-
IntroductionTheMcArthurRiveruraniummineislocatedin
theAthabascasandstonere-gioninthenorthernpartoftheprovinceof
Saskatchewan,Canada.It is
theworld'slargest,high-gradeuraniumde-positwithprovenandprobablereservesofmorethan473millionpoundsU30S.It
ismajorityownedandoperatedbytheCamecoCorporation.
Thispaperpresentsacasestudyoftheengineeringthatwascarriedoutinorderto
artificiallyfreezetheunder-groundorebodyattheminepriortoitbeingmined.Fulldetailsof
theproject
AthabascaSandstone
.~Highpressurewater
~
SandandGravel
--
Case Study: Thermal Analysis of ArtificialGround Freezing at the
McArthur RiverUranium Mine
G.P..Newman
havebeenpreviouslydescribedbyNewmanandMaishman(2000).Theprimaryfocusof
thepaperistheappli-cationofanadvancedthermalanalysistoolthatgreatlysimplifiesthefiniteele-mentmodelingrequiredduringthede-signand.monitoringstagesof
projectssuchas thatcarriedoutatMcArthurRiver.
BackgroundAccordingtoNewmanandMaishman(2000),theorebodyitselfislocated550metersto
620 metersundergroundwheretheground-waterpressureisap-
proximately5500kPa.Duetothepres-enceof
ahangingwallfaultstructure,theorebodyis surroundedon threesidesby
fairlydry,competentground.Theotherthreesidesarecomprisedofhighlyfracturedsandstonewithsignifi-cantamountsof
rubble,flowingsandandclayregions.In
ordertominetheore,itwasnecessarytocreateafrozenwallbarrieraroundthethreepoorsidesoftheorebody.Thefrozenwallbarrierwasdesignedtopermitdrainageofwa-terin
theoreandconsequentlyreducewaterpressurespriorto
mining.Thewallwasalsorequiredtoprovidestruc-
tural supportofweak,clay/ oreground near tomining
cavities.FigureI
showsacross-sectionoftheorebodyandneighboringgeol-ogy.
A mechanicalfreezingsystemiscomprisedof
abrinecoolinganddistributionnet-workplusaseriesof brine
freezepipesinstalledinthegroundto befrozen.
Typicalgroundfreezingapplicationshaveinvolveddrillingfreezeholesfromsurfaceor
near
.........SouthFreezeRow
DryBasementRock
640mLevefDeveloDment
60 GeotechnicalNews, June2003
Figure1.Cross-sectionoftheorebodyandneighboringgeology(afterNewmanandMaishman,2000)
-
~-.....
/1
Low Pressure Side
Brine
72 FreezePipes
~reeze Plant-Conceptualfreezing system
High Pressure Side
Brine
530m L.evel
Figure2. Illustrationof
thehighandlowpressurebrinedistributionnetworks(afterNewmanandMaishman,2000)
surfaceandtheseactivitieshavebeenwelldocumented.Inaddition,thebrinecoolinganddistributionnetworkhastypicallyincludedanammoniacom-pressorwith
ammoniato brineheatexchangers.Theprocessof
installinga"typical"freezingsystematMcArthurRiverwasmademoredifficultduetothelocationof
thefreezepipechamberunderground.
Thefreezingchamberislocated530metersbelowground,whichmeansthatthebrinepressureswithinthefreezepipesandassociatedbrinedistributionnetworkwouldequal5000kPaif
con-necteddirectlytothesurfacerefrigera-tionplant.Thisis
notpracticalfromadesignoroperationsperspective.In or-derto
minimizethebrinefluidpres-sures, the underground
brinedistributionsystemwasisolatedfromthesurfacebrinesystemusingshellandtubebrine-brineheatexchangers.
TheillustrationinFigure2showstherelativepositionofthe800Tonrefriger-ationcapacityfreezeplantonsurface,the12"ID
brinesupplyandreturnlinesinstalledin theshaft,andoneof
fourshellandtubeheatexchangersonthe530m level.The
low-pressurebrinenetworkon the530m leveloperateswithina
150kPato600kPapressurerangeatflowratesrangingbetween130
m3/hrand550m3/hr.Thedesignbrinetemperaturewas-40degreesCelsius.
In orderto determinetheactualgrowthof
thefreezewallitwasneces-saryto
installthermocouplestringsatseverallocationsaroundthefreezingre-gion.Thethermocoupleswereloweredintoacasedholecontainingafinegroutmix
priorto
thegroutsetting.Eachstringwascomprisedoftwelvesensorslocatedatfive-meterintervalsdownanygiventemperaturemonitoringhole.Thisenabledthetemperaturedecaytobe
monitoredoffsetfromthefreezepipesin varioustypesof ground(seeFigure
I for comparisonof
groundtypes).Groundtemperatureswerere-cordedeverysecondday.In
thiscasestudy,theactualgroundtemperaturescanbeusedtoverifythenewthermalanalysis
numerical tool. This isdescribedbelow.
Thermal Modelingand AnalysisTheTEMP/w
two-dimensionalfiniteelement computer
program(GEO-SLOPE,2002)hasbeenusedex-tensivelyby
manypracticinggroundfreezingconsultantsfordesignofartifi-cialgroundfreezingprojects.Inthepastit
wascommonto
applythethermalboundaryconditionforafreezepipebyassuminga
fixedtemperaturedecayfunctionforearlystagesoffreezingfol-
GEOSPEC
lowedby anassumptionthatthepipesurfaceis
ascold,ornearlyascoldasthebrinefortheremainderofthefreez-ingperiod.
Recently,TEMP/w hasbeenmodi-fiedsothattheuseris
abletoapplyaconvectiveheattransferanalysisforthepipe-groundinterface(oranyothersur-faceconvectiveheattransferprocess).Theamountof
convectiveheattrans-ferredbetweenthe
groundandthechilledbrineisdependentonthegroundtemperaturerelativeto
thebrinetem-peratureanditsubsequentlydetermineswhatthenewgroundtemperaturewillbe.Basedonthisapproach,thegroundwillcoolatavariablerateandtoamini-mumvaluethatis
determinedby thebrineflowparametersandthediffer-ence betweenbrine
and groundtemperature.
Convectiveheattransferis com-prisedof two
mechanisms:energytransferduetorandommolecularmo-tion(diffusion)andenergytransferduetothemotionofafluid(advection/con-vection).Inthiscasestudyanalysis,weareconcernedwith
convectionheattransferbetweenfluid in motionin
apipeandthepipewallwhentheyareatdifferenttemperatures.Technically,theconvectiveheattransferisoccurringbe-tweentheinternalpipewall
andthefluidbutitisacceptabletocombinetheconductiveheattransferacrossthesteelpipewallwiththeconvectivecompo-nenttoarriveatacombinedconvectiveheattransfercoefficient.
Regardlessof thenatureof
thecon-vectiveheattransferprocess,theappro-priaterateequationis
q=h(Ts- Tf)whereqistheunitheatflux (W 1m2);h
isthecombinedconvectiveheattransfercoefficient(W/m20C);Tsis
thepipe'sexternalsurfacetemperature(0C);andTf is the fluid
temperature.In theTEMP/w
program,theusermustinputtheoverallheattransfercoefficientaswellasafixedtemperatureortimede-pendenttemperaturefunctionfor
thefluid.Thephysicalsizeof thepipecanalsobespecifiedasanoptionif
theac-tualpipegeometryisomittedfromthefiniteelementmesh.
ThedatainTableI summarizesthethermalpropertiesusedin
theoriginal
GeotechnicalNews. June2003 61
-
GEOSPEC
Temperature Comparison for Different Ground Types
Measured at 2m offset from freeze row In high grade ore/clay-
Predicted Measured at 3.7m offset from freeze row in low grade
sandy/clay- PredictedJj. Measured at 3.5m offset from freeze row in
barren sandstone
4 6 8 10 12
Time (Months)
Figure 3. Comparison of computedand predicted ground
temperaturesfor various rock types(measureddata after Newmanand
Maishman, 2000)
thermalmodelling (NewmanandMaishman,2000)andin
thecurrentanalysis.
It shouldbenotedherethatnoad-
Theseresultsclearlyshowthatthenewconvectiveheattransferboundarycon-ditionoptioncanbeappliedtoartificialgroundfreezingif
oneknowsthebrine
justmentsto
thefreezepipeboundaryconditiontemperaturesweremadeinthemodelwhencalibratingthethermalconductivityof
theclay/
oreground.Theactualfreezepipetemperaturewascomputedasaresultof
theconvectiveheatbeingremovedbythebrine.
Figure3 is a comparisonof
com-putedandmeasuredgroundtempera-turesfor threeof
themainrocktypesandinitialin-situgroundtemperatures.
62 GeotechnicalNews, June 2003
fluidtemperatureandfreezepipediam-eter.Theoverallconvectiveheattrans-fercoefficientcanreadilybecomputedbasedonthebrineflowrateandotherhydraulicflowproperties.
Conclusions
Artificialgroundfreezingof
thehighgradeMcArthurRiveruraniumorede-positwascarriedoutunderextremeconditions.Duetoadvancesinthermal
finiteelementanalysis,it
isnowpossi-bletoaccuratelydetermineheatlossesfromthesoilintothemechanicalfreez-ingsystemwithouthavingtomakeanassumptionabouttheappropriatetem-peratureboundaryconditiontouseintheanalysis.
ReferencesGeo-Slope,2002.TEMP/w Version5
UsersManual,Geo-SlopeInterna-tionalLtd.,Calgary,Canada.
Newman,GregandDerekMaishman.ArtificialGroundFreezingof
theMcArthurRiverUraniumOreDe-posit.Proceedings:InternationalConferenceonGroundFreezingandFrostActioninSoils.Belgium.Sep-tember,2000.
G.P.Newman,P. Eng.,GEO-SLOPEInternationalLtd.,1400,633-
6th.AvenueS\V,Calgary,AlbertaT2P 2Y5,Tel:403-269-2002,Fax:
403-266-4851,http://www.geo-slope.com
25.0
20.0
15.0-0-; 10.0...::s1U 5.0...Q)CoE 0.0Q)I-
-5.0
-10.0
-15.0
0 2
Table1.ThermalPropertiesUsed in Analyses
Conductivity(W/ m0c) BulkVolumetricWaterContent(%)
SilicifiedSandstone 5.2 10
Clav/ Ore 1.2 50
BasementOuartzite 5.6 2
-
IL
BiTechispleasedtopublisht~:nint: annuallistingof ]PhD Thesesin
engineering.- --- - - - --AnkaraUniversityDepartmentof
GeologicalEngineering,06100,Tandogan/ Ankara-Turkey
AyhanKocbayInvestigationof CharacteristicsandAlterationDegreeof
BasaltsOut-croppedAround Osmancik-