1

Thepermeabilityofstylolite-bearinglimestone1

2

MichaelHeap1,ThierryReuschlé1,PatrickBaud1,FrançoisRenard2,3,andGianlucaIezzi43

1GéophysiqueExpérimentale,InstitutdePhysiquedeGlobedeStrasbourg(UMR7516CNRS,Université4

deStrasbourg/EOST),5rueRenéDescartes,67084Strasbourgcedex,France.5

2PGP,TheNjordCentre,DepartmentsofGeosciences&Physics,UniversityofOslo,Norway6

3Université Grenoble Alpes, Université Savoie Mont Blanc, CNRS, IRD, IFSTTAR, ISTerre, 380007

Grenoble,France8

4Dipartimento di Ingegneria Geologia, IV Piano del Plz. Ex-Rettorato, Università degli Studi “G.9

d’Annunzio”,ViaDeiVestini30,66100Chieti,Italy.10

11

Correspondingauthor:M.Heap(heap@unistra.fr)12

13

Abstract14

Stylolitesareplanarfeaturesthatformduetointergranularpressuresolution.Duetotheir15

planargeometryandrelativeabundanceinlimestonereservoirs,theirimpactonregionalfluidflow16

has attracted considerable interest. We present laboratory permeability data that show that17

stylolites can be considered as conduits for flow in the stylolite-bearing limestonesmeasured. A18

combinationofanalysistechniquesshowsthatthisisduetoazonethatsurroundsthesestylolites19

thatismoreporousandcontainslargerporesthanthehostrock.Ourdataalsoshowthatthewater20

permeability of a sample containing a stylolite parallel to fluid flow is typically lower than its21

permeabilitytogas,explainedhereasaresultoftheexpansionofminoramountsofclayfoundin22

thestylolite,andthat,duetotheirmicrostructuralsimilarities,tectonicandsedimentarystylolites23

affect sample permeability similarly. Finally, we show that the permeability anisotropy that24

2

developsintherockmassduetothepresenceofsedimentarystylolitesmakesitappearasthough25

the stylolites are acting as barriers to fluid flow, and may explain the discrepancy between26

laboratorymeasurementsandfield-scaleobservations.Thisapproachcanprovideestimatesforthe27

equivalent permeability, and permeability anisotropy, for stylolite-bearing limestone reservoirs28

worldwide.29

30

Keywords:Stylolite; limestone;permeability;synchrotronX-raycomputedtomography;scanning31

electronmicroscopy32

33

3

Highlights34

35

• Stylolitesinlimestonesareconduitsforflow,notbarrierstoflow.36

• Stylolitesarecharacterisedbyzoneofhigherporositythanthehostrock.37

• Thehigh-porositystylolitezonecontainslargerporesthaninthehostrock.38

• Poreswithinstylolitesarelessspherical:stylolitescreatethehigh-porosityzoneduring39

theirformation.40

• Stylolitescreateapermeabilityanisotropythatmaymakethemfalselyappearasbarriersto41

flow. 42

4

1Introduction43

Stylolites are planes of insoluble minerals that form in rocks as soluble minerals are44

removedbypressuresolution(e.g.,ParkandSchot,1968;Wanless,1979;NennaandAydin,2011;45

Croizéetal.,2013;Toussaintetal.,2018).Theyarecommoninlimestonesduetotherelativelyhigh46

solubilityofcalcite(e.g.,Tondietal.,2006;FabriciusandBorre,2007;BenedictoandSchultz,2010;47

Smith et al., 2011; Rustichelli et al., 2012; Agosta et al., 2012; Laronne Ben-Itzhak et al., 2014;48

Rustichellietal.,2015;Martín-Martínetal.,2018),butarealsofoundinsandstones(Heald,1955;49

Walderhaug,1996;Bjørkumetal.,1998;WalderhaugandBjørkum,2003;Emmanueletal.,2010).50

Stylolitesformperpendiculartothemajorcompressivestressandarecommonlyfoundsub-parallel51

to bedding (formed by overburden stresses; “sedimentary stylolites”), but can form sub-52

perpendiculartobeddingduetotectonicstresses(“tectonicstylolites”;e.g.,RailsbackandAndrews,53

1995;Ebneretal.,2010a).Althoughmacroscopicallyplanar,stylolitesaremorphologicallyvariable54

onthemeso-andmicroscale(e.g.,KarczandScholz,2003;Renardetal.,2004;Schmittbuhletal.,55

2004;Rollandetal.,2012;LaronneBen-Itzhaketal.,2012;Rollandetal.,2014;Koehnetal.,2016).56

Theirroughnessisthoughttobeafunctionofthemagnitudeofthestressunderwhichtheyformed57

(e.g., Koehn et al., 2007; Ebner et al., 2009a; Koehn et al., 2012), the heterogeneity of the host58

material(e.g.,AndrewsandRailsback,1997;Brousteetal.,2007;Ebneretal.,2009b,2010b;Koehn59

etal.,2012),and/orthecompetitionbetweenlong-rangeelasticredistributionandsurfacetension60

forcesalongtheinterface(e.g.,Schmittbuhletal.,2004;Renardetal.,2004).61

Due to their macroscopically planar form, the influence of stylolites on fluid flow and62

reservoircompartmentalisationhasdrawnconsiderableinterest.Ahandfulofexperimentalstudies63

haveshownthatstylolitescanprovideconduitsforflow(Lindetal.,1994;MallonandSwarbrick,64

1998; Heap et al., 2014; Rustichelli et al., 2015), challenging paradigms that stylolites present65

barrierstofluidflow(Dunnington,1967;Nelson,1981;BurgessandPeter,1985;Koepnick,1987;66

Finkel and Wilkinson, 1990; Dutton and Willis, 1998; Alsharhan and Sadd, 2000). Heap et al.67

5

(2014),forexample,measuredthepermeabilityoflimestonesamplesthatcontainednostylolites,68

onestyloliteperpendiculartothe imposedfluid flow,oronestyloliteparallel to flow.Theyfound69

that samples containing stylolites parallel to flow were about an order of magnitude more70

permeable than the stylolite-freematerial. They concluded that this was likely due to a zone of71

enhanced porosity surrounding the stylolite, a conclusion supported by microstructural72

observations(CarozziandvonBergen,1987;RaynaudandCarrio-Schaffhauser,1992;vanGeetet73

al.,2000;Gringasetal.,2002;Padmanabhanetal.,2015).Thehigherporosityzonesurroundinga74

stylolite has also been shown to reduce the uniaxial compressive strength of a stylolite-bearing75

sample (Baudetal.,2016).However, althoughstylolites themselvesmayactas conduits for fluid76

flow(e.g.,Heapetal.,2014;Rustichellietal.,2015),wehighlightthattheyaretheby-productofa77

processwherebydissolvedmaterialsareoftenprecipitatedintotheporespaceoftheadjacentrock,78

therebyloweringtheporosity,andpresumablypermeability,relativetotheoriginalhostrock(see79

Toussaint et al. (2018) for a review). Therefore, formations containing abundant stylolites will80

likely be characterised by lower porosities and permeabilities than neighbouring stylolite-free81

formations.Indeed,stylolitedensityhasbeenmeasuredtobeinverselyproportionaltoporosityin82

somelimestoneformations(e.g.,AlsharhanandSadd,2000).83

WeextendthestudyofHeapetal.(2014)byprovidingnewporosity-permeabilitydatafor84

stylolite-bearing limestones. We also (1) compare the gas and water permeability of stylolite-85

bearing limestones and (2) compare the permeabilities of limestone samples containing86

sedimentary and tectonic stylolites. Our experimental data are supported by microstructural87

observations(scanningelectronmicroscope,SEM),multi-resolution(voxelsizeof6.27and0.7μm)88

synchrotronX-raycomputed tomography(CT),andestimationsof theaverageporeradiusof the89

flowpathusedbygasparticlesdeterminedusingtheKlinkenbergslipfactor.Finally,andusingour90

experimental data,we consider the “upscaled” permeability of a limestone rockmass containing91

stylolites.92

6

93

2Experimentalmaterialsandmethods94

We selected six stylolite-bearing limestones for this study: two from open quarries in95

Burgundy (France) (Corton limestone andComblanchien limestone, both Jurassic) and four from96

coresdrilledaroundtheANDRAUndergroundResearchLaboratorynearBure(France)(onefrom97

the Middle Jurassic “Dogger” series and three from the Late Jurassic Oxfordian stage). The98

porositiesandgaspermeabilitiesofsomeofthesamplesfromBurewerepreviouslypresentedin99

Heap et al. (2014). We provide here new water permeability data on these samples and new100

porosityandpermeability(gasandwater)dataforadditionalsamplestakenfromoneofthecores101

fromBure(fromtheLateJurassicOxfordianstage)andthesamplesfromBurgundy.102

WefirstquantifiedthemineralcontentofourexperimentalmaterialsusingX-raypowder103

diffraction (XRPD). Powdered samples of each of the limestoneswere ground for 10minutes in104

alcoholusinganagatepestleandmortar.TheXRPDanalyseswereperformedonpowderedmounts105

(usingnominallyzero-backgroundSisampleholdersand10-20mgofmaterial)usingaBrukerD-106

5005θ-2θBragg-BrentanodiffractometerequippedwithNi-filteredCuKαradiation.Theobtained107

XRPD patterns were first checked for their crystalline content using search-match comparisons108

with XRPD standards contained in the inorganic crystal structure database (ICSD). The XRPD109

patternswithmorethanonecrystallinephasewerethenrefinedusingthesoftwareEXPGUI-GSAS110

(Larson and Von Dreele 1994; Toby 2001). EXPGUI-GSAS uses the Rietveld method to derive111

crystallographic parameters and phase abundances (wt. %). More detailed descriptions of the112

RietveldrefinementmethodarereportedinIezzietal.(2004;2010andreferencestherein).XRPD113

analysiswasperformedon(1)stylolite-freematerialand(2)onsamplescuttocontainastylolite,114

butwithaslittlehostrockaspossible(inanattempttoidentifythemineralsformingthestylolite).115

In addition, minerals within the stylolites were also identified using energy-dispersive X-ray116

7

spectroscopy (EDS) during our SEM analyses. The mineral content for the Bure samples was117

previouslypresentedinHeapetal.(2014).118

ThefirstoftheOxfordianlimestones(O1;depth=159m)isaheterogeneousallochemical119

limestone that contains ooids, peloids, shell fragments, and fossil foraminifera within a micrite120

matrix (Figure1a). Theooids are typically0.1-0.25mm indiameter (Figure1a).Thepeloids are121

noticeablylargerthantheooids(Figure1a);somepeloidshavediametersgreaterthan1mm.Shell122

fragments inO1 can bemanymillimetres in length.O1 is predominately calcite (99wt.%)with123

subordinatedolomite(<1wt.%)(Table1).ThesecondOxfordianlimestone(O3;depth=174m)is124

awell-sortedallochemicallimestonethatcontainsooids,typically0.25-0.5mmindiameter,within125

amicritematrix(Figure1b).O3ispredominatelycalcite(99wt.%)withsubordinatedolomite(<1126

wt.%),gypsum(<1wt.%),andpyrite(<<1wt.%)(Table1).ThethirdOxfordian limestone(O6;127

depth = 364 m) is a very heterogeneous allochemical limestone that contains peloids, shell128

fragments (>1 mm), and fossil foraminifera (>1 mm) within a micrite matrix (Figure 1c). O6 is129

predominately calcite (99wt.%)with subordinate dolomite (<1wt.%) and pyrite (<<1wt.%)130

(Table 1). The “Dogger” limestone (D3; depth 747 m) is an orthochemical limestone (micrite)131

(Figure1d)composedof93wt.%calcite,4wt.%dolomite,3wt.%quartz,andsubordinatepyrite132

(<<1 wt. %) (Table 1). Corton limestone is an allochemical limestone that contains peloids133

(typically 0.2-1 mm in diameter) within a micrite matrix (Figure 1e). Corton limestone is134

predominately calcite (99 wt. %) with subordinate quartz (<1 wt. %) (Table 1). Comblanchien135

limestone isaheterogeneousallochemical limestonethatcontainsooids,peloids,shell fragments,136

andfossilforaminiferawithinamicritematrix(Figure1f).Allochemsaretypicallybetween0.1and137

1mmindiameter(Figure1f).Comblanchienlimestoneisessentiallyentirelycalciteincomposition138

(99wt.%)(Table1).139

Examples of the stylolites in these materials are shown in Figures 2, 3, 4, and 5.140

Qualitatively, therearecleardifferencesbetween thesedimentarystylolites in termsof thickness141

8

androughness/tortuosity(Figures2,3,and4).Thethickeststylolites(upto2-3mm)arefoundin142

theD3samples(Figure3a).ThestylolitesinCortonlimestonearethemostrough/tortuous(Figure143

4a);theleastrough/tortuousstylolitesarefoundinsamplesO3(Figure2b)andtheComblanchien144

limestone (Figure 4b). Stylolites found in the most heterogeneous limestone—sample 06—are145

correspondingly anastomosing (Figure 2c).We also provide images of tectonic and sedimentary146

stylolites found in sample O3 (Figure 5). There are no discernable differences between the147

thicknessandroughness/tortuositybetweenthetectonicandsedimentarystylolitesinsampleO3148

(Figure5).A combinationofXRPDandEDSanalyses found that the stylolites typically consistof149

dolomiteand/orquartz,withminorquantitiesofpyriteandorganicmatter/clay(Table1).150

Cylindricalsamples(20mmindiameterandnominally40mminlength)werecoredfrom151

theblocks/cores.Sampleswereprepared tocontain (1)onestyloliteperpendicular to theaxisof152

thecore(i.e.perpendicular to the imposed flowdirection), (2)onestyloliteparallel to theaxisof153

thecore(i.e.parallel to the imposed flowdirection),or (3)nostylolite (wherepossible, stylolite-154

free samples were prepared in two or three orthogonal directions). The samples containing no155

stylolites were typically prepared from material 5-10 cm from the stylolite studied. We also156

prepared samples containing tectonic stylolites either perpendicular or parallel to the core axis157

fromoneoftheOxfordianlimestones(sampleO3).Wenotethat,followingsamplepreparation,our158

samplesdidnotcontainanyobviousstylolite-associatedfractures.Representativephotographsof159

the 20 mm-diameter samples prepared for laboratory testing (stylolite-free, one stylolite160

perpendicular to the sample axis, and one stylolite parallel to the sample axis) are provided in161

Figures6and7.Figure8showsphotographsofsamplesofO3containingtectonicandsedimentary162

stylolites.163

The connected porosity of each sample was determined using the triple weight water164

saturationtechnique(GuéguenandPalciauskas,1994).Gas(argonornitrogen)andwater(distilled165

water) permeabilities were then measured in a hydrostatic pressure vessel under a confining166

9

pressureof2MPa.Allmeasurementsofwaterpermeabilitywereperformedusingthesteady-state167

flowmethod.Followingmicrostructuralequilibrium,apressuredifferentialwasimposedacrossthe168

sampleandtheflowratewasmeasuredusinganelectronicbalance(withaprecision±0.0005g).169

Oncesteady-stateflowhadbeenestablished,thewaterpermeability𝑘!"#$% wasdeterminedusing170

Darcy’srelation:171

172

𝑄𝐴= 𝑘!"#$%𝜂𝐿

𝑃!" − 𝑃!"#$ , (1)

173

where Q is the volumetric flow rate, A is the cross-sectional area of the sample, Pup and Pdown174

represent the upstream and downstream pressure, respectively (wherePdown is the atmospheric175

pressure),Listhelengthofthesample,𝑘!"#$% isthepermeabilitytowater,andηistheviscosityof176

theporefluid(takenhereas1.008×10-3Pas).Apressuredifferential(i.e. 𝑃!" − 𝑃!"#$)of0.5MPa177

wasusedforallmeasurementsreportedherein.178

Gas (argonornitrogen)permeabilitywasmeasuredusingeither the steady-statemethod179

(forhigh-permeabilitysamples)orthepulse-decaymethod(forlow-permeabilitysamples).Forthe180

steady-state method, a pressure differential was imposed across the sample (following181

microstructuralequilibrium)andtheoutlet flowratewasmeasuredusingaflowmeter.Sincethe182

pore fluid is compressible, the raw permeability to gas 𝑘!"#_!"# is expressed as (Scheidegger,183

1974):184

185

𝑄𝐴= 𝑘!"#_!"#𝜂𝐿

(𝑃!")! − (𝑃!"#$)! 2𝑃!"#$

, (2)

186

whereη,theviscosityoftheporefluid,wastakenas2.21×10-5and1.78×10-5Pasforargonand187

nitrogen,respectively.Steady-statevolumetricflowrateQmeasurementsweretakenunderseveral188

10

pore pressure differentials (i.e. 𝑃!" − 𝑃!"#$, where Pdown is the atmospheric pressure) to check189

whether any auxiliary corrections were required.We first plot 1/𝑘!"#_!"# as a function of𝑄 to190

check whether the Forchheimer correction is required (Forchheimer, 1901). The correction is191

necessaryifthesedatacanbewelldescribedbyalinearfitwithapositiveslope.TheForchheimer-192

correctedpermeabilityistakenastheinverseofthey-interceptofthebest-fitlinearregressionin193

the plot of1/𝑘!"#_!"# as a function of𝑄. If the Forchheimer correction is not required,we then194

check whether the Klinkenberg correction is required (Klinkenberg, 1941). To do so, we plot195

𝑘!"#_!"# as a function of the reciprocal mean pressure 1/𝑃! , where Pm is the mean pore fluid196

pressure (i.e. (𝑃!" + 𝑃!"#$)/2). TheKlinkenberg correction is required if these data can bewell197

describedbya linear fitwithapositiveslopeand, if true, theKlinkenberg-correctedpermeability198

canbetakenasthey-interceptofthebest-fitlinearregressionintheplotof𝑘!"#_!"# asafunctionof199

1/𝑃! . The Klinkenberg correctionwas required for all samplesmeasured using the steady-state200

method;theForchheimercorrectionwasnotrequired.201

Weusedthepulse-decaymethod(Braceetal.,1968)tomeasurethegaspermeabilityofthe202

low-permeabilitysamples.Followingmicrostructuralequilibriumatthetargetconfiningpressure,203

the decay of an initial pore pressure differential (𝑃!" − 𝑃!"#$ = 0.5 MPa, where Pdown is the204

atmospheric pressure) was monitored using a pressure transducer following the closure of the205

upstreampressure inlet. The gaspermeability𝑘!"#_!"# was thendeterminedusing the following206

relation:207

208

𝑘!"#_!"# = 2𝜂𝐿𝐴

𝑉!"

𝑃!"! − 𝑃!"#$! 𝑑𝑃!"𝑑𝑡

, (3)

209

whereVup is the volume of the upstream pore pressure circuit (7.8 × 10-6m3) and t is time. As210

before, we checked whether these data required any auxiliary corrections (the Forchheimer or211

11

Klinkenbergcorrection).TheKlinkenbergcorrectionwasrequiredforallsamplesmeasuredusing212

the pulse-decaymethod; the Forchheimer correctionwas not required. A detailed description of213

thesepermeabilitymethodsisavailableinHeapetal.(2017).214

215

3Results216

The gas permeability data for the stylolite-free limestones as a function of connected217

porosity are shown in Figure 9a. The data of Lind et al. (1994), a study that alsomeasured the218

permeability of stylolite-bearing carbonate rocks, are also included on Figure 9 because they219

preserveahigherporosity(porosity>0.2)thanthesamplesmeasuredherein.Ourdatashowthat220

gas permeability increases as connected porosity is increased, in accordance with previously221

published studies on the permeability of limestones (e.g., Ehrenberg et al., 2006; Zinszner and222

Pellerin,2007),and that there isnomeasurablepermeabilityanisotropy in thestudiedmaterials223

(Figure 9a contains data on samples cored in orthogonal directions, seeTable 2). The difference224

between permeability to gas and permeability to water in the stylolite-free samples appears to225

dependontheconnectedporosity:permeabilitytogascanbeafactorof4.5higheratlowporosity226

(porosity < 0.05) and the ratio between gas and water permeability is essentially unity at the227

highesttestedporosity(porosity~0.15)(Figure10a).228

The porosity-permeability data for the stylolite-free and stylolite-bearing (perpendicular229

andparalleltoflow)limestonesareshowninFigure9b,togetherwiththehigh-porosity(porosity>230

0.2)dataofLindetal.(1994).Ourdatashowthat(1)thepermeabilitiesofthesamplescontaining231

stylolitesperpendiculartothedirectionofflowaresimilartothoseofthestylolite-freesamplesand232

(2) the permeabilities of the samples containing stylolites parallel to the direction of flow are233

characterisedbypermeabilitieshigherthanthoseofthestylolite-freesamples(Figure9b).Indetail,234

we notice that larger differences between the permeability of the samples containing stylolites235

parallel to the direction of flow and the stylolite-free samples are observed at lower connected236

12

porosities (Figure 9b). For example, the permeability of stylolite-bearing Corton limestone237

(porosity~0.03) canbe twoor threeorders ofmagnitudehigher than the stylolite-freematerial238

(Figure9b).Theratioofgastowaterpermeabilityforallthesamplestested(includingstylolite-free239

samples and samples containing stylolites perpendicular and parallel to flow) is plotted as a240

functionofconnectedporosityinFigure10b.Asforthestylolite-freelimestones(Figure10a),high-241

porosity (porosity ~0.15) samples containing stylolites show little difference between gas and242

water permeability (Figure 10b). The gas permeabilities of the low-porosity samples containing243

stylolites are higher than their water permeabilities; this is especially true for the low-porosity244

samplescontainingstylolitesparalleltoflow(thedifferenceforonesampleismorethananorder245

ofmagnitude)(Figure10b).246

Our data also show that there is essentially no difference between the influence of247

sedimentaryandtectonicstylolitesonthepermeabilityofour limestonesamples(Figures9band248

10b;Table2).249

250

4Discussion251

252

4.1Barrierstoorconduitsforfluidflow?253

OurnewpermeabilitydataareinagreementwiththeconclusionofHeapetal.(2014)and254

Rustichelli et al. (2015): the stylolites measured are not barriers to flow, but conduits for flow255

(Figure9b).Heapetal.(2014)postulatedthatazoneofhigherporositysurroundsastyloliteand256

thatitisthishigh-porosityzonethatenhancesthecirculationoffluids,assuggestedbyCarozziand257

von Bergen (1987), Raynaud and Carrio-Schaffhauser (1992), and Van Geet et al. (2000). The258

greaterincreaseinpermeabilityinthelow-porositysamples(whencomparingthepermeabilityof259

a stylolite-free sample to a sample containing a stylolite parallel to flow) (Figure 9b) is likely a260

consequenceof their lowmatrixpermeabilities.Conduits for flowhaveamuchgreater impacton261

13

the equivalent permeability of low-porosity samples than on high-porosity samples, since the262

matrixpermeabilityofahigh-porositysampleismuchclosertothepermeabilityofthefracture(as263

observed in variably porous fractured materials; e.g., Heap and Kennedy, 2016; Kushnir et al.,264

2018).265

To image the hypothesised zone of higher porosity, we first provide a backscattered266

scanningelectronmicroscope(BSE) imageofastylolitewithinsampleD3,selecteddueto its low267

matrixporosityandpermeability.Thisimageshowsthatthematrix-styloliteinterfaceispopulated268

bynumerousmicropores, typicallyonlya fewmicrons indiameter (whitearrows;Figure11).To269

betterresolvetheporosity,anddistributionofporosity,aroundastylolite,wealsoprovidemulti-270

resolution(voxelsizeof6.27(beamlineMB05)and0.7μm(beamlineID19)andenergyof35keV)271

three-dimensional X-ray tomography imaging performed on a stylolite within sample D3 at the272

EuropeanSynchrotronRadiationFacility(Grenoble,France).Becauseofthehigh-contrastbetween273

theporosityandthemineralsthatcomprisetherock(primarilycalcite,quartz,anddolomite;Table274

1),itisstraightforwardtosegmenttheporositysothatindividualporescanbeimaged(Figure12).275

Thesegmentedimagesshowthatthestyloliteisassociatedwithazoneofhighporosity(Figure12),276

as previously measured by Baud et al. (2016). In particular, we observe that (1) the pores277

surroundingthestylolitearelargerthanthosewithinthehostrock(Figure12)and(2)someofthe278

pores are aligned with the teeth of the stylolite and are characterised by a “finger-like” shape279

(Figure13). Indeed, analysing theX-ray tomographydata (similar toX-ray tomographicanalyses280

performed on intact porous limestones by Ji et al., 2012; 2014) show that the pores are larger281

insidethestylolite(Figure14a)andthattheporeswithinthestylolitearecharacterisedbylower282

valuesofsphericity(where1.0isaperfectsphere;sphericityisdefinedusingtheThermoScientific283

Avizotoolboxas( !!!!"#!"!!

)!/!where𝑆ℎ𝑎𝑝𝑒!"!! = 𝐴𝑟𝑒𝑎!!!/(36 × 𝜋 × 𝑉𝑜𝑙𝑢𝑚𝑒!!!))(Figure14b).284

Thevolumeof an individualporewithin the stylolite varies froma fewμm3up to>105μm3; the285

volume of the pores outside the stylolite are all <104 μm3 (Figure 14a). The average equivalent286

14

diameterof thepores insideandoutside thestylolite is36.5μm(standarddeviationof26.8μm)287

and11.1μm(standarddeviationof4.7μm),respectively.Sphericityinsideandoutsidethestylolite288

varies from0.2to0.4and0.4and0.9,respectively(Figure14b).Sincetheshapeof theporesare289

sometimes linked to the shape of the stylolite (Figure 13), we additionally conclude that such290

porosity is likely the consequence of stylolite formation (in agreement with the conclusions of291

Raynaud and Carrio-Schaffhauser (1992) and Carozzi and von Bergen (1987)), rather than that292

stylolitesformpreferentiallyinazoneofhigherporosity(ashypothesisedbyBraithwaite,1989).293

Tocomplementthesemicrostructuraldata,weusetheKlinkenbergslipfactor,𝑏(whichhas294

theunitsofpressure;Table2)(Klinkenberg,1941), toprovidean independentassessmentof the295

average pore radius used by the gas molecules. Since the pore radius determined using this296

techniqueusesdatafrompermeabilityexperiments,itwillthereforeyieldtheaverageporethroat297

radius(incontrasttotheCTdata,whichprovides informationonthepores).Sincethemeanfree298

pathis inverselyproportionalto𝑃! ,Poiseuille's lawforgasflowinacylindricaltubeandDarcy's299

lawforflowinporousmediayieldsthefollowingrelation:300

301

𝑘!"# = 𝑘!"#_!"# 1 + 𝑏𝑃!

, (4)

302

where 𝑘!"# is the true (i.e. Klinkenberg-corrected) gas permeability (Klinkenberg, 1941). The303

averageporethroatradiusoftheflowpathfollowedbythegasmolecules,𝑟,canbeestimatedusing304

thefollowingrelation(Civan,2010):305

306

𝑟 = 4𝑏𝜂

𝜋𝑅!𝑇2𝑀!

, (5)

307

15

where𝑅!istheidealgasconstant(takenas8.31Jmol-1K-1),𝑇isthetemperature(takenas293K),308

and𝑀! isthemolarmassoftheporefluid(takenas0.03995and0.02802kgmol-1forargonand309

nitrogen, respectively). The Klinkenberg slip factor has previously been used to examine the310

averageporethroatradiusoftheflowpathinrockssuchasshales(e.g.,Helleretal.,2014;Firouzi311

etal.,2014;LethamandBustin,2016)and,morerecently,volcanicrocks(Heapetal.,2018)using312

thesame,orsimilar,method(i.e.Equation(5)).Wefindthattheaverageporethroatradiusofthe313

flow path followed by the gas molecules in the stylolite-free samples, excluding the Corton314

limestone samples, variesbetween~0.05and0.15μm (Figure15a).ExcludingCorton limestone,315

thesamplescontainingthehighestporosities(porosity~0.15)arecharacterisedbythelargestpore316

throat radii (~0.1 to ~0.15 μm; Figure 15a). The average pore throat radius of the flow path317

followedbythegasmoleculesismuchlargerinCortonlimestone,varyingbetween~0.2and0.35318

μm(Figure15a).Althoughnotobviousfromourmicrostructuralobservations(Figure1e),Corton319

limestone must contain larger pore throats than the other limestones measured herein. The320

averageporethroatradiioftheflowpathsfollowedbythegasmoleculesinthesamplescontaining321

stylolites(togetherwiththestylolite-freesamples)areprovidedinFigure15b.Thedatashowthat322

(1)theporethroatradiiformingtheflowpathinthesamplescontainingstylolitesperpendicularto323

flow are similar to the stylolite-free samples and (2) the pore throat radii along the flow path324

parallel to the stylolite are systematically larger than those of the other samples (Figure 15b).325

Thesedatasuggestthatthestylolitesareassociatedwithporethroatswithlargerradiithanthose326

that typify the host rock. This conclusion is in agreement with our X-ray tomography analysis,327

which shows that the pores are larger inside the stylolite than in the host rock (Figure 14a). As328

expected,theradiipredictedusingEquation(5)aremuchsmaller(typically<1μm;Figure15)than329

therangeofradiipredictedfromtheX-raytomographyanalysis(uptoafewtensofmicrons).This330

is because the Klinkenberg analysis (Equation 5) yields the pore throat radius and the X-ray331

tomographyanalysisyieldstheporeradius.332

16

Weconcludeherethatstylolitespresentconduitsfor,ratherthanbarriersto,flow(Figure333

9b)inlimestonesmeasuredherein.Thiscanbeexplainedbyazoneofelevatedporosity(Figure12)334

that contains pores andpore throatswith larger radii than thehost rock (Figures 14a and15a),335

which we conclude must develop around a stylolite during its formation. The development of336

stylolitic porosity is discussed in detail in Carozzi and von Bergen (1987) and is considered the337

resultofgrainscaleheterogeneitiesintherockduringthedissolutionprocess.338

339

4.2Differencesbetweentectonicandsedimentarystylolites340

Ourpermeabilitydatasuggest,forthematerialsstudiedherein,thatthereisessentiallyno341

difference between the influence of sedimentary and tectonic stylolites on the permeability of a342

stylolite-bearing sample: both sedimentary and tectonic stylolites are conduits for fluid flow343

(Figure9b;Table2).This isperhapsnotsurprisingsince theyareverysimilaron themicroscale344

(Figure 5) and on the sample lengthscale (Figure 8). The fact that tectonic stylolites are also345

conduits for flow(Table2) furthersupports thehypothesis thatstylolitescreateazoneofhigher346

porosity during their formation (e.g., Raynaud and Carrio-Schaffhauser, 1992; Carozzi and von347

Bergen,1987),ratherthanthattheyformpreferentiallyinhigherporositylayers(e.g.,Braithwaite,348

1989).349

350

4.3Differencesbetweengasandwaterpermeability351

Differencesbetweenpermeabilitytogasandwateraretypicallyobservedinthepresenceof352

swellingclays(e.g.,FaulknerandRutter,2000,2003;TanikawaandShimamoto2006;Davyetal.,353

2007;TanikawaandShimamoto2009;BehnsenandFaulkner,2011).OurXRPDanalyseshighlight354

that clays are below the detection limit in the stylolite-free material (Table 1). It is therefore355

perhapssurprisingthatweseeaboutafourfolddifferencebetweengasandwaterpermeabilityin356

the low-porosity limestones (Figure 10a). A recent study found that the permeability to gaswas357

17

higherthanpermeabilitytowater intwovolcanicrocks(basaltandandesite)byafactorofupto358

five (Heap et al., 2018). In the absence of significant physicochemical reactions, these authors359

suggestedthatthedifferenceingasandwaterpermeabilities is likelyduetowateradsorptionon360

thesurfaceofthinmicrostructuralelements.Forthestylolite-freelimestones,wefindthatthereis361

essentiallynodifferencebetweenthegasandwaterpermeabilitiesforthesamplescharacterisedby362

the largest average pore throat radii (~0.1 to ~0.15 μm) (Table 2), as determined using the363

Klinkenbergslipfactor.Sampleswithaverageporethroatradiibetween~0.05and0.1μmaremore364

permeabletogasthantowater(Table2).SimilartotheconclusionsdrawnbyHeapetal.(2018),365

weconcludeherethat,intheabsenceofclaywithintheintactmaterials(Table1),thedifferencein366

gasandwaterpermeabilitiesislikelyduetowateradsorptiononthesurfaceofthin(~0.05to0.1367

μm)microstructuralelements.368

Measurementsofgasandwaterpermeabilityonthesamplescontainingstylolitesshowthat369

samplescontainingstylolitesparalleltoflowareoftenmorepermeabletogasthanwater,byupto370

oneorderofmagnitude(Figure10b).Sinceaverageporethroatradiusoftheflowpathfollowedby371

thegasmoleculesisrelativelyhighforthesesamples(upto~1μm;Figure15b),weconcludethat372

thedifferenceingasandwaterpermeabilitiesinthesesamplesareduetominorquantitiesofclay373

foundwithinthestylolite(identifiedbyEDSduringourSEManalyses;Table1).Theexpansionof374

clay minerals in contact with water constricts pore throats and thus reduces permeability (e.g.,375

FaulknerandRutter,2003).376

377

4.4Implicationsforfluidflowinlimestonereservoirs378

LimestoneformsanimportantcomponentoftheEarth’scontinentalcrust(Ehrenbergetal.,379

2006;FordandWilliams,2013)and,asaresult,thepermeabilityoflimestonereservoirsisnotonly380

important for fluid flow and pore pressure distribution within the crust, but also for the381

exploitationofhydrocarbonreserves.382

18

Ourstudyshowsthatstylolitesinlimestonepresentconduitsforflow(Figure9b)duetoa383

zone of elevated porosity, containing poreswith larger radii than the host rock,which develops384

aroundastyloliteduringitsformation(Figures11,12,14,and15).Inordertoconsiderfluidflowin385

stylolite-bearing limestone reservoirs, we must first upscale our laboratory measurements. One386

method to upscale such laboratory data is to first extract the permeability of a stylolite. The387

permeabilityofastylolite,𝑘!"#$% ,canbedeterminedusingatwo-dimensionalmodelthatconsiders388

flowinparallellayers(thesamemodelusedtodeterminethepermeabilityofcompactionbandsin389

Vajdova et al. (2004) and fractures inHeap andKennedy (2016), Farquharson et al. (2016), and390

Kushniretal.(2018)):391

392

𝑘!"#$% = (𝐴 ∙ 𝑘!) − (𝐴!"#$%# ∙ 𝑘!)

𝐴!"#$% (6)

393

where 𝐴 is the cross-sectional area of the sample, 𝑘! is the equivalent permeability (the394

permeabilityofthestylolite-bearingsample),𝐴!"#$%# istheareaofstylolite-freematerial,𝑘! isthe395

stylolite-freepermeability,and𝐴!"#$%istheareaofthestylolite.Forthepurposeofthisexercise,we396

will consider a core of Dogger limestone (D3) taken from the ANDRA Underground Research397

Laboratory at Bure (Figure 16). To calculate 𝑘!"#$% we use the permeability of the stylolite-free398

sampleofD3(𝑘!=3.69×10-19m2;Table2).Theequivalentpermeability, 𝑘! ,ispermeabilityofthe399

D3samplecontainingastyloliteparalleltothedirectionofflow(𝑘! =5.98×10-18m2;Table2),and400

weuseastylolitethicknessof1mm(areasonableapproximationofthethicknessofthestylolitein401

this sample; Figure6d).Using these values, Equation (6) yields a stylolite permeability, 𝑘!"#$% , of402

8.85 × 10-17 m2.We can nowmodel the equivalent permeability of a rockmass populated with403

stylolitesusingourvaluefor𝑘!"#$%andthefollowingrelation:404

405

19

𝑘! = 𝑤!"#$%# ∙ 𝑘! + (𝑤!"#$% ∙ 𝑘!"#$%)

𝑊, (7)

406

where𝑤!"#$%#isthetotalwidthoftheintactmaterial,𝑤!"#$%isthetotalwidthofthestylolites,and407

𝑊 is the length of rock considered (𝑊 = 𝑤!"#$%# + 𝑤!"#$%). The Dogger limestone core sample408

showsthattherearefivesedimentarystylolitesoveralengthofabout25cm(i.e.astylolitedensity409

of20m-1)(Figure16).Althoughstylolitethicknessvaries(Figure16)wewill,forsimplicity,assume410

thatthethicknessofeachstyloliteis1mm.Therefore,accordingtoourmodel(Equation(7);𝑊=411

250mm;𝑤!"#$% = 5mm;𝑤!"#$%# = 245mm;𝑘! = 3.69 × 10-19m2;𝑘!"#$% = 8.85 × 10-17m2), the412

equivalent permeability parallel and perpendicular to bedding for the limestone core shown in413

Figure16is1.80×10-17and3.69×10-19m2,respectively.Inotherwords,the25cm-longsampleis414

50timesmorepermeableparallel tobedding thanperpendicular tobedding.Therefore,although415

stylolitesareconduitsforflow,theycancreateapermeabilityanisotropyinarockunitorreservoir416

thatmaymakeitappearthattheyformbarrierstofluidflow(becausepermeabilityperpendicular417

to bedding is lower than the permeability parallel to bedding) and may therefore explain the418

discrepancybetweenlaboratorymeasurements(thatsuggestthatstylolitesareconduits)andfield-419

scaleinvestigations(thatsuggestthatstylolitesarebarriers).Itisimportanttohighlightthatthese420

equivalent permeability estimates for a stylolite-bearing rockmass are just one snapshot in the421

porosity-permeabilityevolutionof this limestoneformation.Forexample, it is likelythat,priorto422

pressure-solution and the formation of the stylolites, the host rockwasmuchmore porous and423

more permeable. Although not considered in our simple model, we note that the presence of424

stylolites perpendicular to bedding (i.e. tectonic stylolites) will reduce such permeability425

anisotropy.Indeed,ifthenumberoftectonicstylolitesequalsthenumberofsedimentarystylolites,426

no permeability anisotropy will be observed. Although tectonic stylolites are not uncommon in427

limestonereservoirs(RailsbackandAndrews,1995;Ebneretal.,2010a;Figure8),itisdifficultto428

20

assess their density at theANDRAUndergroundResearch Laboratory atBure due to the drilling429

direction(perpendiculartobedding).Thesimplemethodpresentedabovecanbeeasilyadaptedto430

provide estimates for the equivalent permeability, and permeability anisotropy, for stylolite-431

bearing(bothsedimentaryandtectonic)limestonereservoirsworldwide.432

Althoughweconcludeherethatourstylolitesformconduitsforfluidflow,wecannotrule433

out that some stylolites, different to those measured here, may provide barriers to flow. For434

example, (1) stylolites may provide barriers to flow if they are characterised by thick and435

continuous layers of clay-rich material, (2) an abstract by Corwin et al. (1997) suggests that436

stylolites associatedwith a cemented zone could be of lower permeability than the surrounding437

host rock, and (3) the modelling of Koehn et al. (2016) suggests that stylolites with simple438

geometries(e.g.,“simplewave-liketype”)maybemorelikelytoprovidebarrierstoflow.Wealso439

highlight that, due to differences in mineral composition and microstructure, the influence of440

styloliteson thepermeabilityofsandstonemaydiffer fromtheir influenceon thepermeabilityof441

carbonaterocks(e.g.,WalderhaugandBjørkum,2003;Emmanueletal.,2010).442

Therefore, and although we provide laboratory measurements for the permeability of443

stylolite-bearing limestone fromsix formations collected from two locationswithinFrance,more444

laboratory measurements on stylolites that are characterised by thick and continuous layers of445

clay-richmaterial arenow required to further explore the role of stylolites on the regional-scale446

permeability of limestone reservoirs (as concluded by Bruna et al., 2018). Laboratory447

measurementsonstylolite-bearingsandstonesalsoofferaninterestingavenueforfutureresearch.448

449

5Conclusions450

Thesalientconclusionsofthisstudycanbesummarisedthusly:451

(1) Thestylolitesmeasuredhereinareconduitsforfluidflow,notbarrierstofluidflow.452

21

(2) The permeability of a stylolite-bearing sample is lowerwhenmeasuredwithwater than453

withgas.We interpret thishereas the resultof theexpansionofminorquantitiesof clay454

foundwithinthestylolite.Theexpansionofclaymineralsconstrictsporethroatsandthus455

reducespermeability.456

(3) Sedimentaryandtectonicstylolitesaffectsamplepermeabilitysimilarly.We interpret this457

asaresultoftheirsimilarmicrostructures.458

(4) X-raytomographydatashowthatthestylolitesaresurroundedbyazoneofhigherporosity459

that ischaracterisedbypores largerthanthose foundinthe intactmaterial.Thisexplains460

whythestylolitesmeasuredhereinareconduitsforfluidflow.461

(5) The presence of larger poreswithin the stylolite zone is supported by an analysis of the462

Klinkenberg slip factor, which highlights that the average pore throat radius of the flow463

pathfollowedislargerwhenthesamplecontainsastyloliteparalleltoflow.464

(6) X-raytomographydatashowthattheporeswithinthestylolitearemuchlesssphericalthan465

thoseof thehost rockand that theyaresometimesalignedwith the teethof thestylolite.466

Sincetheshapeoftheporesarelinkedtotheshapeofthestylolite,weconcludethatsuch467

porosity is likely the consequence of stylolite formation, rather than that stylolites form468

preferentiallyinazoneofhigherporosity.469

(7) Upscaling our laboratory measurements using a simple two-dimensional model that470

considers flow in parallel layers shows that the equivalent permeability of a stylolite-471

bearinglimestonerockmassishigherparalleltobeddingthanperpendiculartobedding.472

(8) Thepermeabilityanisotropythatdevelopsintherockmassduetothepresenceofstylolites473

makes it appear as though the stylolites are acting as barriers to fluid flow (since474

permeabilityperpendiculartobedding is lowerthanthepermeabilityparallel tobedding)475

and may explain the discrepancy between laboratory measurements and field-scale476

observations.477

22

478

Acknowledgements479

This research was partly funded by CNRS and ANDRA (FORPRO program) and the480

Norwegian Research Council (project ARGUS, grant 272217).Wewould like to thank Alexandra481

Rolland, SilvioMollo, GillesMorvan, andBertrandRenaudié.We thankGilles Jouillerot (Rocamat482

PierreNaturelle)forhishelpselectingthematerialsatComblanchienandElodieBollerforherhelp483

attheEuropeanSynchrotronRadiationFacility(beamlineID19).ThecommentsofEinatAharonov484

andoneanonymousreviewerhelpedimprovethismanuscript.485

486

Dataavailability487

Mostof thedataused inthisstudyareavailable inTables1and2.TheX-raytomography488

datamaybemadeavailableonrequest(toFrançoisRenard). 489

23

Figurecaptions490

491

Figure 1. Optical microscope images of the stylolite-free (host rock) material for the studied492

limestones.(a)SampleO1–OxfordianlimestonefromBure.(b)SampleO3–Oxfordianlimestone493

fromBure. (c) SampleO6–Oxfordian limestone fromBure. (d) SampleD3– “Dogger” limestone494

fromBure.(e)CortonlimestonefromBurgundy.(f)ComblanchienlimestonefromBurgundy.495

496

24

Figure2.Opticalmicroscopeimagesofthestylolites foundwithinstudiedlimestones.(a)Sample497

O1–Oxfordian limestonefromBure.(b)SampleO3–Oxfordian limestonefromBure.(c)Sample498

O6–OxfordianlimestonefromBure.499

500

25

Figure3.Opticalmicroscopeimagesofthestylolites foundwithinstudiedlimestones.(a)Sample501

O6–OxfordianlimestonefromBure.(b)SampleD3–“Dogger”limestonefromBure.502

503

504

26

Figure4.Opticalmicroscope imagesof thestylolites foundwithinstudied limestones. (a)Corton505

limestonefromBurgundy.(b)ComblanchienlimestonefromBurgundy.506

507

27

Figure5. Opticalmicroscope images of (a) a sedimentary stylolite and (b) a tectonic stylolite in508

sampleO3–OxfordianlimestonefromBure.509

510

511

512

513

28

Figure 6. Photographs of the cylindrical samples prepared for laboratory measurements. Three514

representativesamplesareshownforeachlithology:anintactsample(ontheleft),asamplewitha515

styloliteperpendiculartoflow(inthemiddle),andasamplewithastyloliteparalleltoflow(onthe516

right).(a)SampleO1–OxfordianlimestonefromBure.(b)SampleO3–Oxfordianlimestonefrom517

Bure. (c) SampleO6–Oxfordian limestone fromBure. (d) SampleD3– “Dogger” limestone from518

Bure.519

520

29

Figure 7. Photographs of the cylindrical samples prepared for laboratory measurements. Three521

representativesamplesareshownforeachlithology:anintactsample(ontheleft),asamplewitha522

styloliteperpendiculartoflow(inthemiddle),andasamplewithastyloliteparalleltoflow(onthe523

right).(a)CortonlimestonefromBurgundy.(b)ComblanchienlimestonefromBurgundy.524

525

526

527

30

Figure 8. Photographs of the cylindrical samples containing either a tectonic stylolite (the two528

samplesontheleft)orasedimentarystylolite(thetwosamplesontheright).Allsamplesarefrom529

sampleO3–OxfordianlimestonefromBure.530

531

532

533

534

31

Figure9. (a) Gas permeability (measuredunder a confining pressure of 2MPa) as a function of535

connectedporosityforintact(i.e.stylolite-free)limestone.Alldatapointsaboveaporosityof0.2are536

takenfromLindetal.(1994).(b)Gaspermeability(measuredunderaconfiningpressureof2MPa)537

asa functionof connectedporosity for limestone samples containingeithera styloliteparallel to538

floworastyloliteperpendiculartoflow.Thegaspermeabilitiesoftheintactsamples(i.e.thedataof539

panel(a))arealsoplottedinpanel(b).540

541

542

543

32

Figure10.(a)Theratioofgastowaterpermeabilityasafunctionofconnectedporosityforintact544

(i.e.stylolite-free)limestone.(b)Theratioofgastowaterpermeabilityasafunctionofconnected545

porosity for limestone samples containing either a stylolite parallel to flow or a stylolite546

perpendiculartoflow.Theratiosoftheintactsamples(i.e.thedataofpanel(a))arealsoplottedin547

panel(b).548

549

550

33

Figure11.Backscatteredscanningelectronmicroscopeimageofthestylolite-hostrockboundary551

(sampleD3 – “Dogger” limestone fromBure).Quartz grainswithin the stylolite are labelled. The552

whitearrowspointtoporosity(inblack).553

554

555

556

557

34

Figure12.Multi-resolutionX-raysynchrotronmicrotomographyimagesofastyloliteinsampleD3558

(“Dogger” limestone from Bure). (a) An image of the stylolite at a voxel size of 6.27 μm. The559

coloured shapes are pores; individual pores are allocated different colours. (b) An image of the560

stylolite at a voxel size of 0.7 μm. The coloured shapes are pores; individual pores are allocated561

differentcolours.562

563

564

35

Figure 13. X-ray synchrotron microtomography image showing two pores within a stylolite in565

sampleD3 (“Dogger” limestone fromBure). The pores are “finger-like” in shape and are aligned566

withtheteethofthestylolite.567

568

569

570

571

36

Figure 14. X-ray synchrotronmicrotomography data showing (a) the number of poreswithin a572

subvolume of 0.16mm3 as a function of pore size inside and outside of a stylolite (sample D3 –573

“Dogger”limestonefromBure)and(b)theprobabilitydensityfunctionasafunctionofsphericity574

for thepores inside andoutside of a stylolite (sampleD3 – “Dogger” limestone fromBure). Two575

subvolumes of sample D3 with the same volume of 0.16 mm3, one inside the stylolite and one576

outsideofit,wereusedtoperformthesecalculations.577

578

579

580

37

Figure15.(a)Theaverageporeradiusoftheflowpathfollowedbythegasmolecules(calculated581

usingtheKlinkenbergslipfactor;seeEquation5)asafunctionofconnectedporosityfortheintact582

(i.e. stylolite-free) samples. (a) The average pore radius of the pores used by the gas particles583

(calculatedusingtheKlinkenbergslipfactor;seeEquation5)asafunctionofconnectedporosityfor584

limestonesamplescontainingeitherastyloliteparalleltofloworastyloliteperpendiculartoflow.585

Theaverageporeradiusof the intactsamples(i.e. thedataofpanel(a))arealsoplotted inpanel586

(b).587

588

589

38

Figure16.Aphotographof a78mm-diameter core fromBure (sampleD3– “Dogger” limestone590

fromBure).Arrowsindicatethepositionofsedimentarystylolites.591

592

39

Tables593

594

Table1. X-raypowderdiffraction (XRPD)analysis showingquantitativemineral composition for595

thesixlimestonesstudiedherein.MineralcontentsofO1,O3,O6,andD3weretakenfromHeapet596

al.(2014).597

Sample Stylolite-freecomposition

(wt.%)

Mineralswithinthestylolite

O1(Oxfordianlimestone) 99%calcite;<1%dolomite dolomite,clay

O3(Oxfordianlimestone) 99%calcite,<1%dolomite;

<1%gypsum;<<1%pyrite

dolomite,gypsum,pyrite,clay

O6(Oxfordianlimestone) 99%calcite,<1%dolomite;

<<1%pyrite

dolomite,pyrite,clay

D3(“Dogger”limestone) 93%quartz;4%dolomite;3%

quartz;<<1%pyrite

dolomite,quartz,pyrite,clay

Cortonlimestone 99%calcite;<1%quartz quartz,clay

Comblanchienlimestone 99%calcite clay

598

599

40

Table 2. Summary of the experimental data collected for this study. Porosities were measured600

using the tripleweightwater saturation technique.Gasandwaterpermeabilitiesweremeasured601

under a confining pressure of 2 MPa. Gas permeabilities were measured with either argon or602

nitrogengas.Waterpermeabilitiesweremeasuredwithdeionisedwater.Asteriskindicatesthatthe603

connectedporosityandthegaspermeabilitydataweretakenfromHeapetal.(2014).604

Sample Description Connectedporosity

Gaspermeability

(m2)

Klinkenbergslipfactor(MPa)

Waterpermeability

(m2)

Gas/waterpermeability

*O1 Nostyloperp 0.154 7.77×10-17 0.189 7.22×10-17 1.08

*O1 Sedstylopara 0.168 3.29×10-16 0.102 1.81×10-16 1.82

*O1 Sedstyloperp 0.166 1.09×10-16 0.206 1.07×10-16 1.02

*O3 Nostyloperp 0.150 2.25×10-17 0.309 2.37×10-17 0.95

*O3 Nostylopara 0.159 2.20×10-17 0.286 2.54×10-17 0.87

*O3 Sedstyloperp 0.169 4.10×10-17 0.278 4.51×10-17 0.91

*O3 Sedstylopara 0.175 4.65×10-17 0.233 4.78×10-17 0.97

O3 Nostylopara 0.076 1.60×10-18 0.449 8.72×10-19 1.83O3 Nostylopara 0.088 3.63×10-18 0.458 1.98×10-18 1.83

O3 Nostyloperp 0.084 2.05×10-18 0.568 1.08×10-18 1.90

O3 Tectstyloperp 0.075 2.14×10-18 0.452 1.16×10-18 1.84

O3 Tectstylopara 0.080 2.18×10-17 0.079 9.73×10-18 2.24

O3 Tectstylopara 0.082 2.62×10-17 0.078 6.35×10-18 4.13

O3 Tectstylopara 0.077 2.17×10-17 0.089 7.17×10-18 3.03

O3 Sedstyloperp 0.087 2.32×10-18 0.621 1.12×10-18 2.07

O3 Sedstyloperp 0.087 2.61×10-18 0.537 1.51×10-18 1.73

O3 Sedstylopara 0.096 1.75×10-17 0.196 1.34×10-17 1.31

O3 Sedstylopara 0.095 1.84×10-17 0.195 1.15×10-17 1.60

*O6 Nostyloperp 0.067 3.04×10-18 0.335 7.62×10-19 3.99

*O6 Sedstylopara 0.070 5.36×10-17 0.102 3.72×10-18 14.41

*O6 Sedstylopara 0.092 6.58×10-17 0.111 1.07×10-17 6.15

*O6 Sedstylo 0.084 1.51×10-17 0.241 6.04×10-18 2.50

41

perp

*O6 Sedstyloperp 0.086 1.43×10-17 0.252 4.98×10-18 2.87

*O6 Sedstyloperp 0.085 1.39×10-17 0.217 2.73×10-18 5.09

*D3 Nostyloperp 0.034 4.38×10-19 0.557 1.32×10-19 3.32

*D3 Nostylopara 0.034 3.69×10-19 0.447 8.11×10-20 4.55

*D3 Sedstyloperp 0.037 4.88×10-19 0.398 1.61×10-19 3.03

*D3 Sedstyloperp 0.032 3.44×10-19 0.521 7.55×10-20 4.56

*D3 Sedstyloperp 0.029 1.56×10-19 0.847 6.83×10-20 2.28

*D3 Sedstylopara 0.040 5.98×10-18 0.089 7.01×10-19 8.53

COMB Nostylopara 0.031 1.64×10-18 0.270 - -COMB Nostylopara 0.026 7.20×10-19 0.335 - -

COMB Sedstylopara 0.037 1.54×10-17 0.118 - -

COMB Sedstylopara 0.034 3.20×10-18 0.180 - -

COMB Sedstylopara 0.035 4.64×10-18 0.140 - -

COMB Nostyloperp 0.021 2.18×10-19 0.298 - -

COMB Sedstyloperp 0.027 6.95×10-19 0.164 - -

CORT Nostylopara 0.026 1.28×10-18 0.092 - -CORT Nostylopara 0.028 1.07×10-18 0.099 - -

CORT Sedstylopara 0.033 1.98×10-16 0.024 - -

CORT Sedstylopara 0.030 5.22×10-17 0.040 - -

CORT Sedstylopara 0.031 1.31×10-17 0.078 - -

CORT Nostyloperp 0.026 7.85×10-19 0.079 - -

CORT Nostyloperp 0.029 6.58×10-19 0.142 - -

CORT Sedstyloperp 0.027 7.48×10-19 0.085 - -

CORT Sedstyloperp 0.031 9.45×10-20 0.433 - -

605

606

42

References607

608Agosta,F.,Ruano,P.,Rustichelli,A.,Tondi,E.,Galindo-Zaldívar,J.anddeGaldeano,C.S.2012.Inner609

structure and deformation mechanisms of normal faults in conglomerates and carbonate610grainstones(GranadaBasin,BeticCordillera,Spain):inferencesonfaultpermeability.Journal611ofStructuralGeology,45,4-20.612

Alsharhan, A., and Sadd, J.L. 2000. Stylolites in Lower Cretaceous carbonate reservoirs, U.A.E:613SocietyforSedimentaryGeologySpecialPublication,69,185–207.614

Andrews,L.M.andRailsback,L.B.1997.Controlsonstylolitedevelopment:morphologic,lithologic,615and temporal evidence from bedding-parallel and transverse stylolites from the US616Appalachians.TheJournalofGeology,105(1),59-73.617

Baud,P.,Rolland,A.,Heap,M.,Xu,T.,Nicolé,M.,Ferrand,T.,&Conil,N.2016.Impactofstyloliteson618themechanicalstrengthoflimestone.Tectonophysics,690,4-20.619

Behnsen,J.,&Faulkner,D.R.2011.Waterandargonpermeabilityofphyllosilicatepowdersunder620medium to high pressure. Journal of Geophysical Research: Solid Earth, 116(B12), B12203,621doi:10.1029/2011JB008600622

Benedicto, A., & Schultz, R. A. 2010. Stylolites in limestone: magnitude of contractional strain623accommodatedandscalingrelationships.JournalofStructuralGeology,32(9),1250-1256.624

Bjørkum, P.A., Oelkers, E.H., Nadeau, P.H., Walderhaug, O. and Murphy, W.M. 1998. Porosity625prediction in quartzose sandstones as a function of time, temperature, depth, stylolite626frequency,andhydrocarbonsaturation.AAPGbulletin,82(4),637-648.627

Braithwaite,C.J.R.1989.Stylolitesasopenfluidconduits.MarineandPetroleumGeology,6(1),93-62896.629

Brouste,A.,Renard,F.,Gratier,J.P.andSchmittbuhl,J.2007.Varietyofstylolites'morphologiesand630statisticalcharacterizationoftheamountofheterogeneitiesintherock.JournalofStructural631Geology,29(3),422-434.632

Burgess, C. J., & Peter, C. K. 1985. Formation, distribution, and prediction of stylolites as633permeability barriers in the Thamama Group, Abu Dhabi. In Middle East Oil Technical634ConferenceandExhibition.SocietyofPetroleumEngineers.635

Brace,W.,Walsh,J.B.,&Frangos,W.T.1968.Permeabilityofgraniteunderhighpressure.Journalof636Geophysicalresearch,73(6),2225-2236.637

Bruna, P. O., Lavenu, A. P., Matonti, C., & Bertotti, G. (2018). Are stylolites fluid-flow efficient638features?JournalofStructuralGeology,doi:10.1016/j.jsg.2018.05.018.639

Carozzi, A. V., & Bergen, D. V. 1987. Stylolitic porosity in carbonates: a critical factor for deep640hydrocarbonproduction.JournalofPetroleumGeology,10(3),267-282.641

Civan,F.2010.Effectivecorrelationofapparentgaspermeabilityintightporousmedia.Transport642inporousmedia,82(2),375-384.643

Corwin, L.W., Broomhall, R.W., Saidikowski, R.M., &Wooten, J. N. (1997). Stylolites impact the644miscible nitrogen flood in a mature carbonate oil field. In:Middle East Oil Show and645Conference,SocietyofPetroleumEngineers,doi:doi.org/10.2118/37780-MS.646

Croizé, D., Renard, F. andGratier, J.P. 2013. Compaction and porosity reduction in carbonates: A647reviewofobservations,theory,andexperiments.AdvancesinGeophysics,54,181-238.648

Davy,C.A.,Skoczylas,F.,Barnichon,J.D.,&Lebon,P.2007.Permeabilityofmacro-crackedargillite649under confinement: gas andwater testing.Physics and Chemistry of the Earth, Parts A/B/C,65032(8-14),667-680.651

Dunnington,H.V.1967.Aspectsofdiagenesisandshapechangeinstyloliticlimestonereservoirs.In6527thWorldPetroleumCongress.WorldPetroleumCongress.653

43

Dutton, S. P., &Willis, B. J. 1998. Comparison of outcrop and subsurface sandstone permeability654distribution, lower cretaceous fall river formation, South Dakota andWyoming. Journal of655SedimentaryResearch,68(5),890-900.656

Ebner, M., Koehn, D., Toussaint, R., Renard, F. and Schmittbuhl, J. 2009a. Stress sensitivity of657stylolitemorphology.EarthandPlanetaryScienceLetters,277(3),394-398.658

Ebner,M.,Koehn,D.,Toussaint,R.andRenard,F.2009b.Theinfluenceofrockheterogeneityonthe659scalingpropertiesofsimulatedandnaturalstylolites.JournalofStructuralgeology,31(1),72-66082.661

Ebner,M.,Piazolo,S.,Renard,F.andKoehn,D.2010a.Stylolite interfacesandsurroundingmatrix662material:Natureandroleofheterogeneitiesinroughnessandmicrostructuraldevelopment.663JournalofStructuralGeology,32(8),1070-1084.664

Ebner, M., Toussaint, R., Schmittbuhl, J., Koehn, D. and Bons, P. 2010b. Anisotropic scaling of665tectonicstylolites:Afossilizedsignatureofthestressfield?.JournalofGeophysicalResearch:666SolidEarth,115(B6),B06403,doi:10.1029/2009JB006649.667

Emmanuel, S.,Ague, J. J.,&Walderhaug,O. (2010). Interfacial energyeffects and the evolutionof668poresizedistributionsduringquartzprecipitationinsandstone.GeochimicaetCosmochimica669Acta,74(12),3539-3552.670

Fabricius,I.L.andBorre,M.K.2007.Stylolites,porosity,depositionaltexture,andsilicatesinchalk671faciessediments.OntongJavaPlateau–GormandTyrafields,NorthSea.Sedimentology,54(1),672183-205.673

Farquharson, J. I., Heap, M. J., Lavallée, Y., Varley, N. R., & Baud, P. 2016. Evidence for the674development of permeability anisotropy in lava domes and volcanic conduits. Journal of675VolcanologyandGeothermalResearch,323,163-185.676

Faulkner,D.R.,&Rutter,E.H.2000.Comparisonsofwaterandargonpermeabilityinnaturalclay-677bearingfaultgougeunderhighpressureat20°C.JournalofGeophysicalResearch:SolidEarth,678105(B7),16415-16426.679

Faulkner,D.R.,&Rutter, E.H. 2003. The effect of temperature, thenature of the pore fluid, and680subyield differential stress on the permeability of phyllosilicate-rich fault gouge. Journal of681GeophysicalResearch:SolidEarth,108(B5),2227,doi:10.1029/2001JB001581.682

Finkel, E. A., &Wilkinson, B. H. 1990. Stylolitization as Source of Cement inMississippian Salem683Limestone,West-CentralIndiana(1).AAPGBulletin,74(2),174-186.684

Firouzi, M., Alnoaimi, K., Kovscek, A., & Wilcox, J. 2014. Klinkenberg effect on predicting and685measuringheliumpermeability ingasshales. International JournalofCoalGeology,123,62-68668.687

Forchheimer,P.(1901).Wasserbewegungdurchboden.Z.Ver.Deutsch,Ing.,45,1782-1788.688Ford,D.,&Williams,P.D.2013.KarstHydrogeologyandGeomorphology.JohnWiley&Sons.689Gingras,M.K.,MaCmillan,B.,&Balcom,B.J.2002.Visualizingtheinternalphysicalcharacteristicsof690

carbonatesedimentswithmagneticresonanceimagingandpetrography.BulletinofCanadian691PetroleumGeology,50(3),363-369.692

Heald,M.T.,1955.Stylolitesinsandstones.TheJournalofGeology,63(2),101-114.693Heap,M.J.,Baud,P.,Reuschlé,T.andMeredith,P.G.2014.Stylolitesinlimestones:Barrierstofluid694

flow?.Geology,42(1),51-54.695Heap, M. J., & Kennedy, B. M. 2016. Exploring the scale-dependent permeability of fractured696

andesite.EarthandPlanetaryScienceLetters,447,139-150.697Heap,M.J.,Kushnir,A.R.,Gilg,H.A.,Wadsworth,F.B.,Reuschlé,T.,&Baud,P.2017.Microstructural698

and petrophysical properties of the Permo-Triassic sandstones (Buntsandstein) from the699Soultz-sous-Forêtsgeothermalsite(France).GeothermalEnergy,5(1),26.700

Heap,M.J.,Reuschlé,T.,Farquharson,J.I.,&Baud,P.2018.Permeabilityofvolcanicrockstogasand701water.JournalofVolcanologyandGeothermalResearch,354,29-38.702

44

Heller,R.,Vermylen,J.,&Zoback,M.2014.Experimentalinvestigationofmatrixpermeabilityofgas703shales.AAPGBulletin,98(5),975-995.704

Iezzi,G.,Ventura,G.D.,Oberti,R.,Cámara,F.,&Holtz,F.2004.Synthesisandcrystal-chemistryofNa705(NaMg)Mg5Si8O22(OH)2,aP21/mamphibole.AmericanMineralogist,89(4),640-646.706

Iezzi,G.,DellaVentura,G.,Tribaudino,M.,Nemeth,P.,Margiolaki,I.,Cavallo,A.,.&Behrens,H.2010.707Phase transition induced by solid solution: The BCa-BMg substitution in richteritic708amphiboles.AmericanMineralogist,95(2-3),369-381.709

Ji, Y., Baud, P., Vajdova, V., & Wong, T.-f. 2012. Characterization of pore geometry of Indiana710limestoneinrelationtomechanicalcompaction.Oil&GasScienceandTechnology–Revued’IFP711EnergiesNouvelles,67(5),753-775.712

Ji, Y., Hall, S. A., Baud, P., & Wong, T.-f. 2014. Characterization of pore structure and strain713localization in Majella limestone by X-ray computed tomography and digital image714correlation.GeophysicalJournalInternational,200(2),701-719.715

Karcz,Z.andScholz,C.H.2003.ThefractalgeometryofsomestylolitesfromtheCalcareMassiccio716Formation,Italy.JournalofStructuralGeology,25(8),1301-1316.717

Klinkenberg, L. J. (1941). The permeability of porousmedia to liquids and gases. InDrilling and718productionpractice.AmericanPetroleumInstitute.719

Koehn, D., Renard, F., Toussaint, R. and Passchier, C.W. 2007. Growth of stylolite teeth patterns720depending on normal stress and finite compaction. Earth and Planetary Science Letters,721257(3),582-595.722

Koehn, D., Ebner, M., Renard, F., Toussaint, R. and Passchier, C.W. 2012. Modelling of stylolite723geometriesandstressscaling.EarthandPlanetaryScienceLetters,341,104-113.724

Koehn,D.,Rood,M.P.,Beaudoin,N.,Chung,P.,Bons,P.D.,&Gomez-Rivas,E.2016.Anewstylolite725classificationschemetoestimatecompactionandlocalpermeabilityvariations.Sedimentary726Geology,346,60-71.727

Koepnick, R. B. 1987. Distribution and permeability of stylolite-bearing horizonswithin a Lower728CretaceouscarbonatereservoirintheMiddleEast.SPEFormationEvaluation,2(02),137-142.729

Kushnir,A.R.,Heap,M.J.,&Baud,P.2018.Assessingtheroleoffracturesonthepermeabilityofthe730Permo-Triassic sandstones at the Soultz-sous-Forêts (France) geothermal site.Geothermics,73174,181-189.732

LaronneBen-Itzhak, L.L., Aharonov, E., Toussaint,R. and Sagy,A. 2012.Upperboundon stylolite733roughness as indicator for amount of dissolution.Earth and Planetary Science Letters,337,734186-196.735

LaronneBen-Itzhak, L.L., Aharonov, E., Karcz, Z., Kaduri,M. andToussaint, R. 2014. Sedimentary736stylolite networks and connectivity in Limestone: Large-scale field observations and737implicationsforstructureevolution.JournalofStructuralGeology,63,106-123.738

Larson,A.C.,&VonDreele,R.B.1994.GSAS.ReportlAUR,86-748.739Letham,E.A.,&Bustin,R.M.2016.Theimpactofgasslippageonpermeabilityeffectivestresslaws:740

Implications for predicting permeability of fine-grained lithologies. International Journal of741CoalGeology,167,93-102.742

Lind, I., Nykjaer, O., Priisholm, S. and Springer, N. 1994. Permeability of stylolite-bearing chalk.743JournalofPetroleumTechnology,46(11),986-993.744

Mallon,A.J.andSwarbrick,R.E.2008.Diageneticcharacteristicsoflowpermeability,non-reservoir745chalksfromtheCentralNorthSea.MarineandPetroleumGeology,25(10),1097-1108.746

Martín-Martín,J.D.,Gomez-Rivas,E.,Gómez-Gras,D.,Travé,A.,Ameneiro,R.,Koehn,D.,&Bons,P.D.7472018.Activationofstylolitesasconduitsforoverpressuredfluidflowindolomitizedplatform748carbonates.GeologicalSociety,London,SpecialPublications,459(1),157-176.749

Nelson,R.A.1981.Significanceoffracturesetsassociatedwithstylolitezones:geologicnotes.AAPG750Bulletin,65(11),2417-2425.751

45

Nenna,F.andAydin,A.2011.Theformationandgrowthofpressuresolutionseamsinclasticrocks:752Afieldandanalyticalstudy.JournalofStructuralGeology,33(4),633-643.753

Padmanabhan, E., Sivapriya, B., Huang, K. H., Askury, A. K., & Chow, W. S. 2015. The impact of754stylolites and fractures in defining critical petrophysical and geomechanical properties of755somecarbonaterocks.GeomechanicsandGeophysics forGeo-EnergyandGeo-Resources,1(1-7562),55-67.757

Park,W.C.andSchot,E.H.1968.Stylolites:theirnatureandorigin.JournalofSedimentaryResearch,75838(1),175-191.759

Railsback,L.B.andAndrews,L.M.1995.Tectonicstylolitesinthe‘undeformed’CumberlandPlateau760ofsouthernTennessee.JournalofStructuralGeology,17(6),911-915.761

Raynaud, S., & Carrio-Schaffhauser, E. 1992. Rock matrix structures in a zone influenced by a762stylolite.JournalofStructuralGeology,14(8-9),973-980.763

Renard, F., Schmittbuhl, J., Gratier, J.P., Meakin, P. and Merino, E. 2004. Three-dimensional764roughness of stylolites in limestones. Journal of Geophysical Research: Solid Earth,109(B3),765B03209,doi:10.1029/2003JB002555.766

Rolland,A.,Toussaint,R.,Baud,P., Schmittbuhl, J., Conil,N.,Koehn,D.,Renard,F. andGratier, J.P.7672012.Modelingthegrowthofstylolitesinsedimentaryrocks.JournalofGeophysicalResearch:768SolidEarth,117(B6),B06403,doi:10.1029/2011JB009065.769

Rolland,A.,Toussaint,R.,Baud,P.,Conil,N.,&Landrein,P.2014.Morphologicalanalysisofstylolites770forpaleostressestimationinlimestones.InternationalJournalofRockMechanicsandMining771Sciences,67,212-225.772

Rustichelli,A.,Tondi,E.,Agosta,F.,Cilona,A.andGiorgioni,M.2012.Developmentanddistribution773of bed-parallel compaction bands and pressure solution seams in carbonates (Bolognano774Formation,MajellaMountain,Italy).JournalofStructuralGeology,37,181-199.775

Rustichelli,A.,Tondi,E.,Korneva, I.,Baud,P.,Vinciguerra,S.,Agosta,F.,Reuschlé,T.and Janiseck,776J.M.2015.Bedding-parallel stylolites in shallow-water limestonesuccessionsof theApulian777CarbonatePlatform(central-southernItaly).ItalianJournalofGeosciences,134(3),513-534.778

Scheidegger,A.E.1974.ThePhysicsofFlowthroughPorousMedia(No.532.5S21974).779Schmittbuhl,J.,Renard,F.,Gratier,J.P.andToussaint,R.2004.Roughnessofstylolites:implications780

of3Dhighresolutiontopographymeasurements.PhysicalReviewLetters,93(23),p.238501.781Smith,S.A.,Billi,A.,DiToro,G.,&Spiess,R.2011.Principalslipzonesinlimestone:microstructural782

characterizationand implications for theseismiccycle(TreMontiFault,CentralApennines,783Italy).PureandAppliedGeophysics,168(12),2365-2393.784

Tanikawa,W.,&Shimamoto,T.2006.Klinkenbergeffectforgaspermeabilityanditscomparisonto785water permeability for porous sedimentary rocks. Hydrology and Earth System Sciences786Discussions,3(4),1315-1338.787

Tanikawa,W.,&Shimamoto,T.2009.ComparisonofKlinkenberg-correctedgaspermeabilityand788waterpermeabilityinsedimentaryrocks.InternationalJournalofRockMechanicsandMining789Sciences,46(2),229-238.790

Toby,B.H. 2001.EXPGUI, a graphical user interface forGSAS. Journal ofAppliedCrystallography,79134(2),210-213.792

Tondi, E., Antonellini, M., Aydin, A., Marchegiani, L. and Cello, G. 2006. The role of deformation793bands, stylolites and sheared stylolites in fault development in carbonate grainstones of794MajellaMountain,Italy.JournalofStructuralGeology,28(3),376-391.795

Toussaint, R., Aharonov, E., Koehn, D., Gratier, J. P., Ebner, M., Baud, P., ... & Renard, F. (2018).796Stylolites:Areview.JournalofStructuralGeology,114,163-195.797

Vajdova,V., Baud, P.,&Wong,T. F. 2004. Permeability evolutionduring localizeddeformation in798Bentheim sandstone. Journal of Geophysical Research: Solid Earth, 109(B10), B10406,799doi:10.1029/2003JB002942.800

46

Van Geet, M., Swennen, R., & Wevers, M. 2000. Quantitative analysis of reservoir rocks by801microfocusX-raycomputerisedtomography.SedimentaryGeology,132(1-2),25-36.802

Walderhaug,O. 1996.Kineticmodelingof quartz cementation andporosity loss indeeplyburied803sandstonereservoirs.AAPGbulletin,80(5),731-745.804

Walderhaug,O.,&Bjørkum,P.A.(2003).Theeffectofstylolitespacingonquartzcementationinthe805LowerJurassicStøFormation,southernBarentsSea.JournalofSedimentaryResearch,73(2),806146-156.807

Wanless,H.R.1979.Limestoneresponsetostress:pressuresolutionanddolomitization.Journalof808SedimentaryResearch,49(2),437-462.809

Zinszner,B.,&Pellerin, F.M.2007.AGeoscientist'sGuide toPetrophysics. EditionsTechnip. ISBN:8109782710808992.811