A simple estimation method of materials

handlingspecificenergyconsumptioninHPGR
circuits
AlexDoll
Sagmilling.com,Canada

ABSTRACT
Multistagecrushingplants,includinghighpressuregrindingroll(HPGR)circuits,requiremore
materials handling equipment than SAG and ball mill circuits. The energy consumption of
materials handling are sometimes neglected when doing preliminary assessment or desktop
studycomparisonsofHPGRversusSAGmillingcircuitsbecauseestimationofmaterialshandling
power requirements can demand significant general arrangement drafting that is not available
whenperformingpreliminaryassessmentstudies.
The largest component of materials handling power is consumed by conveyors when lifting
materialbetweenstagesofcrushing,meaningthatasimplepotentialenergymodelcanbeusedto
evaluateconveyingspecificpowerconsumption.Thispotentialenergyconsumptionforconveying
canthenbefactoredtoprovideoverallmaterialshandlingspecificenergyconsumptionsuitablefor
desktop studies of HPGR and other multistage crushing circuits. The technique also permits
differentcrushingcircuitflowsheetstobeevaluated,atapreliminarylevel,bysimplycountingthe
numberoftimesaconveyor must "lift" orefrom groundlevelup to abinorother equipment
mounteduphigh.
Somecommentsaboutcircuitdesignandancillarysystemsareofferedwiththegoalofimproving
theeffectivenessofpreliminaryassessmentcircuittradeoffstudiesofHPGRversusSAGmills.

Keywords:Studies,HPGR,conveying,materialshandling

1
INTRODUCTION
Tradeoffstudiesofmultistagecrushingplantsinvolvinghighpressuregrindingrolls(HPGRs)
versussemiautogenousgrinding(SAG)millcircuitstypicallyfocusontheoperatingcostsavings
ofcrushingplantsversusthecapitalcostsavingsofSAGmillplants.Veryhighlevel(ordesktop)
studiesofthistypedonotincludeagreatdealofengineeringwork,meaningthatthedetails(and
costs) of ancillary systems are often not captured when doing the comparisons. The more
complicatednatureofthemultistagecrushingplantsmeanstherearemoreancillarysystemsthan
aSAGmillplant,riskingaskewedanalysisoftheoperatingcosts,benefitingtheHPGRoption.
Oneoftheancillarysystemsthatmaybeoverlookedisthemorecomplicatedconveyingsystems
inherentinmultistagecrushingplants.Duetothemassofmaterialbeinghandledinlargemining
operations,theseconveyorsbecomeasignificantportionoftheoperatingcosts.Asimplewayto
estimate the specific energy consumption of conveying systems without requiring engineering
drawingsandasitelayoutwouldallowhighlevelstudiestocapturethisoperatingcost.
Conventional conveyor power draw calculations are complicated. Examples are the Conveyor
EquipmentManufacturersAssociationmethod(CEMA,2007)orthemethodintheSMEMineral
ProcessingHandbook(Hays&VanSlyke,1985).Whatisneededforhighlevelstudiesisamuch
simplermethodthatgivesacloseenoughapproximationoftheconveyorsystempowerdemand.

METHODOLOGY

ExistingOperations
TwoexistingoperationsareconsideredtobereasonabletemplatesformostbasemetalHPGRplant
designs:CerroVerdeinPeru(CVinFigure1)andBoddingtoninAustralia(BinFigure2).The
crushingcircuitconfiguration atBoddington issuited toanorewithasmallquantity offines,
meaningalltheprimarycrushedoreisfeddirectlytothesecondarycrusher,theproductofwhich
isscreenedwithoversizepassingbacktothesecondarycrusher.TheoreatCerroVerdecontains
morefines,anditisadvantageoustoperformwholeorescreeningwithonlytheoversizepassingto
thesecondarycrusher,theproductofwhichispassedbacktothescreen.

TheHPGRandfineoreunitoperationsoftheBoddingtonandCerroVerdecircuitsarethesamefor
thepurposesofahighlevelstudy. Theliteraturereportsthatthefinescreencirculatingloadat
CerroVerdeis95%110%(Koskietal,2011)andthefinescreencirculatingloadatBoddingtonis
80%(Hartetal,2011).Theseareroundedoffto100%circulatingload(meaningtheHPGRfeedrate
isdoublethefeedratetotheoverallcircuit). Bothoperationsemploywetscreeningoffineore,
meaningthatthefinescreenoversizeconveyors(CV06&B06)arehandlingdampmaterial.This
conveyorwouldhavetobeenclosedandheatedincoldconditions(whicharenotconsideredin
thisstudy).

Someadditionalassumptionscanbeintroducedtosimplifythemassbalancingofthesecircuits.
Thesecondaryscreen(withroughly50mmopenings)willhaveacirculatingloadof66%(meaning
theoversizeflowis0.66timestherateoffreshfeedfromthecoarseorestockpile)regardlessof
whetherthesecondarycircuitisdirect(Boddington)orreverse(CerroVerde).

2
 Figure 1 Block flow diagram of the Cerro Verde circuit, indicating major flows

Figure 2 Block flow diagram of the Boddington circuit, indicating major flows

SimplifyingAssumptions

Onlylargetonnageplantsareconsidered,suchascopperporphyries(layoutsofsmall
plants, such as in the diamond industry, do not fit this assumption). The nominal
throughputisassumedtobe4500t/h,approximately100kt/d.

Only two types of multistage crushing and HPGR plants will be considered,
correspondingtoBoddington(orecontainingfewfines)andCerroVerde(orecontaining
significantfines).

Theprimarycrusherandcoarseorestockpileisneglected,asitisacommonfeatureof
HPGRandSAGmillbasedcircuits.

Thesiteisflat,andalltopsofthemajorstructures(bins,stockpiles,crusherfeedhoppers)
areatthesameelevation.Theoverallliftisassumedtobe30mforallconveyors.

Thesitelayoutiscompactandtheconveyorsarenotusedtotransportoreforsignificant
lateral distances beyond what is necessary to lift the ore into the next unit operation.
Conveyor1isassumedtobe350mlong,conveyors2,3and4areassumedtobe200m
long,andconveyors5and6areassumedtobe400mlong.

Thecirculatingloadofthecoarseorescreen(closesthesecondarycrusher)is66%.

3
 Thecirculatingloadofthefineorescreen(closestheHPGR)is100%.

All conveyor motors are fixedspeed motors with aggregate 90% motor and gearbox
efficiency.

ConveyorPowerDraw

Conveyor sizing calculations were run using a method similar to the SME Mineral Processing
handbook(Hays&VanSlyke,1985).ThisSMEmethodisabitoldandthemaximumbeltsizeitis
capableofusingissixtyinches(1.5m).Multiplebeltsareassumedtobeusedanytimethecapacity
ofastreamexceedsthecapabilityofthatmaximumbeltwidth.Becausetheseresultsareconverted
intoafactoredspecificenergyconsumptionestimate,thefactorsdonotchangewhenmultiplebelts
areused(thereisnooperatingcosteconomyofscalebenefittousingonebeltversustwo;any
suchbenefitswouldapplytothecapitalcost).Conventionalconveyorpowerdrawcangenerallybe
dividedintotwocategories:

Thepowerrequiredtolifttheconveyorloaduptoaspecifiedheight,and

everythingelse.

Theeverythingelsecategoryincludestherollingfrictionofthebelt,bothloadedandunloaded,
thebendingfrictionattheendandtakeuppulleys,thefrictionassociatedwithskirtingandbelt
cleaning,andanyothersourcesofpowerlossnotrelatedtothechangeinelevationoftheconveyor
load.Bymakingsomesimplifyingassumptionsaboutthesitelayout,allthesecanbelumpedintoa
factorappliedtothepowerrequiredtolifttheconveyorload.

The energy required to lift the conveyor load of mass m to a specified height h is simply the
potentialenergyofthatmassatthatheight,givenby

E=mhg (1)

where, E isthepotentialenergyinJoules, m isthemassinkg, h istheheightinm,and g isthe

gravitationalconstant,approximately9.807m/s.Metallurgiststypicallydon'tuseenergyinterms
ofJoules,preferringtheformpowertime:kilowatthours.DividingEasJoulesby3.60givesEas
kWh.

InsertingunitvaluesofliftingonetonnebyonemetreintoEquation1givesaunitenergyof2.724
kWht1m1.Specificenergyconsumption(SEC),askWh/t,isobtainedsimplybymultiplyingby
theheightthataconveyorlifts,thevalueh.

The results of the detailed conveyor sizing calculations compared to the specific energy
consumptionduetoliftingmaterialarepresentedinTable1.Allconveyormotorsweresizedusing
theSME method with an extra25% allowance for ancillary power draw,such as unaccounted
trippersandconveyorsintransferpoints. Thelayoutofconveyor4isverydifferentinthetwo
existingplants:Boddingtonhasasingleconveyordirectlyconnectingthecoarsescreeningareato
theHPGRbins,whereasCerroVerdehasthecoarsescreenoversizeconveyedupandthroughtwo

4
90transfers,thenupagaintodischargeintotheHPGRfeedbins(effectively,doublethedistance
andheightoftheequivalentstreamatBoddington).ThekeyoutputofTable1isthedetermination
ofwhatproportionofaconveyor'spowerisnotrelatedtothechangeinelevationoftheload.The
aggregatetotalsuggeststhatthetotalconveyorpower,asmeasuredatthemotorinput,is1.85times
thepowerrequiredtolifttheconveyorload(alternatively,liftingis55%oftheconveyorpower
draw,everythingelseis45%).

Table 1 Conveyor power draw, SEC versus SEC to lift conveyor load

Motor Power, Throughput, SEC of conveying, SEC of lifting, Factor,

input kW t/h kWh/t kWh/t total SEC:lift SEC
Conveyor Lift,
N m B CV B CV B CV B CV B CV

1 30 701 701 4500 4500 0.16 0.16 0.08 0.08 1.91 1.91
2 30 1069 427 7470 2970 0.14 0.14 0.08 0.08 1.75 1.76
3 30 426 426 2970 2970 0.14 0.14 0.08 0.08 1.75 1.75
4 30/60 636 1271 4500 4500 0.14 0.28 0.08 0.16 1.73 1.73
5 30 1480 1480 9000 9000 0.16 0.16 0.08 0.08 2.01 2.01
6 30 724 724 4500 4500 0.16 0.16 0.08 0.08 1.97 1.97

overall 0.49 0.57 1.85 1.85

based on 0.90 conversion motor input:output based on as-conveyed tonnes.

CompositeCirculatingLoad

Thedegreeofrecirculationwithinthecrushingcircuitwillaffectthepowerrequirement.Asimple
measureofhowmuchmaterialtheconveyoristransportingisneededtoaccountfortheenergy
consumedwhenrecirculatingmaterialwithinthecircuit.

Table 2 demonstrates a throughputindependent proportion of feed rate, effectively, the

circulating load+ 100%for each conveyor. Numerically, thisis thethroughputof aconveyor
dividedbythecircuitfreshfeedrate.Averagingtheproportionvalueofalltheconveyorswithin
thecircuityieldsacompositeproportionthatcanapplytoawholecircuit(thiscanbethoughtofas
thetypicalconveyorintheplantseesthiscompositethroughput).Itturnsoutthatthisvalueisa
property of the circuit flowsheet: a circuit similar to Boddington will have a typical conveyor
carryingafactorof122%timesthecircuitfeedrate,andtheequivalentvalueforCerroVerdeis
105%.

5
Table 2 Conveyor feed rates as proportion of circuit fresh feed

Throughput, t/h Proportion of fresh feed

Conveyor
N B CV B CV

1 4500 4500 100% 100%

2 7470 2970 166% 66%
3 2970 2970 66% 66%
4 4500 4500 100% 100%
5 9000 9000 200% 200%
6 4500 4500 100% 100%

overall 122% 105%

ConveyorSpecificEnergyConsumptionbyaSimplifiedFactoringMethod

The pieces of a simplified factoring method for crushing plant conveyor specific energy
consumptionarenowallinplace. Howhightheconveyorsmustlifttheirload,thespecific
energyconsumptionofthischangeinpotentialenergyfortheload,andtheoverallconveyorpower
factorversuspotentialenergyofliftingisgiveninTable1.Howmuchmaterialisconveyedis
measuredusingtheproportionoffeedisgiveninTable2.CombiningtheseelementsinTable3
givestheestimatedspecificenergyconsumptionofindividualconveyorsandforanentirecircuit.

TheSECofconveying,kWh/tonafreshfeedbasis,iscalculatedusingEquation2:

E = h 2.724 (Proportion of fresh feed / 100) 1.85 1000 (2)

AsignificantobservationisthatEquation2worksforindividualconveyors,andalsofortheoverall
plantifthecumulativeliftofallconveyorstagesisused.Thissimplifiesdesktopstudiesasitisnot
necessarytoaccountforeveryconveyor,onlythesumoftheliftingdonebyallconveyors.

6
Table 3 Conveyor SEC estimated by simplified factoring method

Conveyor Proportion of fresh Conveyor total SEC of conveying,

N Lift, m SEC of lifting feed power factor fresh feed basis

B CV B CV B CV B CV B CV

1 30 30 0.08 0.08 100% 100% 1.85 1.85 0.15 0.15

2 30 30 0.08 0.08 166% 66% 1.85 1.85 0.25 0.10
3 30 30 0.08 0.08 66% 66% 1.85 1.85 0.10 0.10
4 30 60 0.08 0.16 100% 100% 1.85 1.85 0.15 0.30
5 30 30 0.08 0.08 200% 200% 1.85 1.85 0.30 0.30
6 30 30 0.08 0.08 100% 100% 1.85 1.85 0.15 0.15

overall 180 210 0.49 0.57 122% 105% 1.85 1.85 1.11 1.11

Anotherinterestingobservationisthattheoverallconveyingspecificenergyconsumptionwon't
change whether a Boddington or Cerro Verde style circuit is chosen (both are 1.11 kWh/t).
Boddington has less lifting of ore than Cerro Verde (fewer transfer points), but a Boddington
conveyortypicallyagreaterloadthatcompletelycancelsoutthedifferenceinelevationchangeof
ore.

RESULTSANDDISCUSSION
The specific energy consumption of conveying a finespoor ore, such as Boddington, can be
estimatedbyEquation3,andthespecificenergyconsumptionofafinesrichore,suchasCerro
Verde,canbeestimatedbyEquation4:

E = h 1.85 (122% / 100) 2.724 1000 (3)

E = h 1.85 (105% / 100) 2.724 1000 (4)

where,histhesumoftheelevationchangeofallconveyorsdownstreamofthecoarseorestockpile
(theverticallyliftoftheore).Theconstantsare:aconveyorpowerdrawallowance1.85accounting
fornonelevationpowerdraw,acirculatingloadtermof122%or105%torepresenttheaverage
proportionoffreshorefeedratethatisliftedbyatypicalconveyor,andapotentialenergytermof
2.724kWhtoliftatonneoforebyaheightof1m.

Thehtermisthesumoftheelevationchange(verticallift).Boddingtonhasroughlysixconveyor
stageswithaliftof30mineachstage,foratotal h of180m. CerroVerdehasroughlyseven
conveyorstageswithaliftof30mineachstage,foratotalhof210m.Itissuggestedthatthe
stageheightof30misthereasonablemaximumthatshouldbeconsideredforadesktopstudy,and
additionalprocesstrainsof30mhighprocessblocksshouldbeconsideredratherthangoinghigher
insituationswheremorethroughputisdesired.

7
Plantsat least halfthesizeofBoddington andCerroVerdecan usethesamesizeofcrushing
machines(reducingthequantityoflargecrushersineachstageratherthaninstallingthesame
quantityofsmallercrushers),whichwillrequirethesame30mheight.CerroVerdehasfourlarge
secondarycrushers(MP1000)andfourlargeHPGRunits(POLYCOM24/17)totreat5000t/h,so
the30mheightcanbeusedforfeedratesdownto2500t/hbeforesmallerequipmentarerequired
forplantavailabilityreasons.Thefactorsinthismethodwillstillworkforsmallercrushingplants
thatusemoduleheightslessthan30maslongastheactualheightisaccumulatedandenteredinto
theappropriateequations.

Discussion,ComparisontoPublishedData

Publishedconveyorspecificenergyconsumptionvaluesare1.30kWh/tforBoddington(Parkeret
al.,2001)andbetween1.27kWh/t(Vanderbeeketal.,2006)and1.29kWh/t(Koskietal,2011)for
CerroVerde.

The simplified method (Equations 3 and 4) is compared to the published specific energy
consumptionvaluesinTable4.Boththeoperatingplantsmakeasignificantdeviationfromthelist
ofassumptionsusedtodevelopthesimplifiedmethod,soanadjustmentisrequired.Thefineore
binfeedconveyor,number5,liftssignificantlyhigherthantheassumed30m,actuallybeingabout
35matbothBoddingtonandCerroVerde.Thereasonswhythiswasdoneisbelievedtobesite
specificandisnotnecessarilyrequiredifoneassumesaflatlocation.

Afterdeductingtheextraconveyorliftattheoperatingplants,thesimplifiedmethodpredictions
arewithin12%ofthepublishedvalues.

Table 4 Simplified method compared to published plant data

Boddington Cerro Verde

Published conveyor SEC, kWh/t 1.30 1.28

Deduct extra height conveyor 5, kWh/t (-0.04) (-0.04)

Published SEC for comparison, kWh/t 1.26 1.24

SEC by simplified method, kWh/t 1.11 1.11

difference 0.15 0.13

12% 10%

Discussion,ComparisonstoSAGMillCircuits

Theintroductiontothispaperstatesthatthereasonforperformingthissortofcalculationisto
betterrepresentthetotalenergyusedinamultistagecrushingandHPGRplantforthepurposeof

8
comparingtoaSAGmillcircuit.Anumberoftopicsbeyondthescopeofthispaperareneededto
completesuchacomparison:

Dust collection is needed in screening plants and conveyor transfer points. Both
Boddington(Parkeretal.,2001)andCerroVerde(Koskietal,2011)claima0.5kWh/t
specificenergyconsumptionfordustcollection.

Fineorebinswillneedfreezeprotectionandbeltsconveyingdampcrushedorerequire
enclosureandheatinginverycoldclimates(MacLellan,1972).LocationsinthehighAndes
may require additional capital and operating costs to prevent very cold weather from
interferingwithmaterialshandling.

BallmillsformultistagecrushingandHPGR circuitswillbebiggerthanballmillsfor
closedcircuitSAGmillcircuits(Burchardt&Ojeda,2010).Therearetworeasonsforthis:
firstisthattheproductsizeoutofanHPGRcircuit(3mmto6mmaccordingtoBurchardt
etal.,2011)iscoarserthanthetypicalproductsizeofaSAGmillcircuit(mmto3mm
accordingtoMorrell,2011);andsecondisthattheSAGmillsmakemorefinesthanHPGR
circuits(Amelunxen,2013),evenafteraccountingformicrocracking(Dolletal.,2010).A
typical copper porphyry treated by HPGR will see a 0% to 10% reduction in ball mill
operatingworkindexversustheBondballmillworkindextest;atypicalcopperporphyry
treatedbyaSAGmillwillseeaphantomcycloneeffectthatreducestheoperatingwork
index5%to15%versusthelaboratory(Amelunxen,2013).

Thespecificenergyconsumptionoftheballmillcircuitancillarysystems,mostnotablythe
cyclonefeedpump,areprobablyverysimilarwhetherthefeedisderivedfromaSAGmill
oranHPGR.ThesecanbeneglectedforthepurposesofHPGR:SAGcomparisonstudies.

MostcircuitcomparisonsignoretheSAGmillpebblecrushingconveyorspecificenergy
consumption. Themethodpresentedinthispaperisnotsuitableforestimatingsucha
value because theconveyors in apebblecrushing circuit may not include asignificant
degreeoflifting. TheSAGfeedandpebblecrushingconveyorpowerisgenerallysmall,
near0.1kWh/tforasimplerecirculatingbeltsystemwithout apebblecrusher,or0.15
kWh/twithapebblecrusher.TheSAGmillfeedbeltwillconsumelessthan0.06kWh/t,
dependingonthebeltlengthandtheheightoreiselevatedto.

CONCLUSION
Specific energy consumption of conveyors in multistage crushing plants, E as kWh/t, can be
approximated using a simple potential energy model where one only considers how much
material is to be lifted how high. Multiplying the conveyor specific energy consumption
associatedwithliftingthechargebyafactorof1.85providesapredictionoftheconveyorspecific
energyconsumptionconsumedbyallprocesses,includingtheelevationchange.

Different circuit configurations can be approximated as two values so that the specific energy
consumptionisindependentoftheactualfeedrate:

9
 Proportionoffreshfeed:Theeffectiveoverallcirculatingloadofafinesrichore,suchas
CerroVerde,is105%;whereas,theoverallcirculating loadofafinespoorore,suchas
Boddington,is122%.

hCumulativeconveyorlift:Thecumulativeelevationchangeofalltheconveyorsinthe
process;180mforasitelayoutsimilartoBoddingtonor210mforasitelayoutsimilarto
CerroVerde.

E=h1.85(Proportionoffreshfeed/100)2.7241000

Usingtheseinputvalues,themethodpredictsaconveyorspecificenergywithin12%ofpublished
datafortwooperatingHPGRcircuits,BoddingtonandCerroVerde.Themethodissuitablefor
performing a desktoplevel scoping study where the detailed layouts and conveyor sizing
calculationsofafeasibilitystudyarenotavailable,butareasonableassumptionofthebinheight
canbemade.

ACKNOWLEDGEMENTS
This paper was originally published in the Proceedings of the 47 th Annual Canadian Mineral
ProcessorsOperatorsConference,January2015,Ottawa,Canada.

TheauthoracknowledgesthevaluableinsightofferedbyIanOrford(AMECmining)andStefan
Nadolski(BCMiningResearchInc.)duringtheirreviewsofthedraftofthispaper.

NOMENCLATURE
B Boddington(referringtosomeaspectofBoddingtonscircuit,suchasaspecificconveyor)
CV CerroVerde(referringtosomeaspectofCVscircuit,suchasaspecificconveyor)
E (alsoSEC)Specificenergyconsumption,kWh/t
h heightoflift,m

REFERENCES
Amelunxen,P.(2013).TheSAGGrindabilityIndexTest.ShortcoursepresentedatProcemin2013,Santiago,
Chile.
Burchart,E.&Ojeda,R.(2010).HPGRinhardrockapplicationsatechnologythatbrokebarriers. InW.
Kracht, R. Kuyvenhoven & G. MontesAtenas (Eds.), Seventh International Mineral Processing
Seminar,Procemin2010(pp.127139),Santiago,Chile.
Burchardt, E., Patzelt, N., Knecht, J. and Klymowsky, R. (2011). HPGR's in minerals: what do existing
operationstellusforthefuture?InK.Major,B.Flintoff,B.Klein&K.McLeod(Eds.),International
AutogenousGrinding,SemiautogenousGrindingandHighPressureGrindingRollTechnology2011
(Paper108),Vancouver,Canada.
CEMA(2007)BeltConveyorsforBulkMaterials(6thed.2ndprinting),ConveyorEquipmentManufacturers
Association.

10
Doll,A.G.,Barratt,D.J.&Godoy,R.(2010).Microcrackingversusthephantomcyclone;comparingSAGmills
andHPGRonaconsistentbasis.InW.Kracht,R.Kuyvenhoven&G.MontesAtenas(Eds.),Seventh
InternationalMineralProcessingSeminar,Procemin2010(pp.7382),Santiago,Chile.
Hart,S.,Parker,B.,Rees,T.,Manesh,A.,andMcGaffin,I.(2011).CommissioningandrampupoftheHPGR
circuit at Newmont Boddington Gold. In K. Major, B. Flintoff, B. Klein & K. McLeod (Eds.),
International Autogenous Grinding, Semiautogenous Grinding and High Pressure Grinding Roll
Technology2011(Paper041),Vancouver,Canada.
Hays, R.M. and Van Slyke, W.R. (1985). Dry Bulk Material Transport. In N.L. Weiss (Ed), SME Mineral
Processing Handbook (pp. 1032 1064). New York, USA: American Institute of Mining,
Metallurgical,andPetroleumEngineers,Inc.
Koski,S.,Vanderbeek,J.,Enriquez,J.(2011).CerroVerdeConcentratorFouryearsoperatingHPGRs.InK.
Major,B.Flintoff,B.Klein&K.McLeod(Eds.),InternationalAutogenousGrinding,Semiautogenous
GrindingandHighPressureGrindingRollTechnology2011(Paper140),Vancouver,Canada.
MacLellan,J.A.(1972).LayoutandServicesinColdClimates.MiningEngineering,January(pp.5860).
Morrell,S.(2011).TheappropriatenessofthetransfersizeinAGandSAGmillcircuitdesign.InK.Major,B.
Flintoff,B.Klein&K.McLeod(Eds.),InternationalAutogenousGrinding,SemiautogenousGrinding
andHighPressureGrindingRollTechnology2011(Paper041),Vancouver,Canada.
Parker,B.,Rowe,P.,Lane,G.andMorrell,S.(2001).Thedecisiontooptforhighpressuregrindingrollsforthe
Boddingtonexpansion.InD.Barratt,M.Allan,andA.Mular(Eds.),InternationalAutogenousand
SemiautogenousGrindingTechnology2001(pp.III93III106),Vancouver,Canada.
Vanderbeek,J.,Linde,T.,Brack,W.andMarsden,J.(2006).HPGRimplementationatCerroVerde. InM.
Allan, K. Major, B. Flintoff, B. Klein and A. Mular (Eds.), International Autogenous and
SemiautogenousGrindingTechnology2006(pp.IV45IV61),Vancouver,Canada.

11