Shale problems and water-based drilling fluid optimisation in the Hassi Messaoud Algerian oil field

12 Pages·2012·1.17 MB·English

Open Archive Toulouse Archive Ouverte (OATAO)
OATAO is an open access repository that collects the work of Toulouse researchers and
makes it freely available over the web where possible.
This is an author-deposited version published in: http://oatao.univ-toulouse.fr/
Eprints ID: 5843
To link to this article:DOI:10.1016/J.CLAY.2010.06.008
URL: http://dx.doi.org/10.1016/J.CLAY.2010.06.008
To cite this version: Khodja, Mohamed and Canselier, Jean-Paul and
Bergaya, Faiza and Fourar, Karim and Khodja, Malika and Cohaut,
Nathalie and Benmounah, Abdelbaki (2010) Shale problems and water-
based drilling fluid optimisation in the Hassi Messaoud Algerian oil field.
Applied Clay Science, vol. 49 (n° 4). pp. 383-393. ISSN 0169-1317(cid:3)
Any correspondence concerning this service should be sent to the repository
administrator: [email protected](cid:3)
Shale problems and water-based drilling fluid optimisation in the Hassi Messaoud
Algerian oil field
Mohamed Khodjaa,⁎, Jean Paul Canselierb, Faiza Bergayac, Karim Fourard, Malika Khodjae,
Nathalie Cohautc, Abdelbaki Benmounahd
aSONATRACH/DivisionLaboratoires,Avenuedu1erNovembre,Boumerdès35000,Algeria
bUniversitédeToulouse;INPT,UPS;LaboratoiredeGénieChimique(UMRCNRS5503),4alléeEmileMonso,BP84234,F31432ToulouseCedex4,France
cCentredeRecherchesurlaMatièreDivisée,1RuedelaFérollerie,CNRS,Orléans,France
dLaboratoiredesmatériauxcompositesetalliages,UniversitédeBoumerdès,Algeria
eSONATRACH/DivisionForage,10RueduSahara,Algiers,Algeria
a b s t r a c t
Drillingfluidformulationandpropertiesplayafundamentalroleindrillingoperations.Claymineralsbehave
initially as a beneficial rheological adjuvant in drilling muds. Nevertheless, the contamination of oil
reservoirsbyclaymineralspresentinthedrilledgeologicalformation(shales)maygeneratemajorproblems
duringdrillingasplugformation.Inthiscontext,ourstudydealswiththeoptimisationofdrillingconditions
in theHassi Messaoud Algerian field. Themineralogical heterogeneity of this field is first discussed. The
rheological and filtration characteristics of water-based muds with different polymer and electrolyte
concentrationsareinvestigated.Thephysicalandchemicalchangesofbothdrilledformationanddrilling
fluid during the drilling process are studied. Therefore, depending on the clay present in the geological
formation,anoptimiseddrillingfluidsystemusinganewfiltrationprocedureisproposed.Agoodcorrelation
isfoundbetweenfiltration/rheologicalpropertiesandinhibition.
1.Introduction lubricity and temperature stability. However, their use becomes
restrictedbyenvironmentalregulations,sothatthereisagreatneed
Thecomplexdrillingfluidsrepresent15 to18%ofthetotalcost forenvironmentally-friendlyWBMabletoprovidethesameaccept-
(about$1million)ofpetroleumwelldrilling.Theirformulationand ablerequirements(boreholestability)asOBM.
characterisation need various techniques. The classical water-based The drilling performances of 100 Algerian wells located in the
muds(WBM),atleastspudmuds,containedonlywaterandclaybut HassiMessaoud(HMD)fieldwerecompared.Resultsshowthatmajor
theirperformances,directlyevaluatedbythestabilityofthesystem problems occur in the 12"1/4 (actually 16") interphase and are
and the rheological and filtration properties, were generally poor. observed at 700–950m depth as a consequence of variable rate of
CurrenttendenciesaretoincreasetheWBMperformancesortouse penetration (ROP), varying from 8 to 24m/h. Several problems are
biodegradableadditivesinoil-basedmuds(OBM). mentioned such as lost circulation, specially in Turonian and Salt
Wellboreinstabilityisthelargestsourceoftrouble,wasteoftime Senonianformations,shaleinstabilities,holecleaningproblemsdue
andovercostsduringdrilling.Thisseriousproblemmainlyoccursin to a reduction of annular velocities in enlarged hole sections, well
shales (principally clays), which represent 75% of all formations cavingandcollapse.
drilledbytheoilandgasindustry.Theremaining25%arecomposedof DispersiontestswerecarriedoutonAlgeriancuttingssamplesand
othermineralssuchassand,salt,etc.Thewellboreinstabilityisdueto the results are discussed and compared with field experiments.
thedispersionoftheclayintoultra-finecolloidalparticlesandthishas Cuttingssamplescontainingclaywithdifferentcharacteristics(type,
adirectimpactonthedrillingfluidproperties(CharlezandHeugas, size, content, cation exchange capacity) were chosen from various
1991). wells and at different depths. Numerous tests were performed in
The physical properties and behaviour of shale exposed to a differentexperimentalconditions:
drilling fluid depend on the type and amount of clay in the shale.
Generally,OBMprovideexcellentwellborestabilityandaffordgood 1. conventionaldispersiontestswithoriginalcuttingssamples,
2. observation tests of confined shale pellets of different sizes in
contactwithmuds,
⁎ Correspondingauthor.Tel.:[phone];fax:[phone].
3. filtrationteststhroughshalecompactedinAPIcellwithindifferent
E-mailaddresses:[email protected],[email protected]
(M.Khodja). pressureranges,
4. filtration tests through shale and/or core samples compacted in citedabove,areadded.Clay–polymerinteractionsarethusimportant
Hasslercellunderactualfieldconditions(pressureandtemperature). in drilling fluids (Bailey et al., 1994). The challenge is then to find
whichtypeofpolymer-baseddrillingfluidshouldbeused.
Foralltheseexperiments,theinfluenceofparameters,suchasfluid PHPA is a water-soluble anionic synthetic polymer, which is
type,cuttingsmineralogy,pressureandtemperaturewasstudied. commerciallyavailableindry(granularpowder)oremulsifiedform.
Themostcommonlyusedindrillingforboreholestabilisationinshale
1.1.Shaleinstability formationsisthepartiallyhydrolysed(30%)polyacrylamide.PHPA–
clayslurriestendtoformarelativelythinfiltercakeattheborehole
Whenthewellborewallsbecomeunstable,thespillingofcuttings wall, a characteristic often cited as an advantage (Darley and Gray,
causesadisastrouschangeintherheologicalpropertiesofthemud 1988).Moreover,silicate-containingfluidsshowgoodshaleswelling
(Beihofferetal.,1988). inhibition,lowdepletionrateandhighrateofpenetration(ROP)and
Several studies on shale–fluid interactions confirm that various additionallytheyareenvironmentallyfriendly(Wardetal.,1997;Van
causes are at the origin of borehole instability: water adsorption, Oortetal.,1999;TareandMody,2000).
osmoticswellingandcationexchange.DifferentapproachestoWBM InWBM,poly-(glycerols) and poly-(glycols) (abbreviated in the
designare suggested(Boletal., 1992; Cooketal., 1993; Modyand followingasglycerolsandglycols)havebeenwidelyappliedinshale-
Hale,1993;VanOort,2003). drilling fluids (Chenevert, 1989; Bland, 1991; Bland et al., 1995;
Many works also focused on the selection of drilling fluids for Downsetal.,1993;Cliffeetal.,1995;Reidetal.,1993;Twynametal.,
given shale formations (Darley, 1969; Chenevert, 1970; Roehl and 1994).Theypreventcuttingsfromdispersingintothemedium(Bailey
Hackett, 1982; Beihoffer et al., 1988; Zamoraet al., 1990; Hale and etal.,1994).Therefore,theyincreasedrillingrates(Reidetal.,1993;
Mody,1992;Baileyetal.,1994;Simpsonetal.,1994;Durandetal., Cliffeetal.,1995).
1995; Horsud et al., 1998; Pernot, 1999). More recent studies on Moreover, potassium salts have been used for a long time as
shale–fluid interactions suggest a new approach to WBM design swelling inhibitors in WBM. The inhibition is explained by the
(Lomba et al., 2000; Schlemmer et al., 2002; Van Oort, 2003). possiblepenetrationofsmallnon-hydratedionsintotheporosityof
Considerationisgiventomaintainboreholestabilisationinreactive the shale (Simpson et al., 1994), thus forming an effective semi-
shales by reducing hydration (swelling) and/or dispersion. This permeablemembrane.Organic(xanthangum,PAC,PHPAandPAG)or
processisgenerallyreferredas“inhibition”. mineral (sodium silicate) polymers are probably too large to enter
Scarce research and few field tests have been conducted on shale pores. Some theoretical models (Van Oort et al., 1994, 1995,
AlgerianfieldstoinvestigateWBMeffectsondrillingoperations.Van 1999) explain the reduction of the filtrate flow into shale by both
Oort (2003) considered the replacement of OBM, currently used in mechanisms, an increase of the viscosity leading to a reduction of
someAlgerianfields,byWBM.Thisauthorshowsthatthepresenceof shalepermeabilityandaflowofmudfiltrateintotheshaledrivenby
additives in WBM, such as polymer and KCl, aims to reduce shale osmoticpressure.
instability.Claywetabilityandinhibitionpropertieswerestudiedby
analysingthebehaviourofwater–clay–polymer–electrolytesystems. 1.3.Filtrationandinhibition
These properties are connected to the rheological and filtration
characteristicsforbothmudandfiltrate. Theknowledgeofthefiltrationpropertiesisveryimportantinthe
Cuttingscharacterisation is a key parameter to explainhow salt design of drilling fluid formulation. Some works (Loeber, 1992; Li,
added to WBM, affects shale stabilisation. Recovered cuttings, 1996;Argillieretal.,1997;Bennaetal.,1999andBennaetal.,2001)
generally contaminated by drilling fluid, are washed. Specialised have shown that the filtration across the cake depends on several
laboratoriesrecommendcuttingssolventwashing(GuptaandSantos, parameters such as initial clay content, particle or aggregate
2002; M.I. Corporation, 2002). The washing could lead to positive association, water retention and permeability, experimental condi-
effectssuchaspluggingwhichreducespermeabilityandfiltration,or tions,etc.FergusonandKlotz(1954),showthat70%to90%ofthetotal
negative effects in inhibitive tests such as contamination of shale filtratevolume,flowingthroughpermeableformations,occursduring
sampleswhichaffectspolymerevaluationseriously.Thedevelopment mud circulation. During this dynamic filtration, the invasion radius
of a prototypedevice tocollect and preservea washed, continuous reachesavalueof85%.Aconstantflowrateisreachedwhenfiltration
stratigraphicsequenceofdrillcuttingsisworthmentioning. forces, leading to the formation of a mud cake, are balanced by
hydrodynamicforces,i.e.mudcirculationthaterodesthemudcake.
1.2.Roleandcompositionofdrillingfluids
1.4.Shalecharacterisationandinhibitiontechniques
Drillingthewellboreisthefirstandthemostexpensivestepinoil
andgasindustry.AlthoughOBMgivegreatershalestabilitythanWBM Themainmethodsdevelopedforshalecharacterisationandfluid
(Bol et al., 1992), several WBM systems have been developed to inhibitionperformancesdealwithcomposition,reactivity,mechanical
replace OBM in order to respond to environmental regulations andphysico-chemicalpropertiesofshales(orclay):
(Simpson et al., 1994; Friedheim et al., 1999; Patel et al., 2001;
YoungandMaas,2001;Schlemmeretal.,2002). - X-raydiffraction(XRD)analysistodeterminequalitativemineral
For laboratory tests, a typical mud contains several additives at content,
concentrationscommonlyused,includingaviscosifier(xanthangum - Cationexchangecapacity(CEC)andmethylenebluetest(MBT)to
withorwithoutbentonite),afluidlossreducer(polyanioniccellulose: evaluate reactivity and shale factor of drilled cuttings. The MBT
PAC), and different polymeric swelling inhibitors such as partially methodwasrecommendedbyAPI13I,Section11(2003),
hydrolysed polyacrylamide (PHPA), sodium silicate and polyalkyle- - Agravimetricswellingtest(GST),usedtomeasurewaterandion
neglycols(PAGor“glycol”)toimproveshalestability. motionduringshale/mudinteraction(Zhangetal.,2004),
Bentonite, a worldwide-used drilling fluid additive, is added to - Capillary suction time (CST) for determination of filtration
fresh waterto increase hole cleaningproperties and to forma thin propertiesandsaltconcentrationoptimisation(Wilcoxetal.,1987),
filtercakeoflowpermeability.Itsmainfunctionsaretoviscosifythe - Penetrometertoestimatethedegreeandthedepthofsoftening
mud and to reduce the loss of fluids in the formation. In order to (Reid et al., 1993) or “Bulk Hardness Test” designed to give an
stabilise clay particles and to prevent their swelling/dispersion assessmentofthehardnessofshalefollowingexposuretoatest
behaviourinthepresenceofwater,otheradditives,suchaspolymers fluid(Pateletal.,2002),
- Dielectricconstantmeasurement(DCM)toquantifyswellingclay Some shale cuttings and core samples from Hassi Messaoud
contentanddeterminespecificarea(LeugandSteig,1992), Algerianwellshavebeenanalysedasfollows:
- Triaxial test for pore pressure measurements, carried out in
- Cuttingswereair-driedatroomtemperatureandpowderedina
downhole simulation cell (DSC) for compressive stress/strain
porcelainmortar.
behaviour(SalisburyandDeem,1990),
- The moisture content of each shale was measured by drying at
- Oedometer test for pore pressure modification and chemical
105°Cuntilaconstantweightwasobtained.
potentialinfluence(Boletal.,1992),
- Accordingtosomeearlierresultsregardingorganiccontamination
- Slakedurabilitytest(SDT),astandardmethodoriginallyusedin
(Benayadaetal.,2003;Khodja,2006;Khodja,2008),allcuttings
geotechnicalstudieswhenmeasuringtheweatheringandstability
samplesusedwerewashedwithn-hexaneanddriedat105°C.The
ofrockslope:ASTMD4644-97(ASTM,2000), reapproved 1992
absenceoforganiccarbonafterwashingwasnoted.
(Likosetal.,2004),
- ThemineralcompositionwasdeterminedbyXRDanalysis,witha
- Jarslaketesting,aqualitativemethoddesignedtoevaluateshale
PhilipsPW1710diffractometer.
relative durability in contact with a given fluid. Wood and Deo
- Thecationexchangecapacity(CEC)wasdeterminedbyusingthe
(1975), Lutton (1977) describe details of this method using six
cobaltihexamminetrichloridemethod(Chauveteauetal.,1988).
indices,
- DifferentfluidsystemswerepreparedusingAPIequipments(API
- Differential strain curve analysis (DSCA) for in-situ measuring
RP13B-1,2003).
stressorientationandintensity(Fjaer,1999),
- Densities, pH and rheological parameters were determined. The
- Hot-rolling dispersion test (shale disintegration resistance or
rheologicalmeasurementswereconductedatvariablespeed(3to
cuttings dispersion test), the most widely used technique in
600rpm)usingaFann35AviscometergivingvaluesincPorin
optimising drilling fluid. Appreciated for its simplicity, low cost
mPa.s and with using the following formulas from API recom-
and duration, it has been recommended by several laboratories
mended practice for field testing drilling fluids. The numerical
andadoptedbyAPI(1997).Itconsistsofaddingaknownamount
valueoftheplasticviscosity(PVincP)isgivenby:(600rpmdial
ofshalecuttingstoastandardvolumeoftestfluidcontainedina
reading—300rpmdialreading),apparentviscosity(AVincP)is
steelbomb.Thebombisrolledforafixedtime,usually16h,ata
givenby:[(600rpmdialreading)/2],andYieldPoint(YPinPa)by:
given temperature; the shale is then recovered on a sieve. The
2(AV–PV).Otherrheologicalmeasurementsonpolymersolutions
amountofrecoveredshaleisexpressedasapercentageoforiginal
werecarriedoutwithAR2000equipment(TexasInstruments).
weight. High percent recoveries and low moisture contents are
- APIfiltrate,andgel0/10(3rpmdialreadingaftermixingandafter
indicativeofinhibitivefluids.Clearlywithapoorlyinhibitivefluid,
10min)aredeterminedwithusingAPIrecommendations(APIRP
cuttings will disperse into the fluid and zero recovery (and
13B-1,2003).
therefore no moisture content determination) will result. If two
- Cuttingssamplesofdifferentsizes(36to800μm)wereprepared
fluidsgivethesamerecoveryratio,thefluidwhichgivesthelower
forfiltrationoperations,withinthe20–150kPapressurerange.
moisture content is regarded as being slightly more inhibitive.
- Fluid displacement test was determined with Corelab filtration
Indeed,alowerwateruptakebythecuttingsreducestheriskof
system equipment (Argillier and Audibert, 1999; Muniz et al.,
dispersion or swelling in the wellbore. A comparative measure-
2005). Core samples were saturated with synthetic formation
ment of inhibition can be obtained by considering the relative
brinesolutionduring24h,placedinaHasslercellunder14kgf/
cuttingsweightsretainedoneachscreensize,
cm2ofoverbalancepressureunder80°C.Soltrol130wasusedfor
- Shalepelletinhibition(pelletdispersiontest):shalecuttingsare
filtrationtestsinaHasslercelltoevaluateinitialandfinal(after
dried and ground to less than 80 mesh, then mixed into
fluidinjection)permeabilities.Returnpermeabilityorpercentage
homogeneous paste with 10wt.% water. Pellets are made by
ofdamage(D)isdeterminedfromcomparisonofinitialandfinal
pressing approximately 20g of this paste in a carver using a
Soltrolpermeabilitiesinthestablestate.
hydraulicpressunder7000psifor2min(ModyandHale,1993).
- Newfiltrationtest:
Pelletsandfluidareintroducedinasteelbombandprocessedas
In drilling fluid, “inhibition” covers all the mechanisms that can
above(hot-rollingdispersiontest).Forcomparisonandreference,
reduce or/and eliminate swelling, dispersion, and clay–water
anOBMsystemisgenerallyused,
interactions in order to enhance shale wellbore stability during
- Pressuretransmissiontest,usedforconfinedorunconfinedshale
drilling.Inhibitionpercentageisthedifferencebetweeninitialand
(Van Oort, 1994). Muniz et al. (2004) described an apparatus
finalcuttingsweightrecoveryafterfluidcontact.API(1997)give
designedtoevaluateshale–drillingfluidinteractionandestimate
somerecommendationsaboutthoseinhibitionmethods:
shalepermeability,coefficientofreflectivity(membraneefficien-
a. Testisonlyarelativemeasureandshouldbeincludedasapartofa
cy)aswellasionicdiffusioncoefficient,
comprehensive testing program,and as a comparison ofvarious
- Microbitdrillingequipment,requiringcoresampleavailabilityand
wholemudcompositions,
costlyinvestment(Lamberti,1999).
b. Itisstronglyrecommendedthattheshaleismaintainedasnearas
possible its in-situ moisture content and must not to be air- or
2.Materialsandmethods
oven-driedbeforetesting.
c. Thedrillingfluidrheologyparameterhasproventobedifficultto
AlgerianbentonitewasusedinWBMformulation.Theotheraddi-
controlfromtesttotest.Modestchangesintherheologyfromone
tivesprovidedbyMISwacoAlgeriaarei)xanthangumasviscosifier:
fluidtoanothercanstronglyinfluenceshaledispersionfinalresults.
thiswater-solublepolymer,slightlyanionicandhighlybranchedisa
veryeffectivestabiliserforaqueouscolloidalsystems,ii)polyanionic Thecomparisonbetweenvarioustechniquesshowsanimportant
cellulose(PAC)asfluidlossreducer:thiswater-solublepolymeralso contribution of each method. However, these methods are often
actsasviscositymodifierandisavailableintwotypes(high-orlow- criticisedregardingfeasibility,cost,precisionandconditionsused.
viscositygrade),bothofwhichimpartthesamedegreeoffluidloss Inthispaper,weproposeanewmethodcombiningdispersionand
controlbutdifferentdegreesofviscosity,andiii)sodiumsilicateasa pellettests.Byusingthismethodweaimtoprotecttheinitialquality
mineralinhibitor,used toimprovelubricity andshalestability, and ofgotbackcuttings,minimisegrindingandavoidmoistening,while
also two polymers as inhibitors: a partially hydrolyzed polyacryl- opting for a preliminary wash to eliminate the contamination of
amide (PHPA) and polyalkyleneglycol (PAG provided from BASP- cuttingsbytheadditives(polymers,surfactants,etc.)inthedrilling
Baroid,Algeria). fluids.
Table1
Mineralogicalcompositionofcuttings,coressamplesandAlgerianbentonite.
%Clay %Nonclay Kaolinite Illite Chlorite I-Mont.a Montmorillonite Quartz Calcite Dolomite Anhydrite Barite Halite
Turoniantop 0 101 – – – – 0 3 1 7 87 3 –
Turonianmedium 0 100 – – – – 0 3 38 19 36 4 –
Turonianbottom 0 98 – – – – 0 5 52 7 24 6 4
Cenomanianmedium 0 83 – – – – 0 6 45 5 20 3 4
Medium16"intervalb 13 87 5 70 5 20 0 19 6 8 13 11 30
Reservoircore1 6 94 50 5 45 – 0 94 – – – trc tr
Reservoircore2 8 92 95 5 – – 0 92 – – – tr –
Bereasample 19 81 75 15 5 5 0 80 tr 1 tr – –
Algerianbentonited 83 17 – 5 – – 95 13 4 –
a Interstratifiedillite–montmorillonite.
b Sampleobtainedwithmixingcuttingsissuedfromseveraldepthin16"phaseinHassiMessaoudwell.
c Traces.
d Reference.
Our new proposed method combines filtrate data (volume and 3.2.Drillingfluidperformances
rate)withrheologicalandinhibitiveproperties.
The stability of drilling fluids is generally indicated by its visual
3.Results homogeneityafteralongperiodofageing.ForOBMsystems,aphase
separationandadecreaseofviscosityaredirectsignsofdegradation.
3.1.MineralogicalcompositionandCEC In WBM, phase separation is also an indication of mud instability.
Fig.3summarisesthebehaviourofdrillingfluidstateevolution.The
Mineralogicalcompositionsofcuttings,coressamplesandAlger- viscosityofmudaffectsthedispersionandtheswellingofshalesand
ian bentonites are reportedin Table 1. All Turonianshalecomposi- decreasesthediffusionvelocityinporousmedium.Mudswithhigh
tionsaresimilar.Topsampleshowsanhydritepredominance(87%). viscosityandaminimumfiltratevolumearepreferredforinhibition
Theamountofthismineraldecreaseswithdepth,ascalcitecontent efficiency,accordingtoclassicalfiltrationequations.
increasestill52%inthebottomsample.ThesamplenotedbinTable1 In the first part, the rheological behaviour of polymers used in
is representative of an average composition frequently present in conventionaldrillingformulationsisstudied.Fig.4aandbshowshear
Hassi Messaoud shales (Fig. 1). The presence of salt in shales was stressversusshearrateforxanthanpolymersolution(0.005to0.2wt.
confirmed by XRD and Scanning Electron Microscopy (Fig. 2). %)andPAC(0.05to1wt.%),respectively.Asexpected,atlowshear
Contaminationbysaltisprobablyrelatedtothedrillingformulation rate, xanthan solutions show shear-thinning behaviour, but above
used(OBMwithNaCl-saturatedwater).Indeed,saltisaddedtothe 500s−1, fluids behave as quasi-Newtonian (0.2wt.%) or slightly
waterphaseinordertoenhanceemulsionstability(electricaleffect shear-thickening(lowerconcentrations).Thelatterobservationmay
andwateractivity)andtoincreasebothdensityandviscosity. beanartifactduetoturbulenceinducedbyhighshearratesonthe
For reservoir cores and Berea sandstone sample, high quartz low-viscosityfluid.
contentandkaolinitepredominate.Itisnoticedthatquartzispresent Atthesameconcentration,theviscosityofPACsolutionismuch
inallsamples.TheCECofallsamplesarelow,from3.4to6.3meq/ higher than for xanthan solutions. Addition of KCl to a mixture of
100gclayexceptforAlgerianbentonitewhoseCECis60meq/100g. xanthan(0.2%)andPAC(1%)seemspredominantlytodecreasethe
In field conditions, problems occurring during drilling are not viscosityincomparisonwiththeviscosityoftheequivalentPAC(1%)
systematically related to the presence of clay because the samples solution,evenifaslightincreaseisobservedastheKClconcentration
testedinTuronianandCenomaniandonotcontainclays(Table1).For (Fig.4c).Indeed,itisknownthatanionicpolymerPACisparticularly
example, borehole stability problems at the Turonian and Cenoma- sensitivetomonovalentanddivalentcations(K+,Ca2+,Mg2+)and
nian levels are solved by HCl injection for the dissolution of car- may induce polymer dehydration leading to a viscosity loss of the
bonates,especiallyabundantinthoseformations(Table1). polymersolution.
Fig.1.HassiMessaoudAlgerianshaleanalysis.
Fig.2.XRDandscanningelectronmicrographofshale(16"HassiMessaoudsample)andsaltcontaminationcuttings.
Inthesecondpart,therheologicalbehaviouroffourmudsystems gels).AsregardsthePHPAsystem,themudandthefiltrateseemto
(PHPA,PAG,silicate,andspudmud)whosecompositionsarereportedin followdifferentmodels(Table3).
Table2isstudied.Powerlawmodelswereoftenproposedtodefinethe TheselectedPHPAformulationandspudmudwithPHPApolymer
behaviourofKCl/polymertypedrillingfluids(KökandAlikaya,2004) possessacceptablerheologicalandfiltrationcharacteristics(Table4)
butmostdrillingfluidsdonotconformexactlytoanyoftheproposed accordingtoHMDdrillingfluidrequiredproperties(density:1.20to
models. Herschel–Bulkley and Ostwald–de Waele rheological models 1.25,PVb30cP,Gel:7to15lb/100ft2andAPIFiltrateb10mL.PHPAis
wereretainedforthosesystems(mudsandfiltrates).Theestimationof presentin thesamerange ofconcentrationsin themudandin the
the three rheological parameters for Herschel–Bulkley and the two filtratebutdoesnotshowthebestinhibitiverole(Table3andFigs.6
parametersforOstwald–deWaeleequationsareshowninTable3. and 9) whereas for the two other systems, silicate and glycol, the
The PAG and PHPA systems present similar properties, whereas formed cake is less permeable and thus the filtrate is less
the silicate system exhibits the best results (viscosity, filtrate and concentrated. In fact,preliminarydrillingoperationsin HMD, using
silicatehaveshowngoodresultswithbothtendenciestoresolveshale
instabilityandtorespondtoenvironmentalregulations.
TheflowbehaviourofsomemudsystemsisillustratedinFig.5.A
similar trend is observed for all the studied solutions at increasing
shearrate.ThefiltratevolumesobtainedfromPHPA,silicateorglycol
systemsarecomparedinFig.6.Withthesamepolymerconcentration
(Table2),thesilicatesystempresentsahigherviscositythanthePAG
one(Table4andFig.5).Filtratevolumes(Fig.6)increaseintheorder:
silicatebPAGbPHPA.PAGandPHPAshowidenticalfiltratevolumes,
butnotthesamefiltrationvelocity.Thisbehaviourcanberelatedto
polymermolecularweightandshale–polymerinteractiontype.With
the same fixed polymer concentrations (xanthan, PAC, PHPA and
silicate),thePHPAsystemgivesalowerinterfacialtensionthanthe
silicateone,probablyduetoalowerviscosity.Lowviscositypromotes
Fig.3.Representationofdrillingfluiddestabilisation(phaseseparation:lowviscosity wettabilityandencapsulationthatwidelygovernsinhibitivemechan-
and high filtrate). A: Undesirable dispersion with an inhomogeneous additive isms(Khodja,2008).
repartition. Solid–liquid, polymer-solution, dispersed phase-continuous phase are
inhomogeneousandunstable.B:Optimaldispersionwithauniformadditiverepartition.
Themudsystemisstableandexhibitsgoodrheologicalandfiltrationcharacteristics. 3.3.Shaleinhibitionandfiltrationtests
C:Themudsystemisunstableforoneofthefollowingreasons:dramaticfiltration
conditions(pressureandtemperature),useofanincompatibleadditive(contaminant) Shale testing helps to develop inhibitive WBM systems. In our
orofpoorqualityproducts,orconsiderableaging.Solids,polymers,andsaltinWBM,
study,amongallthetechniquesmentionedbefore,twomethodswere
dispersed phase, emulsifiers or other additives in OBM are separated from the
used: hot-rolling dispersion test and pellet test. Moreover, a new
continuousphase.Thesystempresentsaphaseseparation,involvingadegradationof
rheologicalparametersandahighfiltratevolume. procedure for filtration across pellet in an API cell was used to
Table2
Water-basedmudformulations.
Additive Bentonitea KCl Xanthan PHPAborPAGc PACd
(g) (g) (g) orSilicate(g) (g)
Concentration 20 30 1 5 5
a Bentoniteprehydratedinwaterduring24h;otheradditivesintroducedinto1Lof
bentoniteslurry.
b Partiallyhydrolysedpolyacrylamide.
c Polyalkyleneglycol.
d Polyanioniccellulose.
Table3
Herschel–BulkleyandOstwald–deWaeleparameters.
System Herschel–Bulkleyaτ=τ0k⋅γ⋅n Ostwald–deWaeleτ=k⋅γ⋅n
PHPAbmud
τ0 0.686
k 2.654
n 0.538
PHPAfiltrate
k 3.418
n 0.519
PAGcmud
τ0 2.013
k 0.626
n 0.643
PAGfiltrate
τ0 1.459
k 0.220
n 0.702
Silicatemud
k 10.86
n 0.317
Silicatefiltrate
k 0.080
n 0.712
a τ:shearstress(Pa);τ0:shearstressatthethreshold(Pa);k:consistencyindexof
themedium;γ:shearrate(s−1);n:flowbehaviourindex(adimensional,0bnb1).
b Partiallyhydrolysedpolyacrylamide.
c Polyalkyleneglycol.
low cost. Some laboratories recommend using a core sample for
inhibitionevaluationbutcuttingsaregenerallyusedbecauseofcore
unavailability.SeveralcuttingssamplesfromvariousHassiMessaoud
geological formations are used for laboratory dispersion tests on
silicate-orPAG-containingWBMandOBMsystems(Table4).These
tests show that the inhibition percentages obtained with silicate-
WBM(96–98%)aresimilartothoseofOBM,buthigherthanthoseofa
PAG-WBMsystem(78%).Numerousintegritytestsperformedonfield
samples with PAG and silicate systems, lead to select silicate
formulationswhichareoptimisedandproposedforpilottestonthe
Hassi Messaoud field (Khodja, 2003). Hot-rolling dispersion test
clearly shows a similar behaviour for all samples with silicate and
PHPA(Fig.7).
Severalprocedures recommend comparing initial and final sizes
(orweights)forestimatinginhibitionafterfluidcontact.Thequestion
Fig.4.a:Shearstressvs.shearrateforxanthanpolymersolutions.b:Shearstressvs.
Table4
shearrateforPACpolymersolutions.c:Shearstressvs.shearrateforxanthanpolymer PropertiesofthePHPAaformulation.
solutions(0,2%)+PAC(1%)atdifferentKClconcentrations(2to5%w/w).
db PVc AVd YPe Gel0f Gel10f Filtratevolume
(g/cm3) (cP) (cP) (lb/100ft2) (lb/100ft2) (lb/100ft2) (mL)
evaluateshaleinhibitionwithmudsystems,usingdifferentinhibitor 1.03 22 42.5 41 7 8 8.5
polymers(PAG,silicateandPHPA). a Partiallyhydrolysedpolyacrylamide.
b Density.
c Plasticviscosity.
3.3.1.Hot-rollingdispersiontest
d Apparentviscosity.
Thehot-rollingdispersiontestrecommendedbyAPIandadopted e Yieldpoint.
byseverallaboratories,iswidelyappreciatedforitssimplicityandits f ω3aftermixingandafter10min.
HMD geological formations and using PAG-WBM or OBM systems
(Tables5and6).
3.3.3.Newfiltrationtest
The new API water loss (or fluid loss) test was carried out by
replacingWhatman50filterpaperbythepelletintheAPIfiltration
cell(Fig.8).Theslurrywasexposedtoa100psipressurefor30minto
obtainfiltrate.Fig.9showsthatforallsystems,thefiltratevolumeis
proportional to compaction force. Whatever the compaction force,
silicate and PAG systemspresent the lowest filtrate volumes. PHPA
formulationunderdifferentcompaction forces(from60to150kN)
clearlyshowslargefiltratevolume.Evenbelow60kN,thewholefluid
wasfiltered.Thecompactionforce,linkedtothedepositmodeofthe
sediments,hasasignificantinfluenceonthepermeability.
Itisnecessarytounderlinethatthedifferenceofcompactionisone
of the difficulties met during these experiments. The compaction
force,theinitialwatercontentofthepowderandthegrainsizeare
takenintoaccountsincetheseparametershaveaninfluenceonthe
Fig.5.Shearstressvs.shearrateofwholemudsystems. mechanicalbehaviourofthecompactedmaterial.
Fig.10presentsaHasslercelloperatingunderreservoirconditions
is:whichisthemainfactor(claytype,claycontentorcuttingssize) (temperature and pressure) for the filtration of pellets compacted
affectingthedispersionresults? under 10kN. Permeability (k) was calculated using Darcy's law
With all systems, our results show similar, rather high recovery equation.
valuesforlargesize(0.8mm)(Fig.7a)butlowrecoveryvaluesfor
small size (0.100 to 0.315mm) cuttings (Fig. 7b). When using dif- dV=dt=k:ΔP=η:e:A ð1Þ
ferentinhibitivepolymers,almostnodifferenceinrecoveredweightis
noticed between cuttings samples from different geological forma- ΔP:pressure;A:Area(cm2);η:viscosity(cP);V:filtratevolume
tionsandwithdifferentmineralogicalcompositions. aftert(cm3);e:cakethicknessaftert(cm);t:time(s);andk:cake
Ourrecommendationisthentouse,indispersiontests,preferably permeability(1darcy=1μm²).
smallsizecuttings,whichareinclosecontactwithalladditivesusedin Damage(D)isdeterminedbycomparinginitial(k)andfinal(k)
i f
drillingfluidsystems.Moreover,whenusingsmallsizecuttings,clays permeabilitiesinthestablestate.
arefullyexposedtothefluidandaggregationeffectiseliminated.
Xanthan gum or PAC added as a viscosifier, acts synergistically D=100 k–k =ki ð2Þ
! i f"
with PAG and preserves cuttings integrity. To increase glycol
efficiency,aninhibitingion,preferablypotassium,wasused.Forthe
HasslercellresultswithTuroniancuttingspelletsshowclearlythat
silicatesystem,analysesshowhighadsorptionofsilicateiononshale.
PAG exhibits low damage values (D) (Table 7). Higher values are
Theinhibitionmechanismalsodependsonthetypeofpolymerused,
obtainedwithPAC,xanthanandPHPAsolutions.AnalysingDandV
controlled by plugging of clay pores, thus reducing the dispersion
(PAG),orbysurfacecoating(filmformationwithPHPAorsilicate). values,mechanismsforxanthanandPAGsolutionscanbeproposed.
Thesetwopolymersdonotreducethefiltratesufficiently;however
xanthan exhibits a high viscosity and PAG high water absorption.
3.3.2.Pelletdispersiontest
Higher damage value with xanthan (67%) is due to its helical
Ahomogeneouspasteisformedbymixingadried,groundshale
molecularstructure,highmolecularweight(MW)andanioniccharge
(b 80 mesh) with 10% (wt./wt.) water. Pellets are obtained by
comparedwiththelowmolecularweightofelectricallyneutralPAG.
pressing approximately 20g of this paste in the carver hydraulic
XanthanandPAGpreventflocculationthankstostericandCoulombic
pressat7000psifor2min(Fig.8).Theresultsobtainedwiththistest
effects.
andhot-rollingdispersiontestaresimilarforsamplesfromdifferent
Onthecontrary,silicate,PACandPHPAyieldlowfiltratevolumes
but relatively high damage values. The relatively thin filter cake
formedusingPHPAisnotsufficienttopreventexcessivefluidlosses.
The following step concerns the influence of polymers on the
drillingfluidflowthroughnaturalporousmedia:reservoircoresand
Bereasandstonesamples(Table8).Inhibitivepolymers:PHPA,PAGor
silicate,areaddedselectivelytoxanthan-andPAC-containingfluids.
Theresultsrevealthefollowingpoints:
1. forsimilarporosity,Ø,andairpermeability,k ,theadditionofKCl
air
toPHPAreducesdamage(D)andfiltratevolume(V),
2. forsimilarrockandairpermeabilities,thesilicatesystemexhibits
highdamagecomparedwiththePHPAsystem,
3. The PHPA system gives similar filtration properties in the new
filtrationtestinAPIcell(pellet)andinHasslercell(corereservoir)
(Fig. 9 and Table 8). Damage values are 65% (Table 7) and 38%
(Table 8), respectively, for PHPA with pellet and PHPA with core
reservoir.IntermsofdamageratioonBereasandstone,Table8shows
that final Soltrol permeability (k) for PHPA without salt is the
f
highest, due to easier desorption. Paradoxically, PHPA gives the
Fig.6.MudsystemsAPIfiltratevs.time. lowest filtrate volume (0.7mL) with minimum damage (35%)
Fig.7.a:HotRollingdispersiontestwiththesamecuttingsize(800μm)anddifferentmudsystems.b:HotRollingdispersiontestwithasilicatesystem.
Fig.8.Cuttings,coressamples,andprocedureusedinfiltrationmethods.
Table5 (Tab.7and8),butthesilicatesystemgivesminimumfiltrateand
Hot-rollingdispersiontestwithwater-basedmuds(WBM)andoil-basedmuds(OBM) maximumdamage(Tables7and8).
systems.
4. Withasmallmolecularsize,theenvironmentally-friendlysilicate
Sample Polymer- Inhibition Ratio(%) OBM systempresentshighviscosity,lowfiltratevolumeandhighdam-
free-WBM PAG-WBM silicate-WBM age.It isthenrecommendedtouse it in upperlayersbut notin
Albiantop 52 79 98 98 reservoir.
Barremianbottom 51 77 98 98
Turoniantop 53 76 98 99 Practically, drilling engineers need to optimise formulations in
oppositewaysdependingonwhethertheydealwithuppergeological
layers or reservoir formation. In the former case, minimum filtrate,
Table6 optimalviscosityandhighdamagearerequiredinfluidformulation
Pelletdispersiontestwithpolyalkyleneglycol-water-basedmuds(PAG-WBM)andoil- selection. In the latter one, low damage is the principal selection
basedmuds(OBM)systems.
parameter.Generally,asilicatesystemisnotusedinreservoir.Silicate
Sample %Cuttings %Cuttings %Cuttings %Cuttings %Cuttings reacts readily with Ca++ and Mg++ ions. High concentrations of
PAGtest1 PAGtest2 PAGtest3 average(PAG) OBM divalent ions will deplete the effective silicate concentration and
Albiantop 60 62 63 62 99 decreaseitsinhibitiveperformance.However,thein-situgelationof
Barremianbottom 61 63 65 63 98 silicateshasbeenemployedtoreduceaqueousfluidflowinoil-and
Turoniantop 61 63 62 62 98
gas-bearingformations.
Evaluation made on the cuttings samples of HMD, based on
laboratory tests, showed a high percentage of inhibition with the
(Table8).Thisresultisprobablyrelatedtoasurfaceaccumulationof silicate system in comparison with glycol and PHPA systems. The
PHPA, reducing the flow through more compact porous Berea studyofthemechanismsofinhibitionofpolymersshowsthatPHPA
sandstone.Byusingapelletandinthenewfiltrationtest,afiltrate inhibitsclaydispersionbyencapsulationandactsbystericeffectbut,
volumeprovokesthemaximumdamage(Table8andFig.9)dueto for glycol and silicate, electrokinetic effect governs inhibition
swelling(clay–waterinteraction),soliddisintegrationandplugging mechanisms.
phenomenathroughporousmedia.Thecomparisonofdamagefor ThefieldtestsrealisedonfourwellsintheHassiMessaoudfield
both silicate and PAG systems, shows that the PAG system gives withasilicatesystemshowarateenhancementandareductionof
maximumfiltrateandminimumdamagebyusingapelletandBerea lossphenomenaandshaleproblems(Khodja,2008).Inthisformation,
characterisedbyitsaveragereactiveandrichfracturing,itisdesirable
toformulateasilicatedrillingfluidwithaperfectinhibitivecapacity,
ideal rheological properties and good filtration control. After labo-
ratoryandfieldtests,andcollaborativeworkwithoilanddrillingfluid
servicecompanies,anoptimalformulationofasodiumsilicate/KCl/
specificpolymerdrillingfluidsystemwasdeveloped.
Thedrillingrate(ROP)ispracticallyidenticalforallcaseswiththis
silicate drilling fluid (7.23–7.49m/h). One of the key factors to
maintain acceptable rheological properties and filtration control
characteristheproperproportionofeachadditive:KClandsilicate.
This choice of silicate and KCl concentrations is a function of some
parameterswhichcannotbeoptimisedonthelaboratoryscale,such
asreactivityandsizeofshale,drillingparameters(WOB:weighton
bit,ROP,mudpressure,bittype...).Infourdrilledwells,theproportion
of silicate ranges from 40 to 74kg/m3, and KCl concentration
decreasesfrom14%inthefirstwellto4.5%inthethirdone.According
to McDonald et al. (2002), this system presents high inhibition,
offering Health, Security and Environment benefits over traditional
OBM. Moreover, based on the review of various drilling fluid
candidates, it was determined that a silicate-based system was the
Fig.9.Mudsystemfiltratevolumevs.compactionforce. bestchoice.
Fig.10.Filtrationcellforformationdamageevaluation.