lable at ScienceDirect

Marine and Petroleum Geology 59 (2015) 56e71

Contents lists avai

Marine and Petroleum Geology

journal homepage: www.elsevier .com/locate/marpetgeo

Research paper

Texture and diagenesis of Ordovician shale from the Canning Basin,Western Australia: Implications for elastic anisotropy andgeomechanical properties

Claudio Delle Piane a, *, Bjarne S.G. Almqvist b, 1, Colin M. MacRae c, Aaron Torpy c,Arthur J. Mory d, David N. Dewhurst a

a CSIRO Earth Science and Resource Engineering, 26 Dick Perry Avenue, Kensington 6151, Australiab Geological Department ETH Zurich, Switzerlandc CSIRO Microbeam Laboratory, Process Science and Engineering, Bayview Avenue, Clayton 3168, Australiad Geological Survey of Western Australia, Department of Mines and Petroleum, Mineral House, 100 Plain Street, East Perth, WA 6004, Australia

a r t i c l e i n f o

Article history:Received 29 October 2013Received in revised form5 February 2014Accepted 17 July 2014Available online 1 August 2014

Keywords:ShaleDiagenesisCrystallographic preferred orientationAnisotropyCanning Basin

* Corresponding author. Tel.: þ61 8 6436 8716.E-mail address: claudio.dellepiane@csiro.au (C. De

1 Now at Uppsala University Geocentrum, Villav. 16

http://dx.doi.org/10.1016/j.marpetgeo.2014.07.0170264-8172/Crown Copyright © 2014 Published by Els

a b s t r a c t

Microstructural and textural measurements from two Ordovician shale units (Goldwyer and Bongabinniformations) within the PalaeozoiceMesozoic Canning Basin indicate that the former unit was affected bymechanical compaction and clay mineral transformation whereas the latter preserves an early fabric dueto syn-depositional precipitation of authigenic dolomite and anhydrite. Conventional petrographicanalysis coupled with quantitative mineralogy, electron micro probe analyses, X-ray Texture goniometry(XTG) and cathodoluminescence spectroscopy of core samples were used to decipher the post-depositional evolution of marine and supratidal facies in the Goldwyer and Bongabinni formations,respectively. Differences in diagenesis are strongly reflected in the orientation of clay minerals asquantified by XTG: in both cases the c-axes of illite diffract strongest normal to the bedding plane but themeasurements clearly illustrate that shale in the Goldwyer Formation has a stronger preferred orien-tation relative to the Bongabinni Formation, with multiple of random distributions (m.r.d.) values of 5.77and 2.54, respectively.

Laboratory measurements conducted at 10 MPa effective stress also indicate distinct rock physicssignatures: the Bongabinni Formation shows very low anisotropy, whereas the Goldwyer Formationdisplays a higher degree of elastic anisotropy in terms of both P- and S- waves. The crystallographicpreferred orientation of illite, highlighted by the XTG, is likely to contribute to the significant differencein elastic anisotropy observed in the two units. Therefore, the Bongabinni Formation is mechanicallystronger and stiffer than the Goldwyer Formation, likely due to the early dolomite and anhydritecementation of the former providing a rigid microstructure framework.

Crown Copyright © 2014 Published by Elsevier Ltd. All rights reserved.

1. Introduction

Anisotropy in clay-rich sedimentary rocks is receivingincreasing attention, particularly as seismic anisotropy is essentialin prospecting for hydrocarbons, as well as in time lapsemonitoringof fluids in buried reservoirs. It is generally accepted that the

lle Piane)., 752 36 Uppsala, Sweden.

evier Ltd. All rights reserved.

intrinsic anisotropy of shaly facies is the result of both the orien-tation of the rock forming minerals and the presence of alignedelongated pores, plus organic matter when present. Preferredorientation of rock formingminerals, especially phyllosilicates withhigh single crystal elastic anisotropy, is an almost ubiquitousfeature of shales and mudstones and is often reported qualitativelybased on microstructural observations.

From the micro-analytical perspective quantification of thecrystallographic orientation of clay-rich facies is challenging dueto the small grain size and poor crystallinity of clay minerals,which hinders optical and electron microscopy texture

Figure 1. Left: Location and main structural features of the Canning Basin, the approximate location of the Sally-May 2 well is indicated with the white dot. The inset map shows theposition with respect to Western Australia (modified after Haines, 2009). Right: Generalized stratigraphic column of the lower part of the Palaeozoic sedimentary infill (modifiedafter Zhan and Mory, 2013).

C. Delle Piane et al. / Marine and Petroleum Geology 59 (2015) 56e71 57

measurements methods. Whereas X-ray diffraction techniques arehelpful to analyse clay mineral orientations, they are not free oflimitations, such as peak broadening due to small particles size,stacking disorder, polytypism, inter-layering and micro-straincontributing to peak overlaps in the diffraction spectra whichare difficult to deconvolve. Nevertheless, laboratory based X-raytransmission techniques have been developed to study theorientation distribution of basal planes of sheet silicates andsuccessfully applied to shales (e.g. Sintubin, 1994; van der Pluijmet al., 1994; Ho et al., 1999; Aplin et al., 2006; Day-Stirrat et al.,2008a, 2008b). More recently high energy synchrotron X-raydiffraction, has been used to obtain 3D crystal orientation distri-butions but the range of samples analysed to date is limited(Lonardelli et al., 2007; Wenk et al., 2008; Kanitpanyacharoenet al., 2011).

It is generally reported that phyllosilicate minerals show astrong preferred orientation with dominant axially symmetric polefigures typically showing the (001) maximum normal to thebedding plane and rotational freedom of (100) (e.g.Wenk et al.,2007). By comparison, non-clay minerals such as quartz, feldsparand carbonate showclose to random distributions. This observationsupports the assumption of transverse isotropy (T.I.) commonlyused in the description of the elastic properties of shaly rocks(Thomsen, 1986).

The depositional and subsequent mechanical compaction pro-cesses recorded within sedimentary rocks are often responsible forthe observed iso-alignment of phyllosilicates (Bjorlykke, 2013).Diagenesis can enhance or diminish such preferred orientationsthrough dissolutioneprecipitation reactions or by early cementa-tion which preserves the pre-compacted microstructural arrange-ment of the sediment. The texture or microstructure of a rock willdictate its physically measureable properties such as porosity,strength and stiffness, and as such is of primary interest wheninterpreting those measurements from either field, borehole orlaboratory investigations.

This case study presents different microstructural andtextural measurement techniques applied to understand thediagenetic history of shale formations deposited in differentgeological settings from the Canning Basin in Western Australia.

The effect of the different geological evolution of the two shalesis then related to their geomechanical behaviour and rockphysics properties based on high pressure laboratorymeasurements.

2. Samples and geological setting

The shale specimens for this study are from core in an onshorewildcat well, Sally May-2 (Lat: 19 48 5.00; Long: 124 27 27.00;Fig. 1), drilled in 2009 in the Canning Basin (northern WesternAustralia). This large PalaeozoiceMesozoic basin shows a multi-phase depositional history extending from the Early Ordovicianinto the Caenozoic. Deposition in the Canning Basin began inresponse to extensional tectonics and rapid subsidence which wasfollowed by four major and several minor phases of deposition anderosion (Ghori and Haines, 2006). The palaeothermal history of thesuccession is constrained by apatite fission track and vitrinitereflectance analyses, and indicates a maximum burial depth ofapproximately 2500 m in the Early Jurassic and maximum tem-peratures in excess of 100 �C (Duddy et al., 2005; Ghori and Haines,2006).

Recently the subsalt Ordovician stratigraphic section has beenthe target of renewed interest in the area owing to the demon-strated potential for shale gas and oil production from shales in theOrdovician Goldwyer and Bongabinni formations. Core samplesfrom these two formations in Sally May-2 are the focus of this study(Fig. 2).

The Middle Ordovician Goldwyer Formation, varies frommudstone dominated in basinal areas to limestone-dominated insome platform and terrace areas, and is interpreted as of openmarine to intertidal origin (Haines, 2004). The Bongabinni Forma-tion, the basal unit of the Ordovician to Early Silurian CarribuddyGroup, comprises redbed evaporitic mudstone, carbonate andsandstone. The Bongabinni Formation contains mostly oxidisedmarginal marine to supratidal facies deposited under low palaeo-topography conditions and with minimal influx of coarse grainedsiliciclastic detritus into the basin. The lack of marine fauna andbioturbation and the presence of evaporite minerals suggest at leastperiodically hypersaline conditions (Haines, 2009).

Figure 2. Wireline gamma ray log from Sally-May 2 with locations of the cores used for the experimental program. X-ray tomographic images of the 30 cm long cores and of thecore plugs used for geomechanics are also shown. The latter are displayed as two orthogonal slices through the specimens. Bongabinni generally shows a massive unstructuredtexture at core plug scale, whereas the Goldwyer shows strong laminations.

C. Delle Piane et al. / Marine and Petroleum Geology 59 (2015) 56e7158

3. Analytical techniques

The parent shale cores were extracted from depths rangingbetween 1505 and 1852 m (i.e. significantly shallower than theirestimated maximum burial depth) and preserved under oil insealed PVC core liners in order to maintain their native watercontent. Sub-samples for geomechanical testing were plugged andtrimmed using oil as a drilling fluid and stored under light processoil to prevent evaporation of pore fluids. Offcuts of the core plugswere used for characterisation in terms of composition andmicrostructural evaluation.

3.1. Bulk composition: XRD mineralogy

Quantitative mineralogy analysis performed by collecting X-raydiffraction (XRD) patterns from a few grams of powdered samples.Details on sample preparation for XRD are given in Delle Piane et al.(2014); the XRD patterns were recorded with a PANalytical X'PertPro multi-purpose X-ray diffractometer using iron-filtered CobaltKa radiation, 1/4� divergence slit, 1/2� anti-scatter slit and

X'Celerator silicon-strip detector. The diffraction patterns wererecorded in steps of 2q ¼ 0.017� with a 0.5 s counting time per step,and logged to data files for further processing and analysis.

Quantitative analysis of the XRD data was performed using thecommercial package SIRO-QUANT from Sietronics Pty Ltd. The datawere first background subtracted and calibrated for the automaticdivergence slit. The results are normalised to 100% and hence donot include unidentified or amorphous materials.

3.2. Petrography

Thin sections were prepared from offcuts of the core plugs withthe plane of the section normal to the macroscopic lamination forobservation in transmitted light using an Olympus AX70 opticalmicroscope equipped with an Olympus DP71 camera for imageacquisition. More detailed analyses followed by means of electronmicroscopy using a Philips XL 40 and a ZEISS Field emission in-strument. Operational mode was mainly using back scatteredelectrons (BSE), where contrasts in grey level in the reconstructedimages correspond to contrasts in atomic number and therefore

C. Delle Piane et al. / Marine and Petroleum Geology 59 (2015) 56e71 59

chemical composition of the analysed area. Particular areas of in-terest were analysed to retrieve the chemical composition of theregion via energy dispersive X-ray spectroscopy (EDS).

Additional polished sections were used for mineral mappingand cathodoluminescence investigations on a JEOL 8500F-CL fieldemission gun electron probe micro-analyser (FEG-EPMA); at eachpixel of the map, elemental composition was measured by wave-length dispersive x-ray spectrometry (WDS) and EDS, referencestandards were used to calibrate the measurements of eachelement. In parallel to the EDS and WDS measurements, thebackscattered electron signal was measured, as well as the infrared,visible and UV cathodoluminescence (CL) spectra, collected usinga ‘xCLent’ cathodoluminescence spectrometer system. The CL signalcan be used to infer important information on the nature of thevarious mineral phases: in this study we focus on quartz. Previousauthors indicated that bright luminescent grains derive fromigneous and metamorphic rocks, whereas authigenic quartz hastypically less intense luminescence (see for example review byG€otze et al., 2013). Following data acquisition, the mapping resultswere processed in the Chimage package to identify chemical phasesusing a hierarchical clustering algorithm (Wilson et al., 2012).

3.3. X-ray texture goniometry

X-ray texture goniometry (XTG) measurements were performedon samples of the Bongabinni and Goldwyer formations to quan-titatively estimate the orientation of the constituentminerals of therocks. In particular, we focused on the preferred orientation of illite.Quantitative assessment of the alignment of phyllosilicates wasmade using a similar method as that described by van der Pluijmet al. (1994). Pole figures measurements were performed in trans-mission mode using thin sections of the shale samples withthickness of ~200 mm. Transmission mode analysis allows for themeasurements of pole figures at low 2q angles, which is necessarywhen measuring the orientation of the basal plane of clay minerals(i.e. pole to (001) and (002)). The 2q angle for the illite basal plane(001) diffraction peak was first identified for each sample, by 2q-scans. Background measurements for the individual pole figureswere performed in regions of the 2q spectrum without diffractionpeaks.

After corrections of background and absorption effects (Kockset al., 2000), the processed data was imported into the Labotexsoftware (LaboSoft s.c., Poland) for analysis and plotting. The degreeof alignment of the selected minerals is displayed in pole-figurediagrams that show the distribution of crystallographic orienta-tions in the form of poles to crystallographic planes. Pole-figuresallow visualization of the spatial distribution of the X-ray diffrac-tion intensities by displaying contour lines representing the poledistribution of particular crystallographic plane orientations. Thedegree of particle alignment or, in other words, the strength of atexture, is expressed in multiples of a random distribution (m.r.d.)(Wenk, 1985), where higher values reflect higher degree of align-ment. A random texture will produce a diffuse pole figure with nomaximum density with a m.r.d. value of 1.

4. High pressure testing

4.1. Geomechanics

The plugs used for the geomechanical tests were cored withtheir axis perpendicular to the macroscopic bedding identified onthe parent cores, trimmed to obtain parallel flat ends and immersedin jars filled with a light mineral oil to maintain their native hy-dration state until testing. Coring and trimming were also

performed using oil as fluid to minimize water loss during samplepreparation.

Undrained triaxial tests were performed on the preservedshale samples (38 mm in diameter and 77 mm in length) using aTerratek testing machine in the same configuration described byDelle Piane et al. (2011). The equipment comprises a high stiffnessload frame, a pressure vessel (using hydraulic oil as a confiningmedium) and three independent stepping motor pumps forconfining, pore pressure and axial load control. The sample as-sembly includes:

� A cylindrical sample mounted between top and base platens� An impermeable Vitonmembrane (0.75mm thick), isolating thespecimen from the hydraulic and housing ultrasonic P- and S-transducers

� Two steel platens housing ultrasonic P- and S- transducers withprovision for pore pressure measurements placed at both endsof the specimen

� Two diametrically opposed linear variable differential trans-formers (LVDT) clamped on the top and bottom platens tomeasure axial displacements

� A load cell placed underneath the bottom platen.

Following installation in the testing rig the samples underwenta saturation stage at constant confinement and pore pressure. Thepore fluid was synthetic brine reproducing average sea watersalinity (3.5 wt% NaCl) thought to be representative of the depo-sition conditions.

An initial consolidation stage was carried out for the two sam-ples; this consisted in holding a stable isotropic pressure of 15 MPaand a constant pore pressure of 5 MPa at the top and bottom end ofthe sample via pump control. Consolidation stages were monitoredfor stabilization of sample deformation and of ultrasonic signalsand lasted 18 and 19 days for the Bongabinni and Goldwyer samplerespectively; this and the fact that dehydration of the samples waskept to a minimum by immersion in oil from recovery at the wellsite to testing in the laboratory gives us confidence to assume asaturated response of the specimen during the experimentalprogram.

After saturation and sample stabilization, triaxial tests wereperformed at confining and pore pressures of 15 and 5 MParespectively; axial load was applied at a constant displacement rateof 0.07 mm/h (axial strain rate of approximately 2.5 � 10�7 s�1)under undrained conditions until failure. Stress strain curves werethen used to calculate the peak stress and the static Young'smodulus of the samples. Peak stress corresponds to the maximumlevel of differential stress (s1 minus s3) sustained by the samplebefore failure. The Young's modulus (E) was determined as thetangential slope of the differential stress/axial strain curve between40% and 60% of the maximum differential stress. Full details of thetesting methods and processing can be found in Delle Piane et al.(2011).

4.2. Ultrasonic testing

Ultrasonic velocity measurements were taken at increasingvalues of axial load during the undrained triaxial tests using thetransducer configuration described by Delle Piane et al. (2011). Themethod consists of measuring the travel time of an elastic pulsethrough a cylindrical rock sample of known length and diameter.

Velocities were measured along different propagation di-rections through a series of transducers located at the flat ex-tremities or at the circumferential surface of the samples.Measurements of velocities were made at increasing axial load forthe following vibration modes:

Table 2Physical properties of the Bongabinni and Goldwyer Formation samples.

Formation w (wt%) rb (g/cm3) f(%) fhg (%)

Bongabinni 3.8 2.6 9.5 5.7Goldwyer 4.6 2.6 11.3 3.6

w ¼ water content; rb; bulk density; f ¼ porosity calculated from water evapora-tion; fhg ¼ porosity as measured by mercury injection tests.

C. Delle Piane et al. / Marine and Petroleum Geology 59 (2015) 56e7160

� Vpv: compressional wave propagating normal to bedding� Vs1: shear wave propagating normal to bedding with polariza-tion parallel to bedding

� Vph: compressional wave propagating parallel to bedding� Vsh: shear wave propagating parallel to bedding with polari-zation parallel to bedding

Note that while the transducer configuration allows for mea-surement of the off axis quasi-P-wave velocity, the quality of therecorded signal did not allow unambiguous identification of thetime of arrival and is therefore not presented here. The nominalexcitation frequency used in this studywas 0.5MHz. Raw datawerecorrected for the membrane and platen end-cap thickness andconverted into ultrasonic velocity using the length of the specimenat the corresponding pressure conditions.

The measured velocities were used to calculate the anisotropyfactors ε and g (Thomsen, 1986) defined as:

ε ¼V2ph � V2

pv

2�V2pv

�

g ¼ V2sh � V2

s1

2�V2s1

�

5. Results

At the macroscopic scale all the samples are well consolidated,cohesive and dense sedimentary rock. Quantitative XRD results ofthe bulk shale composition are shown in Table 1; details ofanalytical protocol and the clay fraction mineralogy are reported inDelle Piane et al. (2014). Petrophysical characteristics of the sam-ples are reported in Table 2. There is a marked difference betweenporosity measured by the water evaporation method and thatinferred from mercury intrusion. Such a discrepancy can be tenta-tively attributed to the following considerations:

� Mercury only intrudes the interconnected, open porosity; on theother hand water evaporation at 105 �C removes some of theclay boundwater which then is computed as total porosity. Notethat both the Goldwyer and Bongabinni formations samplescontain a certain amount of hydrated minerals (illite and chlo-rite) that could release water when heated;

� Although care was taken to select samples for porosity deter-mination, heterogeneities between the specimens from thesame formation might reflect the observed difference inporosity value assessed from the two methodologies.

X-ray tomograms of core plugs are shown in Figure 2, whilemicrostructures at the whole thin section scale from the twoOrdovician formation specimens are illustrated in Figure 3. The twoformations are devoid of swelling clays and the only clay mineralsidentified by XRD are illite and chlorite.

Table 1Bulk mineralogy of the Bongabinni and Goldwyer Formation shale samples asmeasured by XRD (values expressed in weight %).

Formation Qz Chl Mica/Ill Ab Orth Calc Dol Pyr Hem Anh

Bongabinni 14 1 41 <1 5 29 1 9Goldwyer 25 4 56 1 6 8 1

Qz ¼ quartz; Chl ¼ chlorite; Mica/ill ¼ mica/illite; Ab ¼ albite; Orth ¼ orthoclase;Calc ¼ calcite; Dol ¼ dolomite; Pyr ¼ pyrite; Hem ¼ haematite; Anh ¼ anhydrite.

The Bongabinni Formation is typically red (in the web version)due to the presence of haematite; at the plug scale and under theoptical microscope it appears massive with no evidence of beddingor lamination (Figure 3a). Millimetre-sized green patches of anhy-drite are widespread (Figure 3a). There is no evidence of macro-scopic bedding or pervasive fractures at the core scale (Fig. 2).Overall the grain size of the rock is very fine so that individualgrains can hardly be distinguished under the optical microscope.XRD analyses (Table 1) indicate major mineralogical components asillite/mica and dolomite, with subordinate quartz and anhydriteandminor amounts of haematite. The fine-grained (<2 mm) fractionis dominated by illite/mica with sporadic occurrence of orthoclaseand chlorite.

The Goldwyer Formation samples are dark green to greyish,with very fine lamination down to the millimetre scale, consistingof alternating clay- and silt-rich layers (Figs. 2 and 3b). Some of thelong, thin, planar, parallel fractures observed under the opticalmicroscope appear partially occluded with either crystalline ororganic material. Fossils and organic matter in various shapes andsizes are present particularly in the calcite cemented silt-richlaminae (Figure 3b). Quantitative XRD analyses (Table 1) revealmica/illite as the dominant phase; quartz and orthoclase as majorcomponents, with subordinate calcite, chlorite and pyrite. The finegrained fraction is illite/mica-dominated with chlorite as a secondconstituent. In summary, the Bongabinni Formation shale isdominated by illite and dolomite with subsidiary quartz andanhydrite, whereas shale in the Goldwyer Formation is mainlycomposed of illite and quartz.

SEM images of the Bongabinni Formation samples show equantand rounded detrital grains of quartz and feldspar surrounded by afine-grained matrix. The detrital grains range between tens andhundreds of microns (Fig. 4a) and generally show heavily corru-gated boundaries. Anhydrite crystals constitute a prominentfeature of the microstructure and show two distinct habits:millimetre-sized patches (Fig. 4b) or highly elongated crystals(Fig. 4c) generally showing straight boundaries and occasional in-dents when in contact with other angular clasts (Fig. 4c). Theorientation of these anhydrite grains ranges from random toweakly aligned. Euhedral dolomite crystals are alsowidespread andshow characteristic lozenge shape and variable sizes(1e200 microns), often with concentric zoning (Fig. 4d). Also pre-sent are weathered mica grains and clay minerals (illite and chlo-rite) filling the space between larger grains without showing adiscernible alignment (Fig. 4d); because of their small size indi-vidual clay particles cannot be discerned. Accessory mineralsinclude apatite, haematite and zircons; no organic matter wasrecognized within the analysed sample. Porosity is only visible insub-micron contacts between clay platelets (Fig. 5).

The microstructure of shale within the Goldwyer Formationvaries depending on the grain size of the laminae. Coarse-grainedlaminae contain grains of quartz, K-feldspars and apatite severaltens of microns across; the grains are surrounded by patches ofcalcite. Pores are visible within the calcite patches and as partiallydissolved organic shells; these are at times replaced by quartz andalso host, in their concave side, clusters of framboidal pyrite(Fig. 6a). No particular particle alignment is recognizable other than

Figure 3. Optical microphotographs of whole thin sections in plane polarized transmitted light. Bongabinni Formation (a), and Goldwyer Formation (b). Note the massiveappearance of the Bongabinni Formation shale as opposed to the strongly layered microstructure of the Goldwyer Formation shale.

Figure 4. SEM images of the Bongabinni Formation sample. a) overview of the microstructure with equant and rounded detrital grains of quartz (Qz) and feldspars (light grey in theimage) surrounded by a fine grained matrix b) tight cluster of early diagenetic anhydrite (Anhydr) c) early diagenetic anhydrite needles bent against a rigid detrital grain of quartz.d) association of dolomite (Dol) and anhydrite, note the lozenge shape of the dolomite crystals and the concentric zonations visible in the grain in the middle of the figure.

C. Delle Piane et al. / Marine and Petroleum Geology 59 (2015) 56e71 61

Figure 5. Sub micron porosity in the Bongabinni Formation sample imaged under theSEM (secondary electron image).

Figure 6. SEM images of the Goldwyer Formation sample. a) coarse grained, calcite cemenapatite; Py ¼ pyrite. Dissolved shells are replaced by calcite and quartz cement. Note the seshell, and within the calcite cement just below the apatite clast. Bedding is sub parallel to thlarge apatite clast in the middle of the image; note the thin, elongated fracture (in black) rusurrounded by illite (Ill), note the micro porosity within the apatite d) elongated patch ofgrained clay matrix around the rigid quartz grain indicating mechanical compaction andalteration of the detrital constituents of the rock.

C. Delle Piane et al. / Marine and Petroleum Geology 59 (2015) 56e7162

in the elongated bioclasts of apatite which generally have their longaxis oriented sub-parallel to the macroscopic bedding (Fig. 6a).

The clay rich laminae have a fine matrix mainly composed ofillite with accessory chlorite and microcrystalline quartz. Sub-angular to rounded clasts of quartz and K-feldspar are suspendedin the matrix with elongated crystals of mica and apatite bioclastsoriented predominantly with their long axis sub-parallel tobedding (Fig. 6b). Intra-crystalline micro- and nano-pores charac-terize the structure of the apatite (Fig. 6c). Occasional elongatedpatches of organic matter preferentially aligned parallel to thebedding are visible within the clay-rich laminae (Fig. 6d).

A final comment regards the identification of mica/illite in theBongabinni Formation shale: XRD is not able to distinguish be-tween the two, yet mica is generally of detrital origin in shale de-posits and as such it can be identified based on its size and shape inSEM images. Our investigations show almost no mica, thereforeindicating that the content values reported from the XRD analysiscan be confidently attributed to illite.

5.1. EMPA results

Mineral maps for the Bongabinni Formation (800 � 800 mm,step size 0.5 mm and dwell time of 40 ms) and Goldwyer Formation(900 � 900 mm, step size 0.5 mm and dwell time of 30 ms) samples

ted lamina: Ru ¼ rutile; Cc ¼ calcite; K-f ¼ F-feldspar; Qz ¼ quartz; Bio-Ap ¼ biogeniccondary porosity in the middle of the image corresponding to a partially cement-fillede elongated apatite clast. b) clay rich, fine grained lamina: bedding is sub parallel to thenning across the sample. c) Close-up image at the contact between two apatite clasts,organic matter (Org) aligned within the bedding plane, note the deflection of the finethe corrugated boundaries of the quartz and K-feldspar grains indicating chemical

Figure 7. Bongabinni Formation microstructure as imaged via electron probe microanalyser. a) Back scattered electron image of the region of interest. b) mineral map based onwavelength dispersive x-ray spectrometry. c) panchromatic CL image. d) distribution of quartz grains defined based on the blue region of the scatter plot in e), colour represents theintensity of the CL signal shown in the bar. e) cross plot of CL intensity versus Silicon EDS signal; the blue area represents the quartz grains visualized in d). (For interpretation of thereferences to colour in this figure legend, the reader is referred to the web version of this article.)

C. Delle Piane et al. / Marine and Petroleum Geology 59 (2015) 56e71 63

C. Delle Piane et al. / Marine and Petroleum Geology 59 (2015) 56e7164

Figure 9. Comparison of the characteristic CL signal of quartz grains and quartzcement form the Goldwyer Formation shale. Note the difference in shape between thetwo spectra and the absence of 2.72 eV peak in the quartz cement signal indicating lowtemperature quartz formation (see text for details).

Figure 10. Density contoured pole figures of the normal to the basal plane of illite (001) measured in transmission mode at diffraction angle (2q) of 8.9� . The black thick linerepresents the orientation of the bedding plane (B.P.). The results indicate a much stronger illite alignement in the Goldwyer Formation compared to the Bongabinni Formation shale.

Table 3Summary of mechanical parameters retrieved from triaxial testing.

Formation Peak stress(MPa)

Young's modulus(GPa)

Strainrate (s�1)

Axial strainat failure (%)

Goldwyer 8.76 4.56 2.50E-07 0.17Bongabinni 24.48 7.59 2.48E-07 0.30

C. Delle Piane et al. / Marine and Petroleum Geology 59 (2015) 56e71 65

are illustrated in Figures 7 and 8 together with the BSE andpanchromatic CL images of the same areas.

The larger grains within the microstructure of the BongabinniFormation are detrital quartz, K-feldspar and anhydrite, whereasthe fine-grained matrix is mainly composed of dolomite, illite andchlorite with an area abundance of 27%, 26% and 6.5%, respectively(Fig. 7a and b).

Cathodoluminescence across the specimen, principally fromquartz and K-feldspar are shown in the panchromatic total CLemission image (Fig. 7c). Quartz grains display a range of CL colourand intensity. To highlight the distribution of the weakly lumi-nescent quartz cement, the data was reprocessed to emphasize thequartz grains: by selecting high silicon content (scatter plot Fig. 7e).In the resulting images the quartz grains are brighter with a rangeof colours associated with the stronger cathodoluminescenceemission (Fig. 7d) compared to the weak emission from the quartzcement or overgrowth which has been coloured grey through theuse of select palette. In the Bongabinni Formation sample no quartzbonding phase was detected in the mapped area.

The mapped area of the Goldwyer Formation sample corre-sponds to one of the silt-rich laminae described above and is shownin Figure 8a and b as a backscattered electron image and mineral

Figure 8. Goldwyer Formation microstructure as imaged via electron probe microanalyserwavelength dispersive x-ray spectrometry. c) panchromatic CL image. e) distribution of lumiCL intensity versus Silicon EDS signal; the blue area represents the quartz grains visualizedreferences to colour in this figure legend, the reader is referred to the web version of this

map, respectively. The modal analysis indicates the followingcomposition expressed in area %: quartz 43%, K-feldspar 31%, albite11%, calcite 8.6%, muscovite 2.6% and chlorite 2.1%. The micro-structure is grain supported with quartz and K-feldspar forminginterconnected mono-phase clusters with a thickness of 1 or 2grains but several hundreds of microns long (Fig. 8a). Mica andchlorite grains are isolated in between these clusters and seem tobe aligned parallel to their long axis defining the sedimentarybedding. Calcite is also scattered in themicrostructure, but does notshow any preferred shape or alignment.

Intense luminescence is recordedmainly from the quartz and K-feldspar grains appearing as blue to orange and green to redrespectively in Figure 8c. Figure 8d and f shows the distribution ofquartz cement and overgrowths based on the discrimination of CLcounts and Si Ka radiation illustrated in the blue shaded are of thescatter plot (Fig. 8e). Quartz cementation is widespread and con-

stitutes approximately 7% of the illustrated area. The cath-odoluminescence emission associated with the quartz cement orovergrowth is weak indicating the material is crystalline. A com-parison of the cathodoluminescence spectra from quartz grains andcement regions show the quartz grains are dominated by a peak at1.93 eV with a minor peak at 2.72 eV (Fig. 9). The 1.93 eV peak inquartz is associated with non-bridging oxygen hole centres(NBOHC) while the 2.72 eV has been associated with titaniumincorporation (Leeman et al., 2012). The Ti concentration has beenshown to be a proxy for quartz formation temperature as estab-lished by Wark and Watson (2006). The almost absence of the2.72 eV peak in the cement quartz indicates this was formed at amuch lower temperature and since it is an overgrowth was a latestage development in the formation of this deposit.

. a) Back scattered electron image of the region of interest. b) mineral map based onnescent quartz defined based on the blue region of the scatter plot in e). e) cross plot ofin d). f) close up of the area delimited by the red box in d). (For interpretation of the

article.)

C. Delle Piane et al. / Marine and Petroleum Geology 59 (2015) 56e7166

Table 4Summary of experimental conditions, ultrasonic velocities and anisotropy param-eters during the triaxial tests.

Formation CP(MPa)

PP(MPa)

s1 � s3(MPa)

Vph(m/s)

Vpv(m/s)

Vsh(m/s)

Vs1(m/s)

ε g

Goldwyer 15.02 4.94 0.94 4197 3288 2619 1595 0.31 0.8515.01 5.19 1.95 4295 3402 2600 1591 0.30 0.8315.01 5.52 4.04 4295 3415 2600 1611 0.29 0.8015.00 6.31 7.89 4246 3428 2600 1633 0.27 0.77

Bongabinni 15.04 5.03 1.04 4342 4173 2821 2717 0.04 0.0415.05 6.27 1.95 4342 4218 2821 2717 0.03 0.0415.05 7.08 4.00 4342 4171 2821 2716 0.04 0.0415.05 7.79 7.78 4342 4169 2821 2734 0.04 0.0315.05 8.13 15.54 4342 4187 2821 2760 0.04 0.02

CP ¼ confining pressure; PP ¼ pore pressure; s1 � s3 ¼ differential stress. Velocitiesand ε and g defined in the text.

C. Delle Piane et al. / Marine and Petroleum Geology 59 (2015) 56e71 67

5.2. XTG results

Illite particle orientation was analysed using XTG on thin sec-tions from core of the Bongabinni and Goldwyer formations. Thecrystallographic orientationwasmeasured at diffraction angles (2q)of 8.9� corresponding to the (001) plane. The results are illustratedas contoured pole figures in Figure 10 depicting the distribution ofthe pole to the illite basal plane; the horizontal lines passingthrough the centre of the circles represent the orientation of thebedding plane. The Bongabinni Formation sample shows a verydiffuse distribution, weak maxima are observed orthogonal to thebedding planes for illite, quantified with m.r.d. values of 2.54. Thefabric of the Goldwyer Formation sample is significantly betterdefined and the clay mineral show strongly centred distributionnormal to the bedding plane with m.r.d. of 5.77.

5.3. Geomechanics results

Peak strength and static Young's modulus measurements indi-cate that the Bongabinni Formation is stronger and stiffer than theGoldwyer Formation. Peak values of differential stress weremeasured at 8.4 and 23.0 MPa for the Goldwyer and Bongabinniformations, whereas Young's moduli were 4.2 and 6.9 GParespectively at confining pressure of 15 MPa (Table 3; Fig. 11). TheBongabinni Formation sample failed catastrophically after 0.3%axial strain, similar to the Goldwyer Formation sample whichsustained an axial strain of just 0.17% at failure (Fig. 11 and Table 3).

5.4. Ultrasonic results

Figure 11a and b shows the schematic position of the ultrasonictransducers used to monitor the evolution of the elastic wave ve-locities as a function of the applied stress and examples of arrivalpicks for each propagation mode. Ultrasonic P- and S-wave veloc-ities measured during increasing axial load in different propagationdirections with respect to the macroscopic bedding of the speci-mens (Table 4) are illustrated in Figure 12a and b. The two forma-tions show similar elastic wave velocities along the bedding planes,but the Goldwyer Formation sample has significantly lower P- andS- wave velocities whenmeasured across the sedimentary layering.

Figure 11. Results of the combined geomechanical and rock physics laboratory experiment.triaxial testing at a confining pressure of 15 MPa and initial pore pressure of 5 MPa. Axial loadthe calculation of the tangential Young's modulus. The Goldwyer Formation shale is seenschematic configuration of the ultrasonic transducers used to measure the longitudinal and sthe Goldwyer Formation sample, the vertical solid line in each of the four plots represents thS-waveforms although the changes in frequency and amplitude with the S-wave arrival are

Stress sensitivity of the elastic waves is illustrated in Figure 12c andd, where velocities are scaled to valuesmeasured at the lowest levelof differential stress. A significant difference between the two for-mations is that the Bongabinni Formation shale shows minimal tono stress sensitivity of the ultrasonic velocities and a maximumincrease of 2% for the S-wave travelling along the direction ofapplied maximum principal stress (Fig. 12d). By comparison, theGoldwyer Formation sample displays an increase of the velocitiestravelling normal to bedding (and parallel to s1) of up to approxi-mately 5%, whereas the velocities along the bedding (and normal tos1) show a non constant response to stress with Vph initiallyincreasing and suddenly dropping at the highest stress levels andVsh decreasing and stabilizing to a value lower than the initial one(Fig. 12c).

Anisotropy parameters ε and g also highlight differences in therock physics response of the two shales (Fig. 12e and f; Table 4):both P- and S-wave anisotropy are high in the Goldwyer Formationsamples with values decreasing from 0.31 to 0.27 and 0.85 to 0.77respectively over a differential stress increment between 1 and8 MPa (Fig. 12e). Conversely, the Bongabinni Formation sampleshows negligible values of anisotropy and stress sensitivity(Fig. 12f) with ε and g stable at values of approximately 0.05.

6. Discussion

6.1. Microstructures and diagenesis

Assuming a lithostatic gradient of 22.6 MPa/km as a typicalaverage value (Lyons and Plisga, 1984) and hydrostatic gradient of12.5 MPa/km, the maximum effective stress at 2.5 km is estimatedto be approximately 30 MPa. Experimental compaction of pure clayaggregate indicates that at such stress levels the initial porosity ofan unconsolidated slurry would reduce to values ranging between15 and 40% depending on the dominant clay (Mondol et al., 2007).Obviously these values aremuch larger than those seen in the rocksstudied here, indicating that other processes than purely mechan-ical compaction are responsible for their low porosity.

Diagenetic reactions such as illitization of smectite are oftenreported for shaly facies to be active in deeply buried sections of thesedimentary column where temperatures exceed 70 �C (e.g.Bjorlykke, 2013). Nevertheless, shallow, early growth of diageneticphases has also been reported (e.g. Lash and Blood, 2004). Theprecipitation of diagenetic mineral phases is therefore consideredan important factor in textural alteration and concomitant porosityand permeability reduction of sedimentary facies.

The Bongabinni Formation has been previously interpreted to bedeposited in a low energy supratidal environment with limitedsediment influx. Several lines of evidence support such an inter-pretation; e.g the lack of obvious marine fossil or bioturbation andthe presence of abundant evaporitic minerals indicates a hypersa-line environment akin to a Sabkha in which gypsum (or anhydrite)and dolomite form due to evaporation (Bjorlykke, 2013). Indentedand bent crystals of anhydrite when in contact with detrital quartz(Fig. 4c) indicate they have an early diagenetic origin. The euhedralshape of the dolomite crystals, as well as their concentric zonalgrowth as seen via SEM (Fig. 4d) and cathodoluminescence images,

a) Differential stress (s1 � s3) as a function of axial strain measured during undrainedapplied normal to the macroscopic bedding; the thickened part of the curve is used forto be weaker and more compliant than the Bongabinni Formation. Insert shows thehear elastic wave velocities across the sample. b) Example of waveforms collected frome picked arrival time of the ultrasonic elastic wave. Precursive P-waves are noted on theclear.

C. Delle Piane et al. / Marine and Petroleum Geology 59 (2015) 56e7168

is consistent with early in situ precipitation in a non-compactedsediment rather than having been transported. It is not clearwhether dolomite pre- or post-dates anhydrite formation, althoughsome dolomite crystals impinge upon and bend the latter. Diage-netic anhydrite associated with dolomite has been attributed toreflux of hypersaline brines from overlying evaporite beds post-dolomitization; alternatively, calcium released during dolomitiza-tion has been proposed to explain anhydrite mineralization (Jonesand Xiao, 2005 and references therein). In the case of the Bonga-binni Formation, it is inferred that both phases are the result ofdiagenetic alteration immediately after deposition, but before sig-nificant burial. As such, anhydrite and dolomite acted as a rigidframework during burial resisting the effects of mechanicalcompaction and later diagenetic change similar to the calcitecemented horizons of the Barnett Shale described by Day-Stirratet al. (2008b).

From microstructural observation, it can be inferred that thepost-depositional and diagenetic history of the Goldwyer Forma-tion included episodes of calcite cementation, pyritization oforganic matter and probably illitization of precursor clay. Calcitedissolution is also visible in the partially dissolved shells as well asprecipitation of clay in said shells. As indicated by the cath-odoluminescence images, quartz cementation was also part of thediagenetic history. Quartz cementation could be interpreted as theresult of smectite to illite transformationwhich is known to releasewater and SiO2 as well as under particular chemico-physical con-ditions, i.e. availability of Kþ and temperatures above 70 �C (Bolesand Franks, 1979), which seem realistic given the presence ofaltered K-feldspar in the microstructure of the Goldwyer Formationand the palaeothermal history of the Canning Basin.

Flocculation at the bottom of the sea, as in the case of an openmarine sedimentary environment, usually produces near randomlyoriented texture of clay particles and the lowest observed texturalstrengths in unconsolidated sediments are of the order of1.70e1.90 m.r.d. (Matenaar, 2002). This suggests that the strongparticle alignment observed in the Goldwyer Formation sample isthe consequence of mechanical rotation of the clay particles duringburial due to increasing overburden stress and possibly chemicaltransformation of the clay minerals. Previous studies suggestedthat to achieve a significant illite preferred orientation (m.r.d. >4),diagenetic phyllosilicates reactions, e.g. the smectiteeillite trans-formation, may be critical (see for example Ho et al., 1999; Day-Stirrat et al., 2008a).

The stratigraphic difference in depth between Bongabinni andGoldwyer formations (just under 90 m) cannot explain the majordifference in intensity of the clay fabrics, as the maximum effectivestresses and temperatures the two formations have undergoneduring burial and uplift must have been practically the same.Therefore it is likely that early cementation in the BongabinniFormation due to precipitation of dolomite and anhydritecontributed to the preservation of a very early phyllosilicatecompaction state; similar observations were reported on thecalcite-cemented lithologies of the Barnett Shale studied by Day-Stirrat et al. (2008b) showing significantly less aligned phyllosili-cates inside (2.11 and 2.34 m.r.d.) compared to outside calciteconcretions (4.25 and 4.55 m.r.d.). The interpretation of the postdepositional diagenetic evolution of the two formations is sum-marized in a simplified conceptual model in Figure 13.

6.2. Microstructural effects on macroscopic mechanical and elasticbehaviour

Having established that the two Ordovician shales show sig-nificant differences in their microstructure and diagenetic evolu-tion despite their proximity in depth and geological history, it is of

interest to assess how these differences are reflected in themacroscopic behaviour as measured under controlled conditions inthe laboratory. In this sense, three aspects of the high pressure testresults are particularly interesting:

1. The different stress sensitivities of the elastic wave velocitiesbetween the two formations;

2. The differences in elastic anisotropy;3. The differences in strength and stiffness.

It should be noted that the mineralogy of the Bongabinni For-mation and in particular the high amount of dolomite and anhy-drite and the lack of observable microcracks are responsible for itsstiffer and stronger mechanical response compared to the Gold-wyer Formation sample. Single crystal elastic moduli for the twominerals are significantly higher than those of the other majorconstituents of the rocks (i.e. quartz and illite) and are reported inthe literature as follows: anhydrite bulk modulus ¼ 56.1 GPa,anhydrite shear modulus ¼ 29.1 GPa; dolomite bulkmodulus ¼ 94.9 GPa; shear modulus ¼ 45.0 GPa (Mavko et al.,2009).

The stronger stress sensitivity of the Goldwyer Formation shaleultrasonic velocities and their anisotropy suggests the presence ofaligned micro-cracks parallel to the bedding plane that close withincreasing axial load, as reported for similar experiments ondifferent samples and from inversion of experimental results (e.g.Delle Piane et al., 2011; Sarout et al., 2007). The microstructure ofthe Goldwyer Formation shows the presence of elongated thindiscontinuities sub parallel to the sedimentary layering (Figs. 3 and6b) consistent with our interpretation of the experimental results.Such features are not observed in the Bongabinni Formation sampleand fit with the negligible stress sensitivity of the experimentallymeasured velocities and anisotropy.

In sedimentary rocks, clay minerals are the main contributors toa non-random texture, whereas silicates and carbonates aregenerally reported as nearly randomly oriented (e.g. Valcke et al.,2006). Due to the highly anisotropic nature of clay minerals, thedifference in alignments measured by XTG qualitatively reflects theexpected elastic anisotropy in the Canning Basin samples. It shouldbe noted, however, that themeasured anisotropy is likely due to thesuperimposition of clay particles iso-alignment and laminationsand micro-cracks in the microstructure. So far direct quantificationof microcracks in shale has proven very difficult; however thepresence of low aspect ratio pores in samples of Opalinus Clay hasbeen recently demonstrated by Keller et al. (2011) via nano-tomography and is likely to be a common feature in highly lami-nated shales.]

7. Conclusions

Ordovician shale samples from the Canning Basin have beenanalysed with qualitative and quantitative microstructural tech-niques to assess their post depositional diagenetic history. It isinferred that depositional environment imparts a critical influ-ence on the subsequent microstructural development of theshales. The supratidal Bongabinni Formation was affected by earlycementation by dolomite and anhydrite which provided a stiffcement framework that resisted subsequent mechanicalcompaction during burial, thereby preserving a very earlycompaction fabric in the clay minerals. In contrast, the openmarine Goldwyer Formation shale is dominated by illite andquartz and shows little evidence of early cementation. Itsmicrostructure has strong clay mineral alignment and is charac-terized by late quartz cements.

Figure 12. Results of the ultrasonic measurements collected during triaxial tests on the Goldwyer (left column) and Bongabinni (right column) Formation samples as a function ofdifferential stress. a) and b) ultrasonic velocities, nomenclature of the various velocities is explained in the text. c) and d) scaled velocities normalised by the values recorded at theminimum level of differential stress. e) and f) P-wave (ε) and S-wave (g) anisotropy parameters.

C. Delle Piane et al. / Marine and Petroleum Geology 59 (2015) 56e71 69

The microstructure of the rocks, itself a result of the originaldeposition environment and subsequent diagenesis, exerts a pri-mary control on the macroscopically measured physical propertiesof the rocks. In particular, there is a qualitatively good correlationbetween the degree of alignment of illite and the elastic anisotropyof the rock, as well as between themineralogy and the strength andstiffness of the samples.

This research has demonstrated the significant input thatdetailed microstructural investigations can have in understand-ing the role of compaction and diagenesis on particle alignmentand resulting rock properties in shales. Understanding andquantifying the relationships between microstructural anisot-ropy and rock property anisotropies (strength, elasticity,permeability, etc.) could positively impact practical activities

Figure 13. Conceptual model summarizing the post depositional evolution of the Goldwyer (left) and Bongabinni Formation shales as interpreted from the multi-techniquemicrostructural investigations.

C. Delle Piane et al. / Marine and Petroleum Geology 59 (2015) 56e7170

such as drilling through overburden shales or exploiting gasshale resources.

Acknowledgements

Mike Verrall is thanked for his help with the acquisition of theSEM images from the thin sections carefully prepared by DerekWinchester. AJM publishes with the permission of the Director,Geological Survey of Western Australia.

References

Aplin, A.C., Matenaar, I.F., McCarty, D.K., van der Pluijm, B.A., 2006. Influence ofmechanical compaction and clay mineral diagenesis on the microfabric andpore-scale properties of deep-water Gulf of Mexico mudstones. Clays ClayMiner. 54 (4), 500e514.

Bjorlykke, K., 2013. Petroleum Geoscience e from Sedimentary Environments toRock Physics. Springer, p. 508. ISBN: 978-3-642-02331-6.

Boles, J.R., Franks, S.G., 1979. Clay diagenesis in Wilcox sandstones of SouthwestTexas; implications of smectite diagenesis on Sandstone cementation.J. Sediment. Res. 49 (1), 55e70. http://dx.doi.org/10.1306/212F76BC-2B24-11D7-8648000102C1865D.

Day-Stirrat, R.J., Aplin, A.C., �Srodo�n, J., van der Pluijm, B.A., 2008a. Diageneticreorientation of phyllosilicate minerals in Paleogene mudstones of the PodhaleBasin, Southern Poland. Clays Clay Miner. 56 (1), 100e111. http://dx.doi.org/10.1346/CCMN.2008.0560109.

Day-Stirrat, R.J., Loucks, R.G., Milliken, K.L., Hillier, S., van der Pluijm, B.A., 2008b.Phyllosilicate orientation demonstrates early timing of compactional stabiliza-tion in calcite-cemented concretions in the barnett shale (Late Mississippian),Fort Worth Basin, Texas (U.S.A). Sediment. Geol. 208 (1e2), 27e35. http://dx.doi.org/10.1016/j.sedgeo.2008.04.007.

Delle Piane, C., Dewhurst, D.N., Siggins, A.F., Raven, M.D., 2011. Stress-inducedanisotropy in brine saturated shale. Geophys. J. Int. 184 (2), 897e906. http://dx.doi.org/10.1111/j.1365-246X.2010.04885.x.

Delle Piane, C., Sarout, J., Madonna, C., Saenger, E.H., Dewhurst, D., Raven, M., 2014.Frequency-dependent seismic attenuation in shales: experimental results andtheoretical analysis. Geophys. J. Int. 198 (1), 504e515. http://dx.doi.org/10.1093/gji/ggu148.

Duddy, I.R., Moore, M.E., O'Brien, C., 2005. Thermal History Reconstruction in FiveCanning Basin Wells: Acacia-1, Acacia-2, Kidson-1, Willara-1 & Yulleroo-1,Based on Apatite Fission Track Analysis (AFTA®), Data Report. Petroleumexploration report G3 1772 A1. Western Australia Geological Survey.

Ghori, K.A.R., Haines, P.W., 2006. Petroleum Geochemistry of the Canning Basin,Western Australia: Basic Analytical Data 2004e05. Western Australia GeologicalSurvey, 2006/7.

G€otze, J., Schertl, H.P., Neuser, R.D., Kempe, U., Hanchar, J.M., 2013. Opticalmicroscope-cathodoluminescence (OMeCL) imaging as a powerful tool toreveal internal textures of minerals. Mineral. Petrol. 107 (3), 373e392. http://dx.doi.org/10.1007/s00710-012-0256-0.

Haines, P.W., 2004. Depositional Facies and Regional Correlations of the OrdovicianGoldwyer and Nita Formations, Canning Basin, Western Australia, with Impli-cations for Petroleum Exploration. Western Australia Geological Survey, p. 45.Record 2004/7.

Haines, P.W., 2009. The Carribuddy Group and Worral Formation, Canning Basin,Western Australia: Stratigraphy, Sedimentology, and Petroleum Potential.Geological Survey of Western Australia, p. 60. Report 105.

Ho, N.C., Peacor, D.R., van der Pluijm, B.A., 1999. Preferred orientation of phyllosi-licates in Gulf Coast mudstones and relation to the smectite-illite transition.Clays Clay Miner. 47 (4), 495e504.

Jones, G.D., Xiao, Y., 2005. Dolomitization, anhydrite cementation, and porosityevolution in a reflux System: insights from reactive transport models. AAPGBull. 89 (5), 577e601. http://dx.doi.org/10.1306/12010404078.

Kanitpanyacharoen, W., Wenk, H.R., Kets, F., Lehr, C., Wirth, R., 2011. Texture andanisotropy analysis of Qusaiba shales. Geophys. Prospect. 59 (3), 536e556.

Keller, L.M., Holzer, L., Wepf, R., Gasser, P., 2011. 3D geometry and topology of porepathways in opalinus clay: implications for mass transport. Appl. Clay Sci. 52(1e2), 85e95. http://dx.doi.org/10.1016/j.clay.2011.02.003.

Kocks, U.F., Tome', C.N., Wenk, H.-R., 2000. Texture and Anisotropy: Preferred Ori-entations in Polycrystals and Their Effect on Materials Properties. CambridgeUniversity Press, p. 692. ISBN: 978-0-521-79420-6.

Lash, G.G., Blood, D., 2004. Geochemical and textural evidence for Early (shallow)diagenetic growth of stratigraphically confined carbonate concretions, UpperDevonian Rhinestreet black shale, western New York. Chem. Geol. 206 (3e4),407e424.

Leeman, W.P., MacRae, C.M., Wilson, N.C., Torpy, A., Lee, C.T.A., Student, J.J.,Vicenzi, E.P., 2012. A study of cathodoluminescence and trace elementcompositional zoning in natural quartz from volcanic rocks: mapping titaniumcontent in quartz. Microsc. Microanal. 18 (06), 1322e1341. http://dx.doi.org/10.1017/S1431927612013426.

Lonardelli, I., Wenk, H.R., Ren, Y., 2007. Preferred orientation and elastic anisotropyin shales. Geophysics 72 (2), D33. http://dx.doi.org/10.1190/1.2435966.

Lyons, W.C., Plisga, G.J., 1984. Standard Handbook of Petroleum and Natural GasEngineering. Gulf Professional Publishing, MA, USA.

Matenaar, I.F., 2002. Compaction and Microfabric Rearrangement of Fine-grainedSiliciclastic Sediments. University of Newcastle upon Tyne (Ph.D. thesis).http://ethos.bl.uk/OrderDetails.do?uin¼uk.bl.ethos.394736.

Mavko, G., Mukerji, T., Dvorkin, J., 2009. The Rock Physics Handbook. Tools forSeismic Analysis of Porous Media, second ed. Cambridge University Press,p. 304. ISBN: 978-0-521-54344-6.

C. Delle Piane et al. / Marine and Petroleum Geology 59 (2015) 56e71 71

Mondol, N.H., Bjørlykke, K., Jahren, J., Høeg, K., 2007. Experimental mechanicalcompaction of Clay Mineral aggregatesdChanges in physical properties ofmudstones during burial. Mar. Pet. Geol. 24 (5), 289e311. http://dx.doi.org/10.1016/j.marpetgeo.2007.03.006.

Sarout, J., Molez, L., Gu�eguen, Y., Hoteit, N., 2007. Shale dynamic properties andanisotropy under triaxial loading. Experimental and theoretical investigations.Phys. Chem. Earth 32 (8e14), 896e906. http://dx.doi.org/10.1016/j.pce.2006.01.007.

Sintubin, M., 1994. Clay fabrics in relation to the burial history of shales. Sedi-mentology 41 (6), 1161e1169. http://dx.doi.org/10.1111/j.1365-3091.1994.tb01447.x.

Thomsen, L., 1986. Weak elastic anisotropy. Geophysics 51 (10), 1954e1966. http://dx.doi.org/10.1190/1.1442051.

Valcke, S.L.A., Casey, M., Lloyd, G.E., Kendall, J.M., Fisher, Q.J., August 1 2006. Latticepreferred orientation and seismic anisotropy in sedimentary rocks. Geophys. J.Int. 166 (2), 652e666. http://dx.doi.org/10.1111/j.1365-246X.2006.02987.x.

van der Pluijm, B.A., Ho, N.C., Peacor, D.R., 1994. High-resolution X-ray texturegoniometry. J. Struct. Geol. 16 (7), 1029e1032. http://dx.doi.org/10.1016/0191-8141(94)90084-1.

Wark, D.A., Watson, E.B., 2006. TitaniQ: a Titanium-in-quartz Geothermometer.Contrib. Mineral. Petrol. 152 (6), 743e754. http://dx.doi.org/10.1007/s00410-006-0132-1.

Wenk, H.R., 1985. Measurement of pole figures. In: Wenk, H.-R. (Ed.), PreferredOrientation in Deformed Metals and Rocks: an Introduction to Modern TextureAnalysis. Academic Press, London, pp. 11e48.

Wenk, H.-R., Voltolini, M., Mazurek, M., Van Loon, L.R., Vinsot, A., 2008. Preferredorientations and anisotropy in shales: Callovo-oxfordian shale (France) andopalinus clay (Switzerland). Clays Clay Miner. 56 (3), 285e306.

Wenk, H.R., Lonardelli, I., Franz, H., Nihei, K., Nakagawa, S., 2007. Preferred orien-tation and elastic anisotropy of illite-rich shale. Geophysics 72 (2), E69eE75.http://dx.doi.org/10.1190/1.2432263.

Wilson, N.C., MacRae, C.M., Torpy, A., Davidson, C.J., Vicenzi, E.P., 2012. Hyper-spectral cathodoluminescence examination of defects in a Carbonado diamond.Microsc. Microanal. 18 (06), 1303e1312. http://dx.doi.org/10.1017/S1431927612013578.

Zhan, Y., Mory, A.J., 2013. Structural interpretation of the northern Canning Basin,western Australia. In: West Australian Basins Symposium Perth, 18e21 August2013.