornpilny. whilc "othcr company c'oriricctioris" refbr to business associations spccific l o the decanter husiriess. The datagiven under thc hcading "decaillersales"art! eithttrsiipplicd by ~niin~Il'ilct1irerin qiiestion. or cstimated by thy authors. Set.