You are on page 1of 8

RockMechanics, Aubert/n,Hassan/& M/tn (eds) 1996 Balkema,Rotterdam.

ISBN 90 54 I0 838 X

Usage andapplicability of pseudo-3D stress analysis in borehole stability problems in petroleum drilling andproduction operations
G.G. Ramos & B.S.Wilton

ARCOExploration & Production Technology Co.,Plano,Tex.,USA


A. E Polillo
Petroleo Brasileiro S.A., Rio de Janeiro, Brazil

ABSTRACT: In a drilling/production operating environment, wellbore instabilities arise in allthree stages of a well's lifespan: drilling, stimulation and production phases. Thetiming and severity of theoccurrence of such borehole problems dictate which method ofstability analysis should beused. Pseudo-3D codes areplane-strain, non-isotropic subsets of full3D numerical andanalytical codes. Their speed, portability, andease of usehave popularized them among operations engineers. There are numerous versions ofpseudo-3D stress analysis, from simple linear-elastic to sophisticated poro-elastic-plastic, each withitsownadvantage thatsuits a particular wellbore problem. Simple linear elastic codes are re-emerging in popularity because of ease of useandfieldcalibration schemes.

1 INTRODUCTION

2.1 Drilling stage

Recenttechnological advances are pushing the reachof boreholes beyond25,000 ft in length.

Thisstage warrants anintegrated stability analysis because it is themost capital-intensive. A review of


recent advances in drillingERD wellsis givenby

Highlyinclined, extended-reach wellbores (ERD)


mustremainopenfor prolonged time-periods, not
of a reservoir. In a commercial operating environment, the technical staffmustperform long-

Payne,Wilton and Ramos(1995). The main


concern is to determine the mud composition and

only during thedrilling program butalso in thelife

density whichwill maintain the integrity of the


well, without the loss of drilling fluids.

rangeplanning to avoidpotential drillingand production hazards, andstill be ableto quickly


solve occurrences of unplanned wellbore

instabilities. This paper presents strategies in


encountered in day-to-day fieldactivities.

Conventionally, thechoice of muddensity or well pressure Pw isdictated bythe highest formation pore pressure Pr along the well path,Figure1. Other operational andgeological factors listedbelow
must be considered:

analyzingvarious wellbore stressproblems

Abnormally pressured layers, depleted zones


Fractured formations, lossof circulationzones Penetration rate,mud/ clay compositions Differentialsticking, kicksandblow-outs

2 INSTABILITIES IN FIELD OPERATIONS

Therearethree stages in thelife of a well:


1. dtilling,

Coting recovery, cementing efficiency Formation damage, logging, well tests

2. completion andstimulation, and 3. flow tests, production, anddepletion. Thesedifferentusesandstages in the life of a well
shoulddictatewhich methodof stabilityanalysis is

find the feasible/acceptable limits of the wellbore

appropriate and applicable. Theusual problem isto

Commonly encountered instabilities and their presumed mechanisms of failure arelisted in Table 1, assuming thatmudchemistry hasalready been optimized. Instability simply means thatat some point, therock shear strength ortensile strength has
been exceeded, and its severity ranges from

to collapse. pressure Pwduring all thethree phases. Although negligible good drilling andcompletion plans include rock

mechanicalanalysis, in any field operation, unforeseen instabilities may still arise, and the timing of such events mayalsodictate themanner in which thestability analysis is performed.

or tensile or a mixture of both. Major wellbore


tensile failure.

Rockmechanically, themodes of failureareshear

collapse problems areshear-failure induced, butin case of loss of drilling circulation, themechanism is

1067

Table1. DrillingProblems andSuspected Modes of


Failure

IN-SlTU

Stresses

Observed Problem

Failure Mode

Breakouts

Shear &/or Tensile

Doglegs Stuckpipe Tight spots Ledges Collapse Frequent reaming Largecuttings


Loss of mud volume

Shear &/or Tensile Shear Shear Shear Shear Shear &/or Tensile Shear&/or Tensile
Tensile

Shale E, v, C, and alpha ]


Reservoir E, v, C, and alpha I
Figure 1. Generalized wellbore trajectory

Lossof mudpressure Tensile


2.2 Completionandstimulation

Completion and stimulation engineers have to


ensure that the reservoir can be connected to the

well via perforations,and if required, hydraulic fracture(s) canbe propagated. This operation could fail if the rock adjacentto the cementedcasingis
non-elastic. There are also cases where the

recommendations to be relayed within hours or a few days. The analysthas to decidewhich logs,

bottomhole treatingpressures mustbe minimizedin order to prevent chemical reservoir formation damage. 2.3 Flow tests, production anddepletion Prior to full production,downhole tests that are performed include open-hole logging, fluid sampling, build-up, drawdown, injection, and deliverability tests. It is not unusual to induce failure and/orcollapseduring a well testingphase. An aim of the stability analysis should be to maximize hydrocarbon productionand minimize ground/sandcontrol measuresStability analysis shouldinclude fluid flow and geometricalfactors which are functions of the completion method: open-hole or bare-foot completion, perforated behind cementedcasing, packed with liners or screens, and hydraulically fractured. As the hydrocarbonsare depleted, the drainage region compacts and may fail, compounding the problem of solidsproduction.

cores, seismic data,offset-weB, anddrillingrecords are relevant,if they exist at all, and then gleanout from the availableinformation the necessary input such as elastic modulus, failure strength
xnput data and the need for quick answersmay justify the use of speedy, conventionallinearlyelastic methods.

.parameters, andinsitu stresses. Thelackof quality

On the other hand,in a rigorousfield pre-drilling study,as in multi-well development planningof a 100-million+ barrel oil field, the analysthas the benefitof weeksof lead-time,goodcoredata,pilot hole 'engineering' well data, and a staff of technologists and consultants. An example of an integrated field studyis in the development of the Cusianafield (Columbia), an active thrust-faulting environment (Last andMcLean 1995). Similarly,in a post-mortem failure analysis, applicable toolsvary from the simple to sophisticated 3D. Backcalibration of a basic linear-elastic method is

usually the firststep.


3 STABILITY ANALYSIS METHODS

2.4 Timing,urgency andscheduling


In an operatingenvironment,the need for results from a stability analysis may also be described according to expediency: 1. short-term, immediate, andurgent 2. pre-drilling field study. 3. post-mortem failure analysis. A short-term analysis is usuallyin response to an immediate need to cope with an unexpected, existing or impendinginstability, such as stuck

pipe,lostcirculafion, or fightreaming in a well in progress. An urgent caserequires fastreslonse and

Confronted with various types and severity of instabilities, productionobjectives, expediency, and schedulingconstraints,an operationstechnologist has a choice of analytical tools. Recentreviews of modeling technology, wellbore stability, and drilling advances are given by McLean and Addis (1990), Charlez (1994), and Payneet al (1995). The significant advances are in rock mass characterization, computation, modeling, monitoring, and logging tool developments. The availabilityof desktop andlap-topcomputers (PCs) with programsfor wellbore stress-strain analysis have contributedto wider field applicationsand acceptance of rock mechanicalmodels.Numerical
codes such as finite elements, disllacement

1068

discontinuity elements or boundary elements, which


usedto residein mainframecomputers, can now be executedin PC's. For any borehole of arbitrary

Table2. Commonly UsedComputation Methods Type Comp. Method Output


Full 3D Linear elastic Unix PC Numl. Anal. Realistic Conservative

trajectory, Bradley(1979), Roegiers andDetournay (1988), Aadnoy and Chenevert(1987), Last and McLean (1995) among others, have given expressions for thestresses anddisplacements at the borehole wall, assuming linear elasticity and isotropy, and theseare not presented herefor the sakeof brevity.Figure 1 illustrates the general case of an ERD and the major input requirements in a stabilityanalysis which includeinsitustresses (Sv, Shm, Shmin)' (Pt)' wellborepressure (Pw),cohesive strength (C), andintergranular frictional angle Although a well traversesmultiple layers of varying properties andporepressures (Fig. 1), a full three-dimensional model of the borehole alongits entire lengthis not alwaysnecessary because only the problematic formationneedbe analyzed.And
even if a full 3D numerical model can be created,

Poroelastic

PC

Anal.

Conservative

Elastoplastic Perfectly PC

Anal.
Numl.

Less

plastic
Strain Unix

conservative
Least

softening
Chemical effects
Stress

conservative
Unix
PC

Anal.
Numl.

Prototypes only
Less

dependent

conservative

Not shownin the table are specialcasessuch as


thermo-elastic or viscoelastic codes.

input data requirements would be very intensive. Someof thesewellborestress-strain analysis codes are pseudo-3D programs - a smallersubset of full threedimensional elasto-plastic wellboremodels. A pseudo three-dimensional (P3D or pseudo-3D) type of analysis is a plane-strain methodof estimating stressesin the vicinity of the wellbore in an anisotropic medium.Thus,the P3D models focus only on a given cross-section normalto the bore
axis, and assume that there are no strains

3.1 Applicability and Resurgence of Linear-Elastic


Models

Linear elastic-limit

methods assume that the onset

(displacements) alongthisborehole axis.With P3D (or plane-strain codes), the borehole stability problemis one of computingthe inducednearwellbore stresses which exceedrock strength.The solution to this problem could be performed analytically or numerically, in mainframes(e.g. UNIX) or desktops(e.g. PC), dependingon the additionalassumptions aboutrock behaviorunder stress. Rock response to strainor stress falls into a

of instabilityand collapseis at the elasticlimit, the most conservative option.The applicabilityof this approachseems arguably limited to hard, brittle formations, may be justified in the following
circumstances:

1. High confiningstresses in deepboreholes


enhance linear elastic behavior.

2. Scarcity of laboratory dataon non-linear stressstrainrelationships for shales, let alonein-situor


downhole data.

few categories suchas purely(linearly)elastic(LE), poroelastic (P-E), non-linear elasticandelastoplastic(E-P). Rock strength properties are commonlygiven in terms of the Mohr-Coulomb failure parameters cohesive strength C andintergranular frictionangle (alpha). When stresses exceedrock strength,the rock is described as non-elastic (or plastic).If pore fluid pressure gradient is included in the
calculationsof effective stresses, then the method is

3. Scarcity of datarelatinglaboratory-measured plasticstrains to observed shearfailuremodes. 4. Simplifiedinputandoutputwhichpromotes fastercomputation, parametric analysis, back calculation, andinterpretation. 5. Urgentneeds for results. For example,in an unexpected, urgentproblemof imminentcollapseof a deviated borehole wherethe input in-situ stress and strength propertiesare uncertain, quick but apparently conservative
answers obtained from a linear elastic model would

calledporoelastic. Table 2 compares thesevarious solutionschemes, numerical(Numl.) or Analytical (Anal.), andthe computing platform(Unix or PC). The linearlyelastic method is thebase case because it assumes that failure is equalto the elasticlimit, thus acquiring the label as 'conservative' or pessimistic(Charlez 1994). An elasto-plastic method implies that even after straining the boreholebeyondits elastic limit, the non-elastic region remainsintact and load-bearing,and not necessarily in a collapsed state,and thuslabeled
'less conservative'.

suffice.An engineer facedwith inputdatafrom logs only, may opt for a less data intensive code. Usually,field logs,poreand mud pressures are the only known variables, but with simpler elastic codes,it is fast and effortlessto iterativelybackcalculate the unknown variables (like insitu stress gradientand strength).The use of elastic-plastic modelsin a data-poorfield case would give less
conservative estimates but it also reduces the

engineer's marginof safety.Furthermore, the speed and flexibility of linearly-elastic(L-E) brittle rock modelshave enabledoperators to back-calibrate or performpost-mortems for data-poor field cases,an example of which is given by Zoback and Peska (1995).

1069

A novel approach considers mud and shale interactions in terms of chemical activity coefficients (Wanget al 1994),reduces the strength
of the shale with time, and still retains the elastic-

3.4 Elastic-plastic options


The most common shear failure criterion is the

limit constraints. This is applicablewhen dealing with reactiveshales. Non-linearelasticity in terms of confining-stress dependentYoung's modulus (Santarelli,Brown, and Maury, 1986) also gives more conservative answers andrelieson laboratory
triaxial data.

Mohr-Coulomb,and otherlesspopularcriteria are Von-Mises, Tresca and Drucker-Prager.McLean andAddis(1990) andCharlez(1994) compare these criteria with example applications. The classic Mohr-Coulombcriterion,hasemerged as the most popular because of the availability of data on cohesive and frictionalstrength properties of rocks, and its proven applicabilityin mine/tunneldesign.
The other criteria are more sensitive to the

3.2 Porepressure andfluid flow

Poroelasticity includesthe fluid or pore pressure distribution nearthe borehole in calculating strains and effectivestresses (Cui et al 1995, Detournay, Roegiers,and Cheng 1987, Mody and Hale 1993, among others). This implies that stability is a
function of time. Most of these formulations are still

underdevelopment andrequireextensive field and laboratory validation.During well-completion, it is desirable to perforate into an elasticrock, or a zone with minimalplasticstrain.If hydraulic fracturing is planned, formation breakdown and fracture propagation are the primary concern,and therefore the appropriate analysisis a poroelastic formulation (e.g. Weng 1993, Wang and Dusseault,1991). In a production mode, maximum flow rates can be obtained when the borehole or perforation is allowed to exceedthe elastic-limit (i.e. plastic). Thus,thesedesign stages call for elasto-plastic and poroelastic codes.
3.3 Anisotropy

magnitude of the intermediate in-situ stress and morepopularin numerical methods. A novelelastoplastic or EP (Drucker-Prager) approach includes the effect of water contenton shalestrength (Mody andHale, 1993) andwouldbe an appropriate model for reactiveshales. An exampleDrucker-Prager E-P application, performed in a PC spread-sheet is given by Lal andGuild (1995). Elastic-plastic models extend the stress-strain analysisbeyondthe elasticlimit. A mathematically
convenient formulation is to assume that no stresses

exceedthe elasticlimit, i.e., perfectplasticity,the mostcommonfeatureamongelasto-plasfic models. Analytical andnumericalsolutions canestimate the region aroundthe well that is non-elastic, which is looselyreferredto as plasticized, dilated,disturbed, damaged, or dis-aggregated. Variousinterpretations
of such zones could be the reaction fronts between

drilling mud and shale,spallingskin,breakouts, the damaged region around a perforation, and the ablatedring aroundopen-hole completions, to name
a few.

Like plasticity,transverse anisotropy or othotropic anisotropy is an attractive featurebut not a popular one, owing to lack of field data on coefficientsof anisotropy and the relative insensitivity of the computed outputto the coefficients of anisotropy.
There are information on elastic toodull,

particularlyin the seismicand dynamicdomains, and still less published data regarding static mechanicalproperties. In spite of the lack of field data, formulationsusing anisotropyare easier to implement in P3D thanin FEM. Anisotropyin cohesiveand frictional strength is moredifficultto implement. One usefulapplication of non-isotropicstrength includes the effects of bedding and joints or any plane of weakness.A formulationby Liao and Mear (1991), considers a tensionlessMohr-Coulomb joint, in addition to transverse anisotropy. This simulates a weak plane which could intersect the boreholeat any arbitrary angle.This is a convenient andpowerfulfeature in analyzingthe formationof doglegs, ledgesandloss of circulationalongpre-existingfrictional surfaces. Exampleapplications are in evaluating the impact of clay partings, bedding joints,or fault-crossings.

Again, the acceptability and significance of a plasticzone depends on the stageof life of the well. During drilling of an ERD, a plasticzonemay not be an acceptable risk sinceit is 'unsupported' for periodsof weeks.In this stage, the engineer usually opts for the conservative high-mud wt recommendations from a L-E model, just to ensure the well remains open. However, if the same well has another formation uphole which would breakdown(i.e. fracture)with suchmud pressure, then a less conservativeelasto-plasticmethod should be used. In the stimulation phase, the engineer may look at the plasticzoneas a zoneof
low stress, and it serves to lower the fracture

initiation pressure. And in the acid-stimulation and production stages, the dilatedor non-elastic zone is a region of enhanced permeability, and the developmentof a plastic zone is tolerable, if not highly desirable,up to point of massivesanding. For example,our North Seaoil-fieldexperience has demonstrated thata horizontal well canbe designed for maximizing production,as an elastic-plastic open-hole,especiallyif the input to the model are laboratory and log data (Ramos et al, 1994). McLellan and Wang (1994) givesan exampleof a poro-elastic-plastic application in an acidified
sandstone.

1070

3.5 Data availabilityandsensitivity Availability of, or lack of data can sometimes dictate the course of the stability analysis.The sensitivityof the resultsdepends mainly on two boundary-condition input parameters: insitustress and rock strength. In mostcases, the requiredinput dataare available.However,as a logicalalternative, the method of back-calculating 'unknown' boundaryconditions is becomingpopular,made feasibleby fasterPC's. For example,it takesonly a few seconds to iterativelychange (in the PC) input valuesfor in-situstresses, to comeup with a critical mudwt whichconforms to theknownfield collapse
failure conditions. The same back-calculation

pressures arelowerby 20%. On theotherhand,the well perpendicular to Shmax is more stable, breaking down at CDP=4300 psi and failing in shear with CDP of 710 psi. (The model is not suitable for investigating otherwellboreazimuths.)
The model also calculated the annular volume

whichis plasticized, usually from radialdistances of 0.05r to 1.Sr. The benefits of this kind of analysis arethe calculated magnitudes of the following: the extentof non-elastic'failed' region effectof permeability & pore-pressure gradient materialanisotropy, Kv:Kh effectof residual strength of theplastic zone
effect of non-linear material behavior

scheme is done to derive unknown strength parameters, sincethemajorgeometrical constraints, azimuth, deviation,are alreadyknown. This backcalculation option is commonly performed in interpreting leak-off testsin deviated wells, lossof circulation problems, breakouts, washouts, andeven laboratory hollow-cylinder testdata.

Its disadvantages arethe requirements for: a meshor grid of cellsor elements material properties for eachelement or cell the borehole alignedwith a principalstress mainframe-type computing power voluminous results andprintouts pre- andpost-dataprocessor

With the numericalapproach above,it was not possible to investigate arbitrarily oriented horizontal
wells. In actual field cases, the orientation of a

4 EXAMPLE

COMPARISON

To illustrate the differencesamong some of these

options,we comparea numericalfinite element


method (FEM) to an analytical wellbore model. Horizontalborehole stabilityin shales andreservoir rocks was analyzed with a numerical 2D FEM (Polillo and Crafton, 1991) under plane-strain elastic-plastic andporo-elasfic conditions. The FEM
results summarized in Table 3 will serve as a basis

horizontalor ERD well is the foremost production design criterion. If the 0-deg.azimuthis not stable, the alternative 90-degree azimuth may not be acceptable to the reservoirandfacilitiesengineers. Thus, we use a P3D code (one by Liao and Meat, 1991, amongothers)to determine otherhorizontal azimuthal options.From Figure 2, the analytical
solution shows that for horizontal wellbore

of comparison. The numericalmodel requires that the well is alignedwith one of the principalinsitu stresses, e.g. the azimuthfrom Shmaxis either0 or 90 degrees. The criticaldifferential pressure CDP is

the difference (P, - P.), whichis referred to as


overbalance if positive, or underbalance or drawdown when negative. The two valuesof CDP
are for shear failure and tensile failure.

The2D FEM analysis presented above implies that thehorizontal well maynotfeasible if parallel to Shmax because tensile fracturing initiates at420psi
overbalance, while shear failure starts at

overbalance pressures below 710 psi. With the poroelastic, penetrating fluid case,the breakdown
Table 3. Horizontal Well Linear Elastic and ElasticPlastic Critical Differential Pressures CDPs

(Polillo andCrafton 1991, Table 4)


Azim Shear Tensile Fluid Model

azimuthswithin 40 degrees of Sh .... either shear failure or breakdowndevelops,suggesting major difficultiesof drilling wells within +40 degreesof Sh. Thus,to reachthe desired target,the operator has the option of other more favorable well trajectories, i.e. +50 degrees from Shmin. In serious casesinvolving loss of well due to collapse, the effectsof Sv, Sh .... and Shmin are easilyanalyzed with the analyticalP3D The polar (stereographic) plot of Figure3 illustrates the effectson trajectory of an ERD well (Payneet al, 1995) of two typesof insitu stressregimes.The top quadrant(Fig. 3) is for a normal gravity-stress regime, similar to the previous example (Table 3), showing the minimum P,for shear stability, increasing with deviation and azimuthal proximityto Sh. The lowerquadrant is for a tectonic regionwhere Sh > Sv> Sh.. andimpliesthatthe verticalwell is the leaststable, the 30-degree deviation requires 13.5 ppgwhile the 85-degree deviation requires12.5 ppg. The examplesabove give azimuthsrelative to Sh, whoseorientation and magnitude are not usually known. In an actual field case, to quantify the bounds of insitu stresses and strengths, the logical recourse is to study thepilot-hole's caliper
and leak-off data. With this minimal amount of

deg.
0 90

CDP psi
>710 >710

CDP psi
415 4300

flow
none none

type
L-E L-E

0 90

>710 >710

320 3324

Darcy Darcy

E-P E-P
1 o71

information, a sensitivity analysis or backcalculation scheme can be performed in orderto create a field-validated modelor input-data set.

5.00 4.50

Max,

Hor,

Stress

4.00

3.50
3.00

2.50
2.00

Tensile Failure

Shear Failure
40 60 80
lOO

60

0, Normal .gravitatio

1.50 1.00
0.50 0.00

. 30

20

Well Azimuthfrom Shmax,degrees Figure 2. Critical Pressures for Horizontal


Well
Max. horzl tectonic stress

60
greater than vertical

90

Figure3. Polarstereographic plotshowing muddensity in ppg,asa function of trajectorywhereazimuthis


from Shmax, radius = sineof deviationfrom vertical.

Top quadrant= Normalstress regime Bot. quadrant = Thrust fault stress regime

Table4. Operating Problems, Strategy andTypeof Analysis

Typeof Problem
Geological hazards

Stress-handling Strategy Prevent circulation loss in adjacent normally


pressured sections

Preferred Model Poroelastic

Abnormally pressured
formations

Depleted & Fractured formations Prevent circulation loss in zone Reactive clays Optimize mudwt belowcollapse pressure
Operational Factors Risk of differential sticking
Kicks and risk of blow-outs

Elasto-plastic, Poroelastic Elasto-plastic, Chemical coupling


Linear elastic Poroelastic Poroelastic Poroelastic Poroelastic

Reduce mudwt w/o collapsing adjacent weaker


sections

Optimize mudwt belowcollapse pressure


Reducemud wt to the minimumpossible

Formation damage Poor Coringrecovery


Slow Penetration rate

Mud / clayinteraction
Difficulties while drilling

Optimize mudwt belowcollapse pressure Reduce mudwt to the minimumpossible Optimize mudwt belowcollapse pressure

Poroelastic,Chemical

coupling Linearelastic,Anisotropic
Linear elastic

Optimize mudwt Doglegs, Ledges, Largecuttings Increase mud wt Stuckpipe,Tight spots,


Breakouts

Collapse Bit-bailing, Frequent reaming


Differential sticks

Increase mud wt

Optimizemudwt

Reduce mudwt w/o collapsing adjacent weaker


sections

Elasto-plastic Elasto-plastic, Chemical coupling Elasto-plastic

Loss of mud circulation Well control & kicks

Reduce mudwt w/o collapsing adjacent weaker


formations

Elasto-plastic, Poroelastic Elasto-plastic, Poroelastic


Linear elastic, Poroelastic

Increase mudwt, plankill well


Back-calculate unknown variables

Leakofftestinterpretation

Completion andProduction
Stimulation difficulties Optimize treating pressure Well flow tests sanding, collapse Decrease drawdown pressure Sanding during production, Decrease drawdown pressure collapse
1072

Elasto-plastic, Poroelastic
Elasto-plastic, Poroelastic Elasto-plastic, Poroelastic

5 CONCLUSIONS

Basedon pastandrecent experience in dealing with field casesof wellboreinstabilityproblems, there are numerous toolsand options to help solvesuch problems. Table 4 summarizes commonly encountered difficulties during drilling and production, andpresents strategies andfeatures of thepreferred analytical toolsfor eachsituation. The suggestedapproach could be linearly elastic, poroelasticor elasto-plastic, depending on the perceived cause andproposed strategy. Thereis no singleintegrated model,analyticalnor numerical, applicable to all cases of instability.However, with additional research or field and laboratory data, each model shouldbe revisedto suit a particular field or well, even to the extent of creatingsemiempiricalformulations.
REFERENCES

McLellan,P.J. andWang,Y. 1994.Predicting the effectsof porepressure penetration on the extent of wellborestability:Application of a versatile poro-elastoplastic model. Rock Mech. in Petr.
Eng., Proc. EUROCK 1994, Delft, Rotterdam,
Balkema, 205-214.

Mody, F.K. and A. H. Hale, 1993 A borehole stability model to couple the mechanicsand chemistry of drilling fluid shale interaction, SPE/IADC 25728, Drilling Conf, Amsterdam, Feb. 23-25, pp. 473-489 Payne, M.L., B. Wilton and G. Ramos, 1995. Recentadvances and emerging technologies for ExtendedReachDrilling, SPE 29920, Proc.Intl. Mtg. Petr.Eng.,Beijing,291-308.
Polillo, A. and J.W. Crafton, 1991. Horizontal

wellborestabilityin low-permeability reservoirs, SPE 21864. Proc. Rocky Mtn. Reg. Mtg. Low
Perm Res., Denver, 585-610.

Ramos,G. K. Katahara,J. Grey, andD.J.W. Knox, 1994. Sandproduction in verticaland horizontal


wells in a friable sandstone formation, North

AadnoyB.S. andM.E. Chenevert, 1987.Stabilityof highly inclined boreholes.SPE Drilling Eng.,


364-374.

Sea. Rock Mech. in Petr. Eng., Proc. EUROCK


1994, Delft, Rotterdam, Ballcema,309-315.

Bradley, W.B. 1979Failureof inclined boreholes, J. Energy Resources Tech. Trans., ASME, 101:
232-239

Roegiers, J.C. & E. Detournay1988Considerations on failure initiation in inclined boreholesKey Questionsin Rock Mechanics, Cundal, P. et al eds. Proc. 29th U.S. Rock Mech. Syrup.,

Charlez,P.A.,1994. The impactof constitutive laws on wellbore stability:A generalreview,. Rock Mech. in Petr.Eng.,Proc.EUROCK 1994,Delft,
Rotterdam, Balkema, 239-249

Santarelli-F. J,ET Brown, andV. Maury1986


Analysis of borehole stresses using pressuredependent linear elasticity,.Int. J. Rock Mech. Sci., v.23, pp. 445-449 Wang, Chein Lee., C.H. Yew, and M.E. Chenevert 1994. The stabilityof deviated wellborein shale rocks in Computer Methods in Geomechanics, Rotterdam, Balkema,pp. 2195-2200 Wang, Y. and M.B. Dusseault 1991Boreholeyield and fracture initiation in poorly consolidated
rock strata- Part II Permeable media, Int. J. Rock
Mech. Min. Sci. 28:247-260

Cui, A.H.D. Cheng, D. Leshchinsky, Y. Abousleiman& J. C. Roegiers1995 Stability analysis of an inclinedborehole in an isotropic poroelastic mediumProc. 35th US SympRock
Mech., Rotterdam, Balkema.

Detournay, E., J.C.Roegiers, andA.D. Cheng1987 Some new examples of poroelasticeffects in


rock mechanics. Proc. 28th U.S. Rock Mech.

Symp.,Rotterdam, Balkema. Lal, M. and G.J. Guild, 1995. Very long reach drilling technologyapplicationin North Sea, SPE 29944, Proc.Intl. Mtg. Petr. Eng., Beijing,
587-592.

Weng, Xiaowei 1993. Fracture initiation and propagationin deviated wellbores, SPE 26597.
Proc. SPE Annual Tech. Conf., Houston. Zoback, M.D. and P. Peska, 1995. Insitu stressand

Last, N.C. and M.R. McLean, 1995. Assessing the impact of trajectory on wells drilled in an overthrust region, SPE #30465, Proc. SPE
Annual Techl. Conf., Dallas: 161-171.

rock strength in the GBRN/DOE Pathfinder well, south Eugene Island, Gulf of Mexico, J. Pet. Tech.July.

Liao, S.T andM. Mear, 1991.Stabilityof a highly inclined wellbore, Report to Univ. of TexasAustin Solids Production Consortium, Austin,
Texas.

McLean, M.R. and M. A. Addis 1990 Wellbore

stabilityanalysis: A review of currentmethods analysis and their field application,Proc. SPEHADC 19941, Drlng Conf., Houston,261274

1073

You might also like