3 CAPILLARY PRESSURE
Reservoirrocktypicallycontainstheimmisciblephases:oil,water,andgas.The
forcesthatholdthesefluidsinequilibriumwitheachotherandwiththerockare
expressionsofcapillaryforces.Duringwaterflooding,theseforcesmayact
togetherwithfrictionalforcestoresisttheflowofoil.Itisthereforeadvantageous
togainanunderstandingofthenatureofthesecapillaryforces.
Definition:Capillarypressureisthepressuredifferenceexistingacrossthe
interfaceseparatingtwoimmisciblefluids.
Ifthewettabilityofthesystemisknown,thenthecapillarypressurewillalways
bepositiveifitisdefinedasthedifferencebetweenthepressuresinthenon
wettingandwettingphases.Thatis:
Pc Pnw Pw
Thusforanoilwatersystem(waterwet)
Pc Po Pw
Foragasoilsystem(oilwet)
Pc Pg Pw
What Causes Capillary Pressure?
Capillarypressureisasaresultoftheinterfacialtensionexistingattheinterface
separatingtwoimmisciblefluids.Theinterfacialtensionitselfiscausedbythe
imbalanceinthemolecularforcesofattractionexperiencedbythemoleculesat
thesurfaceasshownbelow.
Formoleculesintheinterior:
Netforces=0sincethereareenoughmolecules
aroundtobalanceout.
Formoleculesonthesurface:
Netresultofforcesisapulltowardtheinterior
causingatangentialtensiononthesurface.
Theneteffectoftheinterfacialtensionistotrytominimizetheinterfacialareain
amanneranalogoustothetensioninastretchedmembrane.Tobalancethese
forcesandtokeeptheinterfaceinequilibrium,thepressureinsidetheinterface
needstobehigherthanthatontheoutside.
Forcesreducinginterfacearedueto:
a)Interfacialtension
b)Externalpressure
Theeffectofinterfacialtensionistocompressthenonwettingphaserelativeto
thewettingphase.Theforcecreatedbytheinternalpressureisbalancingit.
3.1
EXPRESSIONS FOR CAPILLARY PRESSURE UNDER STATIC
CONDITIONS
3.1.1 Pc In terms of radius of capillary tube
Sincetheinterfaceisinequilibrium,forcecanbebalancedonanysegment.The
interfacialforcesareeliminatedbytakingasafreebody,thatpartoftheinterface
notindirectcontactwiththesolid.Aforcebalancewouldgive:
(InternalpressureExternalpressure)*Crosssectionalarea=Interfacialtension
*Circumference
aw
Pnw
air
water
Pw
2
2
Thus, Pnw r Cos 2r Pw r
Pnw Pw r2 Cos 2r
Therefore,
Andsincebydefinition, Pc Pnw Pw ,wehave:
Pc
2Cos
r
Pc
Foranairwatersystem,theairisthenonwettingphaseand
ThisequationisreferredtoasLaplace'sEquationinsometexts.
2 awCos aw
r
3.1.2 Pc In terms of height of fluid column.
Air - water system
air
Pa1
air
Pw1
hw
Pw2
water
(1)
air
Pa2
(2)
Pw2
Pa2 Pw 2 becausethereisnocapillarypressureacrossahorizontalinterface.
Pa1 Pa2 Pw 2
but Pw2 Pw1 h wg
therefore, Pa1 Pw1 h wg
Since Pc Pa1 Pw1 then Pc hw g
Oilwatersystem
oil
oil
Po1
oil
(1)
Pw1
hw
Po2
Pw2
Pw2
water
(2)
Pa2 Pw 2 becausenocapillaryexistsacrossanyinterfacethatishorizontal.
Pw2 Pw1 h wg
Po2 Po1 h o g
Since Pw2 Po2 ,then, Pw1 h wg Po1 ho g
Therefore, Po1 Pw1 w o hg
Thatis, Pc w o hg
CgsUnits:
Pc dynes cm 2
gm cc
g cm sec2
h cm
FieldUnits:
h w o
h
Pc w orPc
psi(or Ib sqin)
144
144
w , o Ib ft 3
h ft
Thetwoexpressionsforcapillarypressureinatube,oneintermsofheightofa
fluidcolumnandtheotherintermsoftheradiusofthecapillarytubecanbe
combinedtogiveanexpressionfortheheightofafluidcolumnintermsofthe
radiusofthetubeasfollows:
h g
2 cos
r
Thereforeforanairwatersystem,
hw
Similarly,foranoilwatersystem,
2 awCos
rwg
hw
2 owCos
r w o g
Thesetwoequationsshowaninverserelationshipbetweenfluidheightand
capillaryradius.Thesmallertheradiusis,thehighertheheightofthefluid
columnwillbe.
3.2
APPLICATIONS OF CAPILLARY PRESSURE EXPRESSIONS IN
POROUS MEDIA
3.2.1 Application to obtain static fluid distribution in porous media
3.2.2 Porous media modelled as a bundle of capillaries
Oneoftheearliestandsimplestdepictionofporousmediawasasabundleof
capillarytubesofarbitrarilyvaryingdiameters.Byapplyingtheapplicableoneof
theequations:
hw
2 Cos ow
2 awCos aw
hw ow
r w o g
r w g
or
Thedifferentwaterheightsinsuchasystemisillustratedinthefigurebelow
whereifthenumberoftubeswerenumerous,asmoothcurvewillresultasshown
inthelowerfigure.Thatfigureisforathreephasegasoilwatersystem.The
figuresalsoshowthedifferencebetweenthewateroilcontact(WOC)andthe
freewatertable.TheWOCisthedepthatwhich Sw 100% begins(moving
downward)whilethefreewatertableisthedepthatwhich Pc 0 .
Gas
Gas
GOC
GOC
Oil
Depth
Oil
Water
WOC
WOC
Water
Water
Free water table
FWT
Water saturation, %
Sg
Depth
So
Gas cap
So Sw
Liquid, %
Sw
Water, %
0.0 Swi
GOC
Oil zone
WOC
Free - water Table
Saturation, %
100
Water
Crossection of reservoir
3.2.3 Porous media modelled as a packaging of uniform spheres
Anevenmorerealisticmodelisthedepictionofporousmediaasapackagingof
uniformspheres.Applyingthetwoexpressionsforcapillarypressureintermsof
theradiioftheinterfaceandintermsoftheheightoffluidcolumn,wehavefor
thissystem:
Sand grain
Sand grain
R1
R2
Non - wetting fluid
Wetting fluid
Fromwhich,
1
1 h w o
Pc
144
R1 R2
1
1
144
R1 R2
h
w o
3
Infieldunits, h ft, Ib in , Ib ft , R1 ft,R2 ft
Unfortunately, R1 andareimpossibletomeasureinporousmediaandsoare
usuallydeterminedempiricallyfromothermeasurementsintheporousmedia.For
thisreason,itismoreconvenienttoexplicitlymeasurecapillarypressureanduse
theequationbelowtocalculatetheheightofthefluidcolumn.
h
144Pc
w o
Example3.3
UsingthedrainagecapillarypressurecurveoftheVenangoCore(shownbelow).
Howmanyfeetabovethefreewatertableisthewater/oilcontact?(1ft=30.48
cm)
w 1 gm cm 3 , o 0.75 gm cm 3 ,1 cm of mercury = 13,322.2 dynes cm
Solution
Fromthefigure,thecapillarypressureatthewateroilcontactcanbereadas4
cm.
Since Pc hw w o g 4 *13,322 dynes / cm
Then,
hw
Pc
413,222.2
217.5 cm
w o g 1 0.75980
217.5
7.1 ft
= 30.48
3.3
Laboratory methods of measuring capillary pressure
Threegenerallyacceptedmethodsofmeasuringcapillarypressureinthe
laboratoryare:
a)ThePorousDiaphragm(orrestoredstate)Method
b)TheCentrifugalMethod
c)TheMercuryInjectionMethod
Allthreetestsareconductedoncoreplugscutfromreservoirwholecoresamples.
Drillingfluids,coringfluids,coringprocedure,corehandlingandtransportation,
storageandexperimentalprocessescanalterthenaturalstateofthecore.
Therefore,specialprecautionsarenecessarytoavoidalteringthenaturalstateof
thecore.Ifthenaturalstateofsaturationofthecorehadbeenaltered,thenitmust
berestoredtoitsnaturalstatebeforeconductinganycapillarypressuretests.
FreshCore:
Samplesfromcoretakenwitheitherwateroroilbasemudsthatarepreserved
(withinvadedfluids)andsubsequentlytestedwithoutcleaninganddryingare
referredtoasfreshcores.
NativeStateCores:
Samplesfromcorerecoveredwithleasecrudeorspecialoilbasefluidsknownto
haveminimalinfluenceoncorewettability,andthataretestedasfreshsamples,
arereferredtoasNativeState.Thesecoresareintheirnativestate(i.e.without
invadedfluids).Suchcorescomingfromabovethetransitionzoneshouldhave
thesamequantityanddistributionofwaterasinthereservoir.Thesesamplesare
preferredforwaterdisplacementtests.
RestoredCores:
Core samples cleaned and dried prior to testing are referred to as restored cores.
An advantage is that air permeability and porosity are available to assist in sample
selection. A disadvantage is that core wettability and spatial distribution of pore
water may not match that in the reservoir.
Thefollowingprecautionscanbehelpfulinobtainingrepresentativecoresifthe
drillingconditionspermit.
1.Useoilbasedrillingmudtominimizeclayswelling
2.Usenonoxidizedleasecrudeasacoringfluid.
3.Suitablestorageproceduresincludesubmersionunderdegassedwater,and
preservationwithsaranfoil,andwax.
Refinedoilversuscrudeoil
Refinedoilsaresuitableformosttests,andarepreferredwhentestsareatambient
conditions.
Crudeoilstobeusedinambienttestsshouldbesampledfromnonwater
producingwellsupstreamfromchemicalorheatertreaters.
Crudeoilsoftenprecipitateparaffinorasphaltenesatambientconditions,
resultingininvalidtestdata.
Reservoirconditiontestutilizinglivecrudeoilatreservoirpressuresand
temperaturesoftenovercomedifficultiesexperiencedwithcrudeatambient
conditions.
Reservoirfluidsamplesforspecialcoretestsmayberecoveredusingbottomhole
samplingtechniques,orrecombinedfromseparatorgasandoilsamples.
3.3.1 Centrifugal Method
1.
Rotateatafixedconstantspeed.Thecentrifugalforcedisplacessome
liquid,which canbereadatthewindowusingastrobelight.Thus,thesaturation
canbe obtained.
2.
Thespeedofrotationisconvertedtocapillarypressureusingappropriate
equations.
3.
Repeatforseveralspeedsandplotcapillarypressurewithsaturation.
3.3.2 Mercury Injection Method
1.
Placecoresampleinachamberandevacuateit.
2.
Forcemercuryinunderpressure.Theamountofmercuryinjecteddivided
bythe porevolumeisthenonwettingphasesaturation.Thecapillarypressureis
the
injectionpressure.
3.
Continueforseveralpressuresandplotthepressureagainstthemercury
saturation.
Advantages:
1.Fast(minutes)
2.Nothresholdpressurelimitation
Disadvantage: 1.Canonlybeusedforshapedcores.
3.3.3 Porous Diaphragm Method
regulator
air supply
P
air
2
core
water
displacing fluid (air)
tissue paper,
pulverised talc,
galena, flour to
maintain capillary
contact
porous diaphragm (filtered glass disc,
cellophane, porcelain) saturated with
displaced fluid
1.
Saturateboththecoresampleandthediaphragmwiththefluidtobe
displaced.
2.
Placethecoreintheapparatusasshown
3.
Applyalevelofpressure,waitforthecoretoreachstaticequilibrium.
Thecapillarypressure=heightofliquidcolumn+appliedpressure
Saturation =
Pore volume - Volume produced
Pore volume
Equilibrium
Production
4.
5.
Time
Increasethepressureandrepeatstep(3)
Plotcapillarypressureversussaturation
Pc
PLc
Sw
Disadvantages:
diaphragm
1.Havetoworkwithinthresholdpressureofthe
2.Takestoolongtoreachtheequilibrium,thereforea
complete
curvetakesfrom1040days
Mercuryinjectiontechniquewasdevelopedtoreducethistime.
3.4
Other methods
Dynamicmethod:
Core
gas
oil
1.
Simultaneoussteadyflowoftwofluidsisestablishedinthecore
2.
Usingspecialwelteddiscs,thepressureofthetwofluidsinthecoreis
measured.
Thedifference=Capillarypressure
3.
Changetherateofonefluidandthesaturationchanges
4.
Plot Pc versussaturation.
3.4.1 Field Method:
Alongcolumnofporousmediumputincontactwithawettingfluidatits
baseandsuspendedintheearthsgravitationalfield.Itislefttoreachequilibrium.
Samplesaretakenatdifferentheightsandthecapillarypressurecalculatedusing
Pc hg
Disadvantage: Maytakeverylongtoreachequilibrium
3.4.2 Capillary Hysteresis
3.4.3 Explanations for capillary hysteresis
1.
Theadvancingandrecedingcontactanglesaredifferent.Ifthecontact
angleduringimbibitionistheadvancingcontactangle,itdiffersfromthecontact
angledrainage(receding).Thismayexplainthephenomenonofhysteresis.
Pw1 Po1
A
advancing contact angle
Oil
Po2 Pw2
R
Receding contact angle
2."Inkbottleeffect
Forporousmediamodelledasabundleoftubeswithvaryingdiameters,agiven
capillarypressureexhibitsahigherfluidsaturationonthedrainagecurvethanon
theimbibitioncurve.
3.4.4 The effect of pore size distribution on capillary pressure curve
Themoreuniformtheporesizes,theflatterthetransitionzoneofthecapillary
pressurecurve.
3.4.5 Conversion of Laboratory Capillary Data to Reservoir Capillary
Data
Water(brine)oilcapillarypressuredataaredifficulttomeasureinthe
laboratory.Generally,airbrineorairmercurydataaremeasuredinsteadandit
becomesnecessarytoconvertthesedatatoequivalentoilwaterdata
P
representativeofreservoirfluids.Ifwedenote Pc aw or c a, Hg as Pc lab ,and
Pc ow as Pc res theconversionequationcanbederivedasfollows:
From
From
Pc lab 2 awCos aw
r
Pc res
r
,weobtain
2 awCos aw
Pc lab
2 ow Cos ow
2 owCos ow
r
Pc res
r
,weobtain
Assumingthatthesameporousmediumappliesinbothlaboratoryandfield,we
equatethe r' s toobtain,
Pc res 2 owCos ow Pc lab
2 awCos aw
ignoringthecontactangles,
Pc res res Pc lab
lab
Anidenticalequationwouldbeobtainedbystartingfromthetwoequations:
1
1
Pc lab
R1 R2 lab
1
1
Pc res
R1 R2 res
Assumingtheradiiofcurvatureinthelaboratoryisthesameasthatinthe
reservoir,theRHS'scanbeequatedand Pc res :
Pc res
res
P
lab c lab
3.4.6 Calculating Average water saturation
If a reservoir average capillary pressure curve (or even a laboratory curve) is
available, it can be converted to a height versus water saturation curve and used to
calculate the average water saturation for any desired interval. One simply needs
to put a new scale for the height on the y-axis of the Pc graph. The average water
saturation between any two height intervals can be evaluated as the area enclosed
between them divided by the distance between the height intervals. An example
illustrates the procedure.
Example3.4
Forapayzonewhosetopandbottomare45ftand25ftfromthefreewatertable,
usethelaboratoryPcgraphbelowtocalculatetheaveragewatersaturationforthis
payinterval.
a / w lab 70 dynes/cm, o / w res 35 dynes/cm, w 62.4 lb/cu ft, o 48 lb/cu ft
Solution to Example 3.4
First, convert the Pc lab to Pc res:
Pc res
res Pc lab (35)Pc lab
0.5Pc lab
lab
(70)
Next, convert Pc res to height above the free water table and plot on the right axis by
putting
a new scale on the RHS for h. Its scale is 5.0 the scale for Pc lab. This is
shown below.
144Pc res
144(0.5)Pc lab
5 Pc lab
( w o )
(62.4 48)
Mark off the top (45 ft) and bottom (25 ft) on the h-axis.
Shaded area
Sw
(45 - 25)
1
(0.31 0.25)(45 25)
2
Since the area in this case can be approximated by a single trapezoid, the shaded
area
Therefore,
Sw
0.5 0.31 0.25 45 25
45 25
= 0.28
The area under more complex shaded areas can be calculated after sub-division
into a number of trapezoids and applying the trapezoidal rule.
3.5
Averaging Capillary Pressure Curves
Considerareservoircrosssectionfromwhichfourcoresamplesaretakenat
differentdepthasshownbelow.Eachcorewillgenerateitsowncomplete
capillarypressurecurveinthelaboratorywhichcanbeconvertedtoareservoir
capillarypressurecurve.Thusfourdifferentlaboratorycapillarypressurecurves
areobtainedasshownbelow.Thequestionthenarises:
Howdowegetasingle Pc or height vs Sw curvetorepresentthereservoir?
TheansweristousetheLeverettJfunction
(1)
(2)
(3)
(4)
3.5.1 Leverett J-function
TheLeverettJfunctionisdefinedas:
Pc K
J
Cos
1
2
K =permeability
=interfacialtension
=contactangle
=porosity
TheJfunctionhastheeffectofnormalizingallcurvestoapproachasinglecurve
andisbasedontheassumptionthattheporousmediumcanbemodelledasa
bundleofnonconnectingcapillaries(Sliderpp279280).Obviouslythemore
capillarybundleassumptiondeviatesfromreality,thelesseffectivetheJfunction
correlationbecomes.Thiscorrelationisnotunique,butseemstoworkbetter
whentherocksareclassedastorocktypes,eg;limestone,dolomite,etc.
Givenseveralcapillarypressurecurves,withcorrespondingvaluesof
permeability K andporosity ,theprocedureforobtainingJfunctioncurveis
where,
asfollows:
a)
Pc .
b)
c)
Pickseveralvaluesof Sw from0to1andreadthecorrespondingvaluesof
Therewillbeasmany Pc valuesastherearecurves.
Foreach Pc value,calculateJandplotversus Sw .Repeatforall Sw values.
Putyourbestcorrelationcurvethroughthedata.
ThisJCurveisnowamastercurvethatcanbeusedtorepresentthatreservoirand
intheabsenceofotherdatacanbeusedforotherreservoirsofsimilarrocktype.
Thegraphsbelow,takenfrom(Amyxet.al.)showstheJfunctioncurveforthe
Edwardsformationshowingclassificationastorocktypes.
Fig.JfunctioncorrelationofcapillarypressuredataintheEdwardsformation,
JourdantonField.Jcurvefor(a)allcores;(b)limestonecores;(c)dolomitecores;
(d)microgranularlimestonecores;(e)coarsegrainedlimestonecores.(Source
Amyxet.al.)
3.5.2 How to use the Leverett J-function to calculate Average Water
Saturation
Valuesforaverageinitialorconnatewatersaturationarerequiredinmany
petroleumengineeringcalculations.Examplesare:(a)averagewatersaturationin
asectionofreservoirinordertofixeffectivefluidpermeabilities, k g ,k o ,k w, and
(b)averagewatersaturationinthewholereservoirinordertofixtheinitial
7758Ah 1 Swi
N
Boi
hydrocarbonvolumeinplace,
Undercapillaryequilibriumconditions,thewatersaturationofaparticularpiece,
orsampleofrocknotdependsonseveralfactors.Ithasbeenshownwithcertain
limitations,thataproperlydeterminedLeverettJfunctionversuswatersaturation
curvecanbeusedtoobtainanaveragewatersaturationfromanumberof
capillarypressurecurves.ItisassumedthataLeverettJfunctioncurveis
availableandappliestothereservoir.Theobjectivehereistoshowhowtousethe
Jfunctiontoobtainthebestpossibleestimateofaveragesaturation.Recallthat
theJfunctionisdefinedas:
k
Pc
J
Cos
P
Byexpressingthe c termintermsofheightandfluiddensitiestheequivalent
equationis:
k
h w o
J
144Cos
Itisimportanttonotewhileapplyingthisequationthatitsunitsarenotimportant.
Mixedunitscanbeusedwithoutappropriateconversionfactors.Itisonly
importanttobesuretousethesameunitsthatwentintodeterminingthevaluesof
Jmakinguptheoriginalplot.Inotherwords,findoutwhatunitswereusedto
calculatetheJfunctioncurveandstayconsistentwiththoseunitswhetherthey
aremixedornot.
NotealsothatJ=constant*h.Therefore,theshapeofaJfunctionversus Sw curve
wouldbesimilartothatofaheightversus Sw curve.Thedifferenceisa
displacementbyafactorequaltotheconstant.Thus,a Pc curvecanbeconverted
toaheightcurvesimplybyaddinganewyaxishavingitsabscissaequaltothe
constant* Pc .
3.5.2.1
Case 1: Permeability, Porosity, and Elevation are known for
each sample
K1 , 1
1
2
K2 , 2
K3 , 3
K4 , 4
3
h2
4
h4
Pc 0
h1
h3
Datum
Thisfigureillustratesfourreservoirsampleshavingdifferentvaluesof
permeabilityandporosityandlocatedatdifferentheightsabovea Pc 0 datum.
Assumingfluidpropertiesarethesameinallpieces,theJfunctionequationcan
besimplifiedto:
k
o
J ch
where c is the constant w
144Cos
3.5.2.2
The correct method
Thecorrectmethodofobtainingtheaveragesaturation, Swi forthefourpiecesis
tocalculateJforeachpiece,determinethecorrespondingwatersaturation, Swi of
eachpiecebyusingtheJcurveandthentakingthearithmeticaverageofthe
saturationswiththeequation:
S wi
jN
1
Swi
N
j 1
j
Notethatthisprocedurecorrectlytakesintoaccounttheverticalpositionofthe
piecesandtheircorrespondingpermeabilityandporosity.
Less correct methods
Thesemethodsfirstcalculateaveragevaluesof k, , and h ,substitutetheminto
theJequationtogetanaverageJvalue,andthenreadtheaveragewatersaturation
S wi fromtheJfunctionversus Sw graph.Theonlyadvantageofthesemethodsis
thattheamountofcalculationsisreduced.Theresulting S wi willalwayshave
errorinit.Howmucherrordependsonthespecificconditionbeingcalculated.
Thefigurebelowillustratestheconceptbehindusingaveragevaluesinorderto
obtainanaverageJvalue.
k
1
2
3
h1
h2
h3
h4
Pc 0
Therearetwoways:
K
Method(a): Calculate foreachsampleandobtainthearithmeticaveragefor
K
allfour.Also,obtainthearithmeticaverage h .Itisassumedthattheaverage
islocatedattheaverageheight h .TheaverageJfunctionequationthenbecomes:
J ch
where,
k
where c is a constant
1 N
h h j
N j 1
k 1
N
k j
j
j 1
N
Thisistheeasiestoftheaveragingmethodstodo.
Method(b):Thegeometricaveragepermeabilityandporosityareusedtogetthe
averageJfunction:
K
J ch G
G
K G =geometricmeanpermeability=
1 N
Antilog log K j N K1 K2 ....KN
N j 1
1 N
Antilog log j N 1 2 ..... N
N j 1
G =geometricmeanporosity=
Zerovaluesof K j and j arenotpermittedwhenevaluatingthegeometric
averages.Becauseporosityvaluesusuallyshowverylimitedrange,thegeometric
averageporosity, G ,canbereplacedbytheeasiertocalculatearithmetic
average, A ,withlittlelossofaccuracy.Therefore,theformusedbymost
engineersis
J ch
where,
3.5.2.3
KG
A
1 N
A
j
N
j1
isthearithmeticaverage
Errorsduetousingaveragevaluesof K and
Standing(4)discussestheamountoferrorin S w introducedbyusingaverage
valuesof k, , and h andstatesthattheerrordependsonseveralfactors.
Onefactoristhedistributionof k' s inaverticalsense.Ifthe k' s aredistributed
randomly,noerrorwillbeinvolved.Ontheotherhand,ifhighpermeabilities
predominateinoneportionofthesectionandlowpermeabilitiesinanother,some
errorwillbeintroduced.
Asecondfactoristheshapeofthe J vs. Sw curve.WherelogJislinearto Sw ,no
errorwillresultfromgeometricaverage k .WhereJislinearwith Sw ,someerror
willresult.
Athirdfactoristherangeofpermeabilityvalues.Littleerrorisintroducedwhen
therangeissmallandmoreerrorisintroducedwhentherangeislarge.Thebest
waytominimizeerrorsofaveragingisnottoaverage.Usethecorrectmethod.
3.5.2.4
Case 2: Permeability and porosity are unknown as functions
of elevation. Distance from Pc 0 (distance from free-water
table) is known
Thepetroleumengineeroftenneedstodevelopavalueforaveragewater
saturationbutdoesnothavedetailedinformationonpermeabilityandporosityas
afunctionofelevation.(Manywellsarenotcoreanalyzed).However,hemay
knowfromresultsofpressurebuildupteststhattheaveragepermeabilityinthe
regionofthewellboreis,say,100md.Also,hemayknowfromwelllogsthatan
averageporosityis,say18%.Withtheseaveragepermeabilityandporosityvalues
plusinformationonthedistancetotheappropriate Pc 0 datumandinformation
onfluidproperties,hecanmakeareasonablecalculationoftheaveragewater
saturation.
Jtop
gas
GOC
oil zone
oil
fault
Jbottom
WOC
free water
Sw
Toillustratethemethodofgetting S wi ,considerthesketchabove.Atthewellbore
location,thebottomandtopoftheformationare hbottomand htop fromthe Pc 0
datum.Forgivenvaluesof K, , w , o , Cos , and h ,calculate Jtop and J bottom.
Shadetheareaenclosedby Jtop and Jbottom onthe J curveandcalculatethe
averageinitialwatersaturation S wi .
Thesimplestwayofdetermining S wi isbygraphicalintegration.Thus,determine
theareaunderthecurve,dividethisareabythevalue
willgive S wi .Thatis:
S wi
Example3.5
Jtop
Jbottom
SwidJ
J top J bottom
top
Jbottom
andtheresult
100
%
ExamplecalculationoftheuseofCapillaryPressureDatatoObtainAverage
WaterSaturationUsingJFunction
Itisdesiredtocalculatetheinitialoilinplaceforanoilreservoirhavingagascap
asillustratedbelow.ThereisnopriorJfunctioncurveavailableandnowelllogs
togivepermeability,porosity,andsaturationdatawithdepth.Allwehaveareold
coresfromstorage.
Thebulkvolumeoftheoilzoneis1,000acreft.Thethicknessoftheoilzoneis
20ft.Fourcoresamplesweretakenfromtheoilzoneinthemiddleof5ft.
intervals.Fromlaboratorymeasurementsofporosityandpermeability,thedata
are:
Intervaldepth
4,0004,005
4,0054,010
4,0104,015
4,0154,020
Permeability
11.2
34.0
157.0
569.0
Porosity
0.147
0.174
0.208
0.275
well
gas
GOC
oil zone
WOC
fault
free - water table
Thefreewatertableisatadepthof4030ft.Inadditiontoporosityand
permeability,thecapillarypressureforeachsamplewasmeasuredusingair
displacingwaterinacentrifuge.Theselaboratoryderivedcapillarypressure
curvesareshownbelow.Thewater/oilinterfacialtensionforthisreservoiris
estimatedtobe28dynes/cm,thereservoir(water/oil)wettingangleis0.0.The
air/waterinterfacialtensionis70dynes/cmwithawettingangleof0.0also,
w 62.4 Ib ft 3 , o 50 Ib ft 3 .Calculatetheaveragewatersaturationandthe
initialoilinplace.
Solution
a) Convertthe Pcla b datato Pcres dataandcalculatetheJfunctioncurveusing:
Pcres
res cos res
Thishasbeencalculatedandplottedbelow.
K
where Pcres Pclab res
lab
b)Calculatethevalueof J ateach" h "ofeachcoreandreadthecorresponding
watersaturationfromthe J curve.
k
62.4 50
J h
144 28
J
0.738
0.967
1.478
1.748
h
27.5
22.5
17.5
12.5
Sw
0.37
0.35
0.29
0.27
c)Obtainthearithmeticaveragewatersaturation.
Theaveragewatersaturation
Sw
1
0.37 0.35 0.29 0.27 0.32
4
UsingtheLesscorrectmethod
1
K G 11.2 34 157 5694 76.37
1
A 0.147 0.174 0.208 0.275 0.201
4
h 20
76.37
0.8
0.201
Therefore, S w 0.34 byreadingontheJfunctioncurveat J 0.8
J 0.00307 20
Only K and available
Supposetherearenocores,therearenowelllogs,anaverage K isavailablefrom
welltests,and canbeestimatedfromcorrelations.
Inthiscase,usetheheightsofthetopandbottomofthepayzonefromthefree
watertabletoobtain Jtop and Jbottom .
Jtop
Similarly,
htop w o K
144Cos
hbottom w o K
144Cos
Jbottom
Plottheseandfindtheaveragewatersaturationgraphically.
Exercises
1.Giveapossiblereasonwhy forasoapsolutionisabout40dynes/cmandnot
about70dynes/cmaswouldbethecaseforfreshwater.
2.Showthattheexpression
thegeneralexpression:
2Cos
r
derivedforatubeisaspecialformof
1
1
R1 R2
Pc
Pc
3.Giventhat air / water 70 dynes/cm and oil / water 23 dynes/cm ,wouldyou
expectawater/airinterfaceoranoil/waterinterfacetohaveasmallercontact
angleassumingthesamecapillarypressureappliestoboth,inthesamecapillary
tube.
4.Calculatetheentrypressurefornaturalgasintoaporethroathavingthe
followingsizesandshapes.
a)Acylindricalshapeporethroatof0.0001inchdiameter
b) Anellipticalshapeporethroatofd1=0.0001inchesandd2=0.001inches
d2
d1
c)Aninfinitehorizontalfractureoffracturewidth=0.0001inches.
Use=35.2dynes/cm,and=0.0
References
1.Clark,NormanJ."ElementsofPetroleumReservoirs"HenryL.Doherty
Series,SocietyofPetroleumEngineersofAIME,Dallas,1960.
2.SliderH.C.,WorldwidePracticalPetroleumReservoirEngineeringMethods,
PenwellBooks,1983
3.Wilhite,G.P.:Waterflooding,SPETextbookSeries,Vol.3,1986.
4.Standing,M.B.:Lecturenotes,StanfordUniversity,1977
5.Amyx,J.W.,Bass,Jnr.D.M.,Whiting,R.L.:PetroleumReservoir
Engineering,McGrawHill,1960
