Mechanical properties of North Sea Tertiary mudrocks investigations by triaxial ... · PDF file · 2018-02-09Tertiary mudrocks" investigations by triaxial testing of...

Clay Minerals (1998) 33, 171-183

Mechanical properties of North Sea Tertiary mudrocks" investigations by

triaxial testing of side-wall cores

L. W E N S A A S l , P . A A G A A R D , T . B E R R E * AND E. R O A L D S E T t

Department of Geology, University of Oslo, P.O. Box 1047 B lindern, N-0316 Oslo, *Norwegian Geotechnical Institute, P.O. Box 3930 Ullevhl Hageby, N-0806 Oslo, and tDepartment of Geology and Mineral Resources Engineering,

Norwegian University of Science and Technology (NTNU), N-7034 Trondheim, Norway

(Received 28 August 1996; revised 2 August 1997)

A B S T R A C T : In the North Sea Tertiary section, wellbore instability problems are frequently reported in Palaeocene-Early Oligocene smectite-rich mudrocks. Analysis of the mechanical properties of these Tertiary mudrocks is generally hampered by the lack of suitable core material. This study represents an attempt to study the geomechanical behaviour of mudrocks by triaxial tests of side-wall cores obtained from the borehole wall. The tests performed include measuring the changes in pore pressure during shearing and undrained shear strength in specimens initially consolidated to in situ effective stress levels. The coefficients of permeability (kf), estimated from the consolidation time behaviour range from 2.6 x 10 - n to 2.4 x 10 -12 m/s. The tested cores behaved like slightly overconsolidated to normally consolidated materials with an initial near constant volume (elastic behaviour) for low deviatoric load followed by an increasingly contractant behaviour approaching failure. Compared with results from onshore analogues, the strength properties of the investigated mudrocks appear to be related to their content of expandable clay minerals. A wellbore stability chart to forecast adequate drilling fluid pressures for future wells has been developed by the use of linear (Mohr-Coulomb) failure criteria based on the peak strength data. It is demonstrated that side-walt cores can provide satisfactory test materials for rock mechanical analysis, and their use may serve to improve our knowledge of the rock mechanical behaviour of typically troublesome mudrocks for which no conventional cores are available.

Drilling problems caused by compressive or tensile failure o f the borehole wall often result in substantial loss of drilling time and increases in rig time costs. Poor hole conditions (tight hole, hole enlargement, hole collapse, hole fill etc.) may create problems in running the casings and reduce the quality of petrophysical log data. Drilling and borehole stability problems are related both to chemical and mechanical processes. Chemically inf luenced swell ing of clayey rocks can be reduced by the use of oil-based drilling fluids or by modifying the chemical activity of water-based

1 Present address: Statoil Research Centre, Postuttak, 7005 Trondheim, Norway.

drilling fluids. The latter includes the use of high- concentration KC1 water-based drilling muds (e.g. Steiger, 1982; Chenevert, 1989), or the use of various water soluble semi-synthetic or synthetic polymers (e.g. Bruton & McLaurine, 1993). The various inhibitor systems have generally been proven to reduce wellbore instability. Problems may still be experienced even with the most inhibitive drilling fluids, suggesting a mechanical nature of the instability problems.

The main objective of wellbore stability analysis is to determine the critical upper and lower drilling- fluid pressure limits ('drillability window') to avoid compressive and tensile failure around the borehole wall. Ideally, borehole stability should be handled as a combined approach including both chemical

�9 1998 The Mineralogical Society

172 L. Wensaas et al.

and mechanical processes. Physicochemical processes include the interchange of fluid and ions between the drilling fluid and the surrounding rock which may modify both the stress conditions and the mechanical properties of the mudrocks (Bol et al., 1994; Mody & Hale, 1993). The transport processes (e.g. osmotic flow, diffusion, hydraulic flow) between the formation and the drilling fluid are complex. Present conceptual models are weakened by their qualitative nature, where several processes have to be calibrated empirically (Warpinsky, 1992).

In the North Sea Tertiary section, instability problems are frequently reported in Palaeocene- Early Oligocene smectite-rich mudrocks, however, the necessary input data for modelling of physico- chemical processes in typical heterogeneous North Sea Tertiary mudrocks are not yet available. A better understanding of the mechanical properties of these rocks has become increasingly more important due to the extensive drilling of deviated wells. But due to the lack of adequate core material, very few rock mechanical tests on North Sea Tertiary mudrocks have been undertaken and even fewer are reported in the literature.

The main objectives of the paper are twofold: (1) to determine whether carefully selected side-wall cores can represent appropriate test material for rock mechanical testing, and (2) to evaluate whether calculated wellbore stability from the obtained peak strength data and a simple linear (Mohr-Coulomb) mechanical approach can be correlated with the observed drilling problems in the studied mudrock sequence. The test results are compared with published results from triaxial tests of onshore mudrocks (Weald Clay, London Clay and Fullers Earth) and the rock strength behaviour has been considered in relation to mineralogical composition.

G E O L O G Y OF T H E S T U D Y A R E A

Well 25/7-2 is located close to the eastern boundary of the southern Viking Graben on the north-western flank of the Utsira High (Fig. 1). The studied Palaeogene sequence (c. 1200-2600 mKB) consists of large Palaeocene submarine-fan complexes (Ty and Heimdal Formations) interfingering with marine hemipelagic mudrocks (Lista and Sele Formations) derived from deltaic progradation eastward from the Shetland Platform into the Viking Graben (Isaksen & Tonstad, 1989). Late Palaeocene to earliest Eocene deposits were

dominated by ash-fall deposits (Balder Formation) derived from the volcanic activity associated with the opening stage of the Norwegian Greenland Sea. Renewed Eocene subsidence formed deep depocen- tres along the Viking Graben (Jordt et al., 1995) which were filled gradually by Middle-Late Eocene hemipelagic muds mixed with minor sands (Lower Hordaland Group) gradually changing to Early Oligocene interbedded silty/sandy mudrocks and muddy diatomites towards the top of the studied interval (Middle-Upper Hordaland Group).

S A M P L I N G A N D A N A L Y T I C A L T E C H N I Q U E S

The investigated side-wall cores were sampled by shooting a small hollow bullet into the borehole wall. A small cylindrical piece of rock with an axis normal to the well is thereby retrieved using a wireline tool. Side-wall cores selected for triaxial strength tests (Fig. lb) were cleaned gently, wrapped into cellophane and covered in wax (offshore) to preserve their water content. The samples were stored in glass jars at a temperature of -4~ until the testing took place. Opening, sample description, and sample preparation (trimming) were performed at high humidity (>98%) condi- tions. The tests were performed using a triaxial cell (70 MPa cell pressure capacity) modified as part of this study to handle cores with a diameter of 25 mm (by T. Berre, Norwegian Geotechnical Institute). The test cell had both external and internal load transducers, two axial strain measuring devices, two radial strain measuring callipers, units for measure- ment and control of cell pressure, and a unit for volume change measurement and for measuring and control of back pressure. During the experiments, axial load, confining pressure, radial and axial displacement, pore pressure and back pressure, were continuously monitored and recorded with an automatic data-acquisition system.

The triaxial tests were run as isotropically consolidated undrained (CIU) tests, equivalent to published triaxial tests on shales (e.g. Steiger & Leung, 1992) and onshore analogous tests (Nakken et al., 1989; Marsden et al., 1989, 1992). The CIU tests are divided into two parts: (1) drained isotropic consolidation to the level of the calculated in s i tu stress (Table 1); and (2) undrained (deviatoric compression) with a recording of the pressure build-up. During the isotropic consolida- tion the samples were drained from both ends

Mechanical properties of mudrock~ 1 7 3

e ~

D D D r l D D O O O O 0 [ ] D D [ ] 13 ~ n ~ ~ [ ] O n O 0 [ ] [ ] D [ ] [ ] [ ]

;;44 ,I, ,~ ;

I # , 1 u~ I I LU - - - - . INI~ I E: L~I ~ 191

�9 �9 I " I - I ' 1

o o o o o o o o o o

(83~u) q),dao

0 0 '''~

e~ e,i'~ , ~ r ' - ,.~

I

z C ~ ~

[ ] : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : , :::::::::::::::::::::::::::::::::::::::::: :: :::::::::::::::::::::::::::::::::::::

~1 [~ :m:m:m:m:::m:::r :: ::: :!: K:! :!:1%!:K:!%1: i: i ~ [i:!:!:!:!:!:!:i:!:!:!:!:i:i:!:!:!:!:!:i:i:!:!:i:!:!:i:i:!:i:!:~ ::::~ ~! ~#~i e~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i?i~i~i<~i~h h ?i~i(~i~iiiiiiiiiiiiiiiiiiiiiiiiii~ 1 ~ [ ~ |:F~<ilFiiiiiii%ii! i

1 t ~ ! : ::: . . . . . . . ~ ~:::::::::::::~::::::::::::::::::::: ~::<:::::::: ::::::::::g::::::::::~:~::::~:::::::::::::: :::::: ::: ::::::::::::::::::::::::::::::::::::::::::::::::::: :: : : : : : : : ::::::::::::::::::::::::::::::::::

:i~i~i~ ........... ~i~i~i~ ~ ~;;;;~;m~i~i~i~i~!~i~i~i~L>~i~i~K~i~ ......... ~ ! ~ : ~ : ; ~ : <:~;;~m~m~m~:~;~ ......... ~=::~:~=::~:~:~:~:~::~:~:~:~:~:~:: ........... ~:::~:~::::i:

~ .... ~ ~iiiiiiiii~i~i,i,i~i~i~i,~,i,i~i ~ ~::~:::~:, ............. ~,~,~,~,~,~,~,~,~,~, ~<:~i~i~i~%i~i~i~i~i:i~i~::~i~i~i~<:~iiiiiiii~i~/:i~ii:i~iii

~ ~:~:iiii ~ i i L ~ii i i~i i i i i i i~i~ii i i i i i i i i~i~i C~ !:i:i : ! i i i i ! i ~ i L : . . . . . . . . . . q i ! i ! i i i i i i i i i i i i i i i i i i i ! i i ............... i i i i i iL !~:::::::ii:iiii~ ~ ~:Ui~iiiiiiiiiii~ii~i~:iii~iii!~i~ii~i~<~ii~i~i~i~iiii~i~iiiiiiiiiii ijiiiii!i~ii!i!i~iiii!!iiii

::u<::3::i:;i:i:iiil :::: ~ ~ :: : : -:::::::!: ~ iiiiiii:iiiiiiii::iijjiiiiiiiiiiiiiii:iiiiiiiiiiii:iiiiiiiii:::::i:: : ii::::i:::: NO i k.._, ..... ...................

r-u.K. ~ ....

! ~ ~ ~ ~

I ~ ~ ~ ~_/14 ~ l O l ~ l ~ :~:::: :#:~:~:~ ~ ' ~ - : ; . . . . . ~ . '

174 L. Wensaas ct al.

r~

_ ~

~g

.~.~ : 8 o o

i

o

0 o

o

o

0

o

o

0

o

~ d d d d

o

o

o

~ ' ~

o o

0 0 0 , ~

c~ r~ " ~ ~ 0

Mechanical properties of mudrocks 175

against a constant backpressure (6 MPa). Side drains to assist the drainage were not used. The consolidation serves three purposes: to rescind the effects of sample disturbance (e.g. micro-cracks): to consolidate the sample to saturation: and to dissolve any air inside the pore-space (Steiger & Leung, 1988, 1992). All specimens were saturated with weak brine (52.0 g/1 KC1) at the consolidation stage, in order to prevent swelling during consolida- tion. The degree of initial saturation (Swi) was proved by performing a 'B' value test (Bishop, 1973). The strain rate used during the shearing phase was calculated using the results of the consolidation data, according to the recommended procedure given by the British Standards Institution, London (BS 1377: Part 8: 1990). Peak strength was generally achieved after a testing time (tf) of 14-40 h of deviatoric loading. All tests were run at room temperature. An overview of the sample data and the testing programme is given in Table 1. To obtain some broader information about the test material, sample 8 (1540 mKB) was consolidated to only 10% of its in situ stress level and sample 9 (1361 mKB) was run as an extension test. The permeability of the specimens was estimated from the consolidation behaviour according to the method of Head (1986).

The mineralogy of the bulk samples and also of the <2 ~tm fractions was determined by X-ray diffraction (XRD) analysis of unoriented (powder) mounts and oriented specimens (inverted Millipore slides), respectively. The fine fractions were scanned after the following treatments: (1) Mg- saturation; air dried; (2) ethylene glycolation; and (3) preheating (550~ h). All analyses were performed on a Phillips PM 1700/1710 diffract- ometer with Cu-Kct radiat ion (1.54060 A), connected to a PDP 11/53 computer system with a Phillips APD software.

R E S U L T S

Physical properties and mineralogy

All results of the geomechanical testing are summarized in Table 1. The initial water content (wi) ranges between approximately 38 and 20 wt%, decreasing with increasing burial depth. The corresponding calculated initial (unstressed) porosity (ni) ranges between 57-36 vol%. The initial degree of saturation (Swi) is between 83-95%. The mineralogy of drill bit cuttings and

side-wall cores is presented elsewhere (Tyridal, 1994) and only a summary for the tested specimens is included here. The XRD analysis confirmed the mudrocks to be homogeneous containing the following: 71-82% (average 75%) clay, 5-14% (average 7%) quartz, 0 - 4 % (average 2%) K- feldspar, 2 -7% (average 4%) plagioclase, 1-10% (average 6%) pyrite, 3 - 8 % (average 5%) calcite, and 0 -2% dolomite and siderite. The fine fraction (<2 ~tm) contained between 35-65% (average 46%) smectite and mixed-layer I-S, 10-30% (average 16%) illite, 0 -50% kaolinite and 0-25% chlorite.

Consolidation

During the drained isotropic consolidation, an equal all-round effective stress produces a change in sample volume, and for fully saturated samples, this volume change is directly related to the drained volume of water and hence the change in porosity (Table 1). High volumetric strains (EvoD and octahedral strains (Coot) (14-23%) during consoli- dation indicate that the specimens have been affected by relaxation prior to analysis (e.g. swelling/microfracturing). Values for drained bulk modulus (B) for the isotropic consolidation were obtained at maximum effective confining pressure. The coefficients of permeability (kf) estimated from the consolidation time behaviour (Table 1) range from 2.6 x 10 - l l to 2.4 x 10 -12 m/s (2.7 x 10 -6 to 2.5 x 10 -7 Darcy), which is within the micro- nannodarcy range reported by Steiger & Leung (1990). Similar permeabilities are also reported from fractured shales (10-s-10 -s Darcy), whose permeability of the intact rock is several orders of magnitude lower (Chenevert & Sharma, 1991). Thus, if the measured high strains during consolida- tion reflect closing of micro-cracks created by previous sample relaxation, the estimated perme- abilities may have been overestimated. However, large consolidation strains and subsequent large reductions in porosity have also been reported from CIU triaxial tests of onshore analogous clays (London Clay and Fullers Earth) at offshore stress conditions (Marsden et al., 1992). Hence, large strains during consolidation seem to be typical for soft, smectite-rich mudrocks.

Shearing

Figure 2 shows the deviatoric stress-strain response and the change in pore-fluid pressure (u)

176

en

O

L. Wensaas et al.

(edl~l) e J n s s e J d e J 0 d 0

, i 0 i , �9 i i , �9 . i . �9 , i , , . i , , o i

.,,:1- to

, ~

N O I S S ~ d d l A I O D

o ~, ~, i �9 �9 ' �9 , �9 | �9 �9 �9

A

q \

/ J

N O I S N ~ I X ~

~ / / I I

. - ~ ~ ; ~

N O I S S ~ P d d l A I O D

( \ ( ( I

O *~ " t -

O

"r"

( e d N ) sseJ1S o!JOle!AeO

0 . . . . . 0 . . .

~. r

O3 \ -

c -

J, o . m

'~ 0

. i I

N O I S N ~ t l X ~ / ' . o ~

o.

\ ": t'-- �9 . . m

..i.--, ~ : (/-)

/ 0

�9 0 , - - . . . .

~ o = ~

. =o ~

o g ~.= ~ o

0

'L ~0 ~3

- I 2; tO "-" - .= "~

m ~, ~ . ~

Od ~ o~ u

o

~. " ~

,

~ , . ~

Mechanical properties of mudrocks 177

during the undrained shearing phase. The stress- strain behaviour of all samples during deviatoric loading is basically the same; any difference is in magnitude rather than nature. At low deviatoric stresses (q = Ol - or3) and low strains the specimens show a close to linear behaviour. The curves show a gradual transition into an upward convex shape with a plateau of near constant stress in the post-failure region. Such overall stress-strain performance is typical of plastic/ductile behaviour. The undrained shear strength is reached after low (2.5-3.5%) axial strain (1.4-2.0% lateral strain).

The pore pressure in undrained compression tests (Fig. 2) continued to show a small but steady increase in the post-failure region. The extension test showed a minimum pore pressure at about the peak stress stage, from which it increased towards zero during post-failure deformation. Values of the secant Young's modulus (E�89 and Poisson's ratio (~tlf) from deviatoric stress vs. axial strain curves obtained at half the peak deviatoric stress are given in Table 1. The samples still appear to be within the linear range at half their peak load. Young's modulus (stiffness) represents the slope of the change in deviatoric stress against change in axial strain, while the dimensionless Poisson's ratio (~t) is the negative ratio of the horizontal and vertical strain for the given stress range. Young's modulus increases with increasing effective mean stress, while Poisson's ratio shows no variation with stress. The undrained shear modulus (G) represents the samples' resistance against shear deformation (modulus of rigidity). The initial values (Gi) measured during undrained shearing phase, vary from 0.2 to 2.1 GPa (Table 1). Despite the change from an initially stiff to a gradually more soft behaviour approaching failure, the material largely retains its ability to carry a load in the post-failure situation. Furthermore, high shear stresses in the post-failure region suggest that the frictional resistance along the failure surface is high for the entire stress range studied (i.e. for stress conditions corresponding to typical wellbore situations).

Stress paths and peak strength

Figure 3a illustrates the stress paths in the q-p' diagram for the triaxial compression tests. The failure stress (undrained strength) is taken as the maximum principal stress difference qf = ( ~ 1 - - O '3 ) f

(Table 1), or the minimum stress difference in the extension test (positive deviatoric stress values

correspond to shortening of the sample). As the pore pressure increases throughout the tests (Fig. 2b), the stress paths get curved to the left (Fig. 3a). Pore pressure increases in undrained tests because the specimen would reduce its volume due to the contractant behaviour, but contractancy is prevented by the test configuration. Initially, the mean effective stress p ' appears constant (i.e. path close to vertical) as expected for undrained constant volume shearing of an isotropic elastic material. Based on the calculated pore pressure parameter A (Skempton, 1954) and dilatancy (D) (Janbu, t985) at failure (Af, Dr in Table 1), the mudrocks tested in compression behave like normally consolidated to lightly overconsolidated materials (Af between 0.56 and 0.79) with an initial near constant volume and elastic behaviour for low deviatoric load followed by an increasingly contractant behaviour (De < O) approaching failure.

Values for the cohesion e' (1.1 MPa) and the angle of friction qb' (17 ~ were obtained by fitting the Mohr-Coulomb criterion to a linear regression of the peak strength data in the p'-q diagram (Fig. 3b). A non-linear failure criterion proposed by Marsden et at. (1989) is shown for comparison.

A P P L I C A T I O N TO W E L L B O R E S T A B I L I T Y

The following discussion is based on the assump- tion that the wellbore instability in the studied well is purely mechanical in nature. Evaluation of mechanical stability of the wellbore consists of two parts; calculation of the stress state around the borehole and consideration of a failure criterion. The vertical overburden stress (Fig. 4a) was determined from integration of the rig bulk density measurements on drill bit cuttings. The 'best empirical solution' of the minimum horizontal stress (sh) was taken as the 'minimum envelope' of leak-off-test (LOT) data from 21 nearby wells (Fig. 4a). No borehole breakout data were available for determination of horizontal stress anisotropy and the two horizontal stresses (O" h and ~rt) were assumed to be similar in magnitude. The pore- fluid pressure estimated from wire line logs and drilling parameters indicate a maximum absolute overpressure in the investigated Tertiary section of -4.7 MPa.

Failure normally implies the condition in which the specimen can sustain no further increase in stress. The linear Mohr Coulomb criterion (Fig. 3b)

178

. Q

oO

O

L. Wensaas et al.

I i

I I I I I I ~ O ~r) 0 i ~

(BclIN) b sseJls o!Jole.~ea

O

Oo~OOO.O ~ e []

~ ~ I~

(D c:) v

,.._: v co

I I I I I I I i

(odiN) b ssOJ:lS oiJo:lg.U~e(]

(:2)

M~

O

0 c'O

CO C'J

" ,0

C~I

a o

, , 0

"::::i"

O4

0

',,0

C~

Q..

u') u)

u')

._>_ U

r 0.) t'-

0.)

:>,.~

b ' 7

0 -}-

II ,=i

u~

u~

o ~

...-.. o.~

0.) = ~., .> -~

E

~u

6 ~

e~

t~

Mechanical properties of mudrocks

._= = ~ i

-~S o

,~ - - i 1 ~ ..'~i::iii: ' o , o . I ,.",~iiiiiiiiiii~ I /

. o o ., :::::::::::::::::::::::::::::::::::::::::::::: / / \ \ v

v . ::::::::::::::::::::::: �9 \~ . v o o ~.. .............................................. k l , ~ ~ , , ~. ~ . ::::::::::::::::::::::::::::::::::::::::::::::::::: v . ,~,~ _~ �9 I , ~ , " ::::::::::::::::::::::::::::::::::::::::::::::::::

{~. ~ ~./ �9 ::::::::::::::::::::: i,a .- ~$!:!:~:i:!:i:i:i:i::~i:i:::~ �9 t - }

1" "" x ~ N ~ z~::~::~i~::~ii:::~ ~ ~ : , . ~.~....!~.~,,

I I "' "':::::::~ ~ I ~ " / ,~i::iiiiiiii::~ ~*~ . . . ~ ' ~ " ~

. / , . , . / ~ . ~ w I

o o . . . . ~ 1 ; . . . . ; . . . . ; t ; . . . . o o _ 0 0 0 0 0

(8)t'") q;dea

I O . i = , I ' ~ . / . . , ' F ~~

I ,-,'~" ._" ,.-'" ~ " / ~ I ~.- _..,- : , > / ~_ o ~ _ .., ' i1 . ,

;"~" - : ' . .gr3 ~ -

I ~ "-"- ' . - - '~"" F . / I f

,a ~ 1 7 6

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

( 8 N t u ) q l d e o

. ~ o - ~ " ~ o

o ~ ~ ~

~ ' 0 . 0 0 J j = ~j

o _ ~ ~

~ ' = ~ ~ e~ ~ ~ . ~

= - ~ c ~ ' ~

- ~ ~

, ~ 0 ~.~ ,

m

~.~ . .

179

180 L. Wensaas et al.

incorporates the maximum (g~) and minimum (or3) principal stresses, and because of its simplicity it has been extensively used in the petroleum industry (Aadnoy, 1988; Aadnoy & Chenevert, 1987; McLean & Addis, 1990a,b). A failure criterion defines the critical state boundary line between possible and impossible stress states (Fig. 3b). Major shear deformation will only occur if the material is loaded in such a manner that it attains this critical state. Here plastic shear deformation may continue without change in volume (Fj~er et

al., 1992). In assuming a linear relationship, the error appears negligible, at least for stress conditions corresponding to typical wellbore situa- tions (<14 MPa; McLean & Addis, 1990a).

Failure of the borehole wall is induced by drilling fluid pressure which is either too high (tensile failure) or too low (compressive shear failure). In this paper, calculations of the upper and lower limits of drilling-fluid pressures were performed according to Fjmr et al. (1992) (eqns. 1 to 4) assuming a vertical hole with impermeable walls and a poro- elastic material behaviour. Provided that isotropic stress conditions prevail, the stresses around a borehole can be expressed as (Fj~er et aL, 1992):

err =Pw (I0 = 2Crh -- Pw (1) (~z = ~ v

where c~0, clz, and ~r represent the tangential, vertical and radial stresses, respectively, around the borehole, and Pw the pressure of the drilling fluid (well pressure). With the relative magnitudes of the stresses being; c~0 < ~z < clr, the lower critical dril l ing-fluid pressure (i.e. compressive shear failure) becomes:

Pw ~ [2~h + ~xpf(tan2~ -- 1) -- Co]/(tan2~ + l) (2)

where cz is the Blot's constant (assumed = 1), Pr is the formation fluid pressure, 13 the angle of the failure plane to the major principle stress (c~e), and where the uniaxial compressive strength Co is defined as:

Co = 2c 'tan~3 = 2c 'tan[(~b/2) + 45 ~ (3)

The upper critical drilling-fluid pressure (i.e. tensile failure) becomes:

Pw ~> 2crh - cxpf + To (4)

where To is the tensile strength, which due to small fractures and cracks at the borehole wall, is assumed to be zero.

The results of the stability calculations are given in Fig. 4. Although stability models based on linear behaviour, typically tend to provide a conservative estimate of the 'drillability window', there is a good agreement between the results of the stability calculations and the reported hole problems between 1000 and 1900 mKB (e.g. twist off, tight hole, pipe stuck, hole pack off). In this interval the drilling fluid pressure is slightly lower than the line defining the calculated lower pressure l imit (Fig. 4b). After an increase in the drilling fluid pressure at 1900 mKB it entered the stable region ( 'dril labili ty window') and hole stability was reported to improve. The results also imply that the drilling-fluid pressure can be further increased in the troublesome Tertiary interval without any major risk of tensile failure. Moreover, drilling with a low (1.07 g/cm 3) drilling-fluid weight down to 1500 mKB in a nearby well resulted in stuck pipe and the hole being packed off at 1133 mKB and then having to be side-tracked. After increasing the drilling fluid weight to a similar level as in the studied well (1.32 g/cm 3) the well was successfully drilled throughout the Tertiary interval.

D I S C U S S I O N

As eqns. 2 and 4 indicate, the calculated wellbore stability is very sensitive to the uncertainty in the input parameters such as field stresses and pore- fluid pressure. Pore-fluid pressures of 1.3 g/cm 3 MWE (mud weight equivalents) (12.8 kPaJm) are typically reported in the Lower Tertiary shales. Quantification of shale pore-pressures (by indirect methods) has been shown to be strongly influenced by l i thological and mineralogical variat ions (Wensaas et al., 1994), and these variations may introduce an uncertainty in the estimated shale pressures. Calculations assuming hydrostatic pres- sure conditions cause the lower critical pressure line to decrease (max. 5%) and the drilling fluid pressure tends to enter the 'drillability window'. This suggests that the reported drilling problems may essentially be driven by overpressuring.

The oh trend (Fig. 4a) is very similar to the trend ob ta ined from the Cent ra l Nor th Sea by Gaarenstroom (1993). Calculations with a higher oh trend as proposed by Breckels & van Eeklen (1982) or Kwakwa et al. (1991), cause the lower critical pressure line to increase. This will result in a conservative estimate relative to the field observations. The uncertainty related to the

Mechanical properties o f mudrocks 181

60

50

0 40 t3_

r 30

o

-~ 20

C3

]0

1)

2)

3)

j / \ SMECT,TECO.TE.T

I I I

0 10 20 30 40

Mean effective stress p' (MPa)

1) Weald Clay

2) Weald Clay

3) Weald Clay

4) SWC

5) London Clay

6) Fullers Earth

FIG. 5. A comparison with peak strength data from onshore analogues; (1), (2), (3) Weald Clay; smectite content 0-1% (Nakken et al., 1989; Marsden et al., 1989; 1992); (4) North Sea Tertiary mudrocks (this study); smectite content 40-60%; (5) London Clay; smectite content 35-45% (Marsden et al., 1992); and (6) Fullers Earth;

'pure' smectite (Marsden et al., 1992).

determination of the Oh trend appears to be more critical to the stability forecast than the uncertainty related to overpressure.

The undrained strength of shales and mudrocks is not a unique parameter, but depends upon several factors, where we regard the most important to be material type (e.g. mineralogy, texture, fabric, water content), present stress level, previous stress history, anisotropy and testing conditions. By comparing the side-wall core data with data from onshore analogues, a correlation between rock strength and smectite content could be anticipated (Fig. 5). The strength of the samples decreases with increasing smectite content, from Weald Clay with negligible smectite content to Fullers Earth with close to 100% smectite. Similar correlations have been reported by Shell and Exxon between low shale strength and high water contents (Mody & Hale, 1993), or high surface areas (Steiger & Leung, 1990; Ewy et al., 1994), both being typical properties of smectite-rich rocks. Moreover, from

CIU triaxial tests of various 'pure' clay mineral samples, Wang et al. (1980) found that the peak strength of montmorillonite (smectite) was about half of the strength of other clays at a given confining pressure.

By performing the test exclusively on a set of carefully selected, full recovery cores (without any visible anisotropy), we anticipate that any influence from sampling distortion (e.g. micro-fracturing, swelling) on the obtained rock strength data is kept to a minimum. Interestingly, both the recovery and quality of the side-wall cores were found to correlate broadly with variations in lithology and mineralogy. Low recovery cores are typically fractured and consist of silty and/or calcareous mudrocks, whereas full recovery cores, with no visible cracks, consist of fine-grained smectite-rich mudrocks. Mineral reactions which tend to strengthen the mudrocks, e.g. smectite diagenesis (>1800 mKB), carbonate cementation and Opal CT to quartz appears to be poorly represented in the

182 L. Wensaas et al.

selected test material (according to unpublished pe t rographica l data). This impl ies that the instability calculations based on the selected core specimens are probably only representative of the most fine-grained and smectite-rich formations (i.e. 'worst case scenario').

C O N C L U S I O N S

(1) Triaxial tests of side-wall cores have been shown to represent a suitable method for determina- tion of the mechanical properties and stress-strain behaviour o f weak, smecti te-r ich shales and mudrocks. No tests were performed on hard (cemented) or silty Tertiary shales.

(2) The rock mechanical properties of the Tertiary mudrocks from Well 25/7-2 identify them as weak rocks with low values of the strength parameters c ' (cohesion) and ~)' (angle of friction). Within the test conditions applied, the mudrocks exhibit a ductile (non-brittle) behaviour.

(3) The measured rock strength increases almost linearly with increasing confining stress.

(4) Compared with the results from onshore analogues, the strength properties of the investi- gated mudrocks are most likely to be related to their content of expandable clays.

(5) Wellbore stability calculations based on a linear peak strength criterion (Mohr-Coulomb) and estimated field parameters (stresses and pore-fluid pressure) were used to calculate the drilling-fluid pressure required to prevent compressional shear and tensile failure of the borehole wall. The predicted instability generally shows a good correlation with reported drilling problems. In forecasting, the upper and lower critical drilling fluid pressures appear to be very sensitive to variations in the assessment of the horizontal minimum stress trend. Overall, the stress conditions should be better defined because instability models are highly affected by the accuracy of field stress data.

ACKNOWLEDGMENTS

We would like to thank the participants in licence PL 103 (Conoco, Statoil and Amerada Hess) for allowing us to publish these results. The authors acknowledge the financial support provided by the Conoco Norway INC. project CNRD 25-6 on Tertiary Claystones on the Norwegian Shelf. This study is a contribution to the joint University of Oslo/University of Trondheim R&D Programme on Clays Claystones and Shales in

Petroleum Geology. Constructive criticism from V. Feeser and F. Madsen on an earlier version of the manuscript is much appreciated.

REFERENCES

Aadn~y B.S. ( 1988) Modelling of the stability of highly inclined boreholes in anisotropic rock formations. SPE Drilling Engineering 3, 259-268.

Aadnoy B.S. & Chenevert M.E. (1987) Stability of highly inclined boreholes. SPE Drilling Engineering 2, 364-374.

Bishop A.W. (1973) The influence of an undrained change in stress on the pore pressure in porous media of low compressibility. Geotechnique, 23, 435-442.

Bol G.M., Wong S.W., Davidson C.J. & Woodland D.C. (1994) Borehole stability in shales. SPE Drilling & Completion, 87-94.

Breckels I.M. & van Eekelen H.A.M. (1982) Relationship between horizontal stress and depth in sedimentary basins. J. Petr. Tech. September, 1982, 2191-2199.

Breton J.R. & McLaurine H.C. (1993) Modified poly- amino acid hydration suppressant proves successful in controlling reactive shales. SPE Paper 26327, Proc. SPE flnn. Tech. Conf Exh., Drilling, Houston, Texas.

Chenevert M.E. (1989) Glycerol mud additive provides shale stability. Oil Gas J. July 1989, 60-64.

Chenevert M.E. & Sharma A.K. 099 l ) Permeability and effective pore pressure in shales. Paper IADC/ SPE 21918.

Ewy R.T., Ross G.D., Gast M.R. & Steiger R.P. (1994) North Sea case histories of wellbore stability predictions for successful high-angle Nelson Field wells. IADC/SPE 27496 paper, 1994 IADC/SPE Drilling Conference, Dallas, 1994, 589-598.

Fj~er E., Holt R.M., Horsmd P., Raaen A.M. & Risnes R. (1992) Petroleum Related Rock Mechanics. Developments in Petroleum Science 33. Elsevier, Amsterdam.

Gaarenstroom L., Tromp R.A.J, de Jong M.C. & Brandenburg A.M. (1993) Overpressures in the Central North Sea: implications for trap integrity and drilling safety. Pp. 1305-1314 in: Petroleum Geology of the Northwest Europe: Proc. 4th Conf. (LR. Parker, editor). Geological Society, London.

Head K.H. (1986) Manual of Soil Laboratory Testing 3: Effective Stress Tests. Pentech Press, London.

Isaksen D. & Tonstad K. (1989) A revised Cretaceous and Tertiary Iithostratigraphic nomenclature for the Norwegian North Sea. Norwegian Petroleum Directorate Bulletin No. 5.

Janbu N. (1985) Soil models in offshore engineering, 1985. Rankine Lecture. Geotechnique, 35, 241-281.

Jordt H., Faleide J.I., Bjorlykke K. & lbrahim M.I.

Mechanical properties of mudrocks 183

(1995) Cenozoic sequence stratigraphy of the central and northern North Sea Basin: tectonic development, sediment distribution and provenance areas. Mar. Petrol. Geol. 12, 845-879.

Kwakwa K.A., Batchelor A.S. & Clarc R.C. (1991) Assessment of the mechanical stability of high-angle wells in block 22/11, Nelson Field discovery. SPE Drilling Engineering, 25-30.

Marsden J.R., Wu B., Hudson J.A. & Archer J.S. (1989) Investigations of peak rock strength behaviour for wellbore stability prediction. Pp. 753-760 in: Rock at Great Depth (V. Maury & D. Foumaintraux, editors). Balkema, Rotterdam.

Marsden J.R., Holt R.M., Nakken S.J. & Raaen A.M. (1992) Mechanical and petrophysical characteriza- tion of highly stressed mudstones. Pp. 51-56 in: Rock Characterization, Proc. ISRM Conf. Europe 92 (J.A. Hudson, editor). Chester, England.

McLean M.R. & Addis M.A. (1990a). Wellbore stability analysis: A review of current methods of analysis and their field application. IADC/SPE Paper 19941, IADC/SPE Drilling Conf., Houston, Texas, 1990.

McLean M.R. & Addis M.A. (1990b) Wellbore stability: The effect of strength criteria on mud weight recommendations. SPE Paper 19941, 65th Ann. Tech. Conf. Exhibition Soc. Petrol. Engineers. New Orleans, 1990.

Mody F.K. & Hale A.H. (1993) Borehole stability model to couple the mechanics and chemistry of drilling fluid shale interaction. SPE/IADC 25728 paper 1993 SPE/IADC Conf. Amsterdam, 1993, 473-490.

Nakken S.J., Christensen T.L., Marsden J.R. & Holt R.M. (1989) Mechanical behaviour of clays at high stress levels for wellbore stability applications. Pp. 141-148 in: Rock at Great Depth (V. Maury & D.

Foumaintraux, editors). Balkema, Rotterdam. Skempton A.W. (1954) The pore pressure coefficients A

and B. Geotechnique, 4, 143-147. Steiger R.P. (1982) Fundamentals and the use of

potassium/polymer drilling fluids to minimize dril- ling and completion problems associated with hydratable shales. JPT, 1661 - 1670.

Steiger R.P. & Leung P.K. (1988) Quantitative determi- nation of the mechanical properties of shales. SPE Paper 18024, SPE Ann. Tech. Conf. Exh., 1988.

Steiger R.P. & Leung P.K. (1990) Lecture: Predictions of wellbore stability and shale formations at great depths. Pp. 1209-1218 in: Rock at Great Depth (V. Maury & D. Foumaintraux, editors). Balkema, Rotterdam.

Steiger R.P. & Leung P.K. (1992) Quantitative determi- nation of the mechanical properties of shales. SPE Drilling Engineering, 181 185.

Tyridal D.S. (1994) Litologisk og mineralogisk sam- mensetning av slamsteiner i relasjon til loggrespons. Cand. Scient. (MSc) Thesis, Univ. Oslo, Norway.

Walker G.R. (1961) Vermiculite minerals. Pp. 279-324 in: The X-ray Identification and Crystal Structures of Clay Minerals (G. Brown, editor). Mineralogical Society, London.

Wang C., Mao N. & Wu F.T. (1980) Mechanical properties of clays at high pressure. Geophys. Res. 85, 1462-1468.

Warpinski N.R. (1992) Rock mechanic issues in completion and stimulation operations. Proc. 33rd U.S. Symposium, Rock Mechanics, 375-386.

Wensaas L., Shaw H.F., Gibbons K., Aagaard P. & Dypvik H. (1994) Nature and causes of overpressur- ing in mudrocks of the Gulifaks area, North Sea. Clay Miner., 29, 439-449.