porosity of devonian and mississippian new albany shale across a maturation gradient: insights from...

23
GEOHORIZON Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion Maria Mastalerz, Arndt Schimmelmann, Agnieszka Drobniak, and Yanyan Chen ABSTRACT The evolution of porosity in shales with increasing maturity was examined in a suite of five New Albany Shale samples spanning a maturity range from immature (vitrinite reflectance, R o 0.35%) to postmature (R o 1.41%). Devonian to lower Mis- sissippian New Albany Shale samples from the Illinois Basin used in this study contain marine type II kerogen having total organic carbon contents from 1.2 to 13.0 wt. %. Organic petrol- ogy, CO 2 and N 2 low-pressure adsorption, and mercury intru- sion capillary pressure techniques were used to quantify pore volumes, pore sizes, and pore-size distributions. Increasing maturity of the New Albany Shale is paralleled by many changes in the characteristics of porosity. The total porosity of 9.1 vol. % in immature New Albany Shale decreases to 1.5 vol. % in the late mature sample, whereas total pore volumes decrease from 0.0365 to 0.0059 cm 3 /g in the same sequence. Reversing the trend at even higher maturity, the post- mature New Albany Shale exhibits higher porosity and larger total pore volumes compared to the late mature sample. With increasing maturity, changes in total porosity and total pore vol- umes are accompanied by changes in pore-size distributions and relative proportions of micropores, mesopores, and macropores. Porosity-related variances are directly related to differences in the amount and character of the organic matter and mineralogical AUTHORS Maria Mastalerz Indiana Geological Survey, Indiana University, 611 North Walnut Grove Ave., Bloomington, Indiana; [email protected] Maria Mastalerz received her Ph.D. in geology from the Silesian Technical University in Gliwice, Poland, in 1988 and did her postdoctoral fellowship at University of British Columbia in Vancouver, Canada. She is currently a senior scientist at the Indiana Geological Survey, Indiana University. Her research interests include coal geology, organic matter petrology and geochemistry, coalbed methane, shale gas, and CO 2 sequestration. Arndt Schimmelmann Department of Geological Sciences, Indiana University, 1001 East 10th St., Bloomington, Indiana; [email protected] Arndt Schimmelmann received his degree in chemistry from the Technical University at Braunschweig, Germany (1979), and his Ph.D. in geochemistry from the University of California at Los Angeles (1985). He is currently a senior scientist at Indiana University. His research interests include light stable isotopes in organic materials, isotopic shifts in sedimentary organic matter during maturation, and Holocene paleoceanography. Agnieszka Drobniak Indiana Geological Survey, Indiana University, 611 North Walnut Grove Ave., Bloomington, Indiana; [email protected] Agnieszka Drobniak received her Ph.D. (2002) in earth sciences from AGH University of Science and Technology, Kraków, Poland. She has been working with the Indiana Geological Survey for the past 11 years. Her research interest includes coal geology, coalbed methane, CO 2 sequestra- tion, trace elements in coal, and coal petrography. She is specializing in geographic information sys- tems and databases. Yanyan Chen Department of Geological Sciences, Indiana University, 1001 East 10th St., Bloomington, Indiana; [email protected] Yanyan Chen received her masters degree in geology from Indiana University at Bloomington, Indiana (2012). She is currently a Ph.D. candidate in the Department of Geological Sciences at Indiana University. Her research interests include coal geology, organic matter petrology and geochemistry, and shale gas. Copyright ©2013. The American Association of Petroleum Geologists. All rights reserved. Manuscript received November 13, 2012; provisional acceptance February 21, 2013; revised manuscript received March 5, 2013; final acceptance April 1, 2013. DOI:10.1306/04011312194 AAPG Bulletin, v. 97, no. 10 (October 2013), pp. 1621 1643 1621

Upload: yanyan

Post on 18-Dec-2016

214 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

AUTHORS

Maria Mastalerz � IndianaGeological Survey,Indiana University, 611 North Walnut Grove Ave.,Bloomington, Indiana; [email protected]

Maria Mastalerz received her Ph.D. in geologyfrom the Silesian Technical University in Gliwice,Poland, in 1988 and did her postdoctoralfellowship at University of British Columbia inVancouver, Canada. She is currently a seniorscientist at the Indiana Geological Survey, IndianaUniversity. Her research interests include coalgeology, organicmatter petrology and geochemistry,coalbed methane, shale gas, and CO sequestration.

GEOHORIZON

Porosity of Devonian andMississippian New Albany Shaleacross a maturation gradient:Insights from organic petrology, gasadsorption, and mercury intrusion

2

Arndt Schimmelmann � Department of

Maria Mastalerz, Arndt Schimmelmann, Geological Sciences, Indiana University, 1001 East Agnieszka Drobniak, and Yanyan Chen 10th St., Bloomington, Indiana;[email protected]

Arndt Schimmelmann received his degree inchemistry from the Technical University atBraunschweig, Germany (1979), and his Ph.D. ingeochemistry from the University of California atLos Angeles (1985). He is currently a seniorscientist at Indiana University. His research interestsinclude light stable isotopes in organic materials,isotopic shifts in sedimentary organicmatter duringmaturation, and Holocene paleoceanography.

Agnieszka Drobniak � Indiana GeologicalSurvey, Indiana University, 611 NorthWalnut GroveAve., Bloomington, Indiana; [email protected]

Agnieszka Drobniak received her Ph.D. (2002) inearth sciences from AGH University of Scienceand Technology, Kraków, Poland. She has beenworking with the Indiana Geological Survey forthe past 11 years. Her research interest includescoal geology, coalbed methane, CO2 sequestra-tion, trace elements in coal, and coal petrography.She is specializing in geographic information sys-tems and databases.

Yanyan Chen � Department of GeologicalSciences, Indiana University, 1001 East 10th St.,Bloomington, Indiana; [email protected]

Yanyan Chen received her master’s degree ingeology from Indiana University at Bloomington,Indiana (2012). She is currently a Ph.D. candidate

ABSTRACT

The evolution of porosity in shales with increasing maturitywas examined in a suite of five New Albany Shale samplesspanning amaturity range from immature (vitrinite reflectance,Ro 0.35%) to postmature (Ro 1.41%). Devonian to lower Mis-sissippian New Albany Shale samples from the Illinois Basinused in this study contain marine type II kerogen having totalorganic carbon contents from 1.2 to 13.0 wt. %. Organic petrol-ogy, CO2 and N2 low-pressure adsorption, and mercury intru-sion capillary pressure techniques were used to quantify porevolumes, pore sizes, and pore-size distributions.

Increasing maturity of the New Albany Shale is paralleledby many changes in the characteristics of porosity. The totalporosity of 9.1 vol. % in immature NewAlbany Shale decreasesto 1.5 vol. % in the late mature sample, whereas total porevolumes decrease from 0.0365 to 0.0059 cm3/g in the samesequence. Reversing the trend at even higher maturity, the post-mature New Albany Shale exhibits higher porosity and largertotal pore volumes compared to the late mature sample. Withincreasing maturity, changes in total porosity and total pore vol-umes are accompanied by changes in pore-size distributions andrelative proportions of micropores, mesopores, and macropores.Porosity-related variances are directly related to differences in theamount and character of the organic matter and mineralogical

in the Department of Geological Sciences atIndiana University. Her research interests includecoal geology, organic matter petrology andgeochemistry, and shale gas.Copyright ©2013. The American Association of Petroleum Geologists. All rights reserved.

Manuscript received November 13, 2012; provisional acceptance February 21, 2013; revised manuscriptreceived March 5, 2013; final acceptance April 1, 2013.DOI:10.1306/04011312194

AAPG Bulletin, v. 97, no. 10 (October 2013), pp. 1621– 1643 1621

Page 2: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

ACKNOWLEDGEMENTS

This study received support from U.S. Departmentof Energy, Basic Energy Sciences, Grant No. DE-FG02-11ER16246. We thank three anonymousreviewers for their comments and suggestions.The AAPG Editor thanks the anonymous reviewersfor their work on this paper.

1622 Geohorizon

composition, but maturity exerts the dominant control uponthese characteristics. We conclude that organic matter transfor-mation due to hydrocarbon generation and migration is a pivotalcause of the observed porosity differences.

INTRODUCTION

Producibility of shale gas from unconventional reservoirs in low-permeability shales critically depends on the pore systems forstoring and releasing hydrocarbon gas. The behavior of shales as areservoir rock for gas is influenced not only by storage mecha-nisms operating via mineral components (e.g., displacementphenomena and hydrodynamic solubility, Bruant et al., 2002),but even more importantly by parallel characteristics of organicmatter offering sorption sites on the organic surface area ofmesopores or “volume filling” inmicropores (Bustin et al., 2008).In contrast to organic macerals that commonly preserve cav-ernous microstructures from their biological precursor tissues,little original porosity exists within most mineral grains. Be-cause of the coexistence of minerals and organic matter, gas inshales is generally stored both in (1) an adsorbed state in con-tact with organic matter, and (2) in a free gas state in inter-granular porosity and fractures (e.g., Jarvie et al., 2007). Thisstands in contrast to conventional sandstone reservoirs that con-tain gas mostly in a free state, and it also differs from coal wherecoalbed gas is predominantly sorbed on abundant organic mat-ter (Clarkson and Bustin, 1996). Moreover, because of differ-ent pore sizes, gas flow in shales is expected to occur via bothKnudsen diffusion and slip flow in nanometer-size pores andDarcy-like flow in larger pores (e.g., Kuila and Prasad, 2011) asopposed to Darcy-like flow in micrometer or even larger poresizes in sandstone.

The determination of porosity in shales is difficult, in partbecause no single fluid-probe-based porosimetric technique cansatisfactorily measure the wide range of pore volumes and pore-size distributions. Instead, a combination of various porosimetrictechniques must be applied synoptically. Recent comparativestudies on porosity-related characteristics used various tech-niques todocumentpore-sizedistribution in shales from less-than-2-nm micropores to macropores measuring more than 50 nmto micrometer size (Kuila and Prasad, 2011; Chalmers andBustin, 2012;Chalmers et al., 2012;Clarkson et al., 2012, 2013).Nelson’s (2009) compilation of pore-throat diameters in shalesbased on mercury intrusion capillary pressure (MICP) indicatestypical pore-size values from 5 to 50 nmwith rare larger outliers.MICP’s lower detection limit of approximately 3.6 nm renders

Page 3: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

this technique unsuitable for detecting less-than-2-nm pores (i.e., micropores), although the pres-ence of small micropores was documented in shalesvia alternate techniques, such as gas adsorption(Kuila and Prasad, 2011; Chalmers et al., 2012;Mastalerz et al., 2012). In comparison, pore-throatdiameters in coarser siliciclastic rocks are muchlarger and span a continuum from 20 to 0.005 mm(5 nm). Reservoir sandstones and tight-gas sand-stones have typical pore-throat-size values of morethan 2 mm (Nelson, 2009) and from 2 to 0.03 mm(30 nm), respectively.

Focused ion beam (FIB) milling coupled withthe high magnification of scanning electron micro-scopy (SEM) enabled the imaging of nanometer-scale pores in shales (Schieber, 2010; Curtis et al.,2011a, b; 2012; Chalmers et al., 2012; Fishmanet al., 2012; Loucks et al., 2012). Although thishighly specialized technique is hampered by thespatial limitation of the observed area and thusmay never be able to achieve representative im-aging for reservoir characterization, importantly,it proves the occurrence of nanopores in shales.These FIB-SEM findings confirm the utility of othertechniques such as gas adsorption and neutronscattering, which have provided earlier indirect evi-dence for nanometer-size pores in shales but wereunable to offer visualization. These techniques havealso allowed classification of shale pores into dis-tinct domains that were either associated with min-erals (interparticle pores and intraparticle pores) orwith organic matter (Loucks et al., 2012).

Pores in organic matter attract interest becauseof the growing awareness that the dominant gasstorage capacity of shale resides in the kerogen net-work (Ross and Bustin, 2009; Strąpoć et al., 2010).Modica and Lapierre (2012) suggested that the“water-wet” (i.e., water-logged) matrix and poreswithin minerals (Passey et al., 2010) may not con-tribute at all to the available hydrocarbon storagevolume or transmissibility, because relatively largecapillary adhesion forces (relative to buoyancy for-ces) in water-logged systems make the replacementof water by gas rather unlikely (McAuliffe, 1979;Vandenbroucke, 1993). The modeling study ofModica and Lapierre (2012) further suggests thatrelevant storage capacity evolves as a result of

thermal decomposition of organic matter duringcatagenesis.

Evolution of organic matter porosity with ma-turation is not well understood, although it is men-tioned with increasing frequency. The recent em-phasis on documenting pores of various sizes inshales has not been matched with adequate effortstoward understanding porosity evolution with matu-ration. Studies of coalbedmethane (CBM) provideda wealth of literature on the evolution of porosity inmacerals with increasing coal rank (e.g., Gan et al.,1972; Clarkson and Bustin, 1996; Levy et al., 1997).Those and other studies have documented that in-creasing coal rank is accompanied by decreasingvolumes of macropores and increasing volumes ofmicropores. The total porosity of vitrinite-rich coalshaving carbon contents more than 80 wt. % (i.e.,medium volatile bituminous rank and higher) ismainly caused by micropores, whereas microporesand mesopores jointly account for the bulk of po-rosity in coals having carbon contents between 76and 84 wt. % (i.e., high volatile bituminous rank),and macroporosity is dominant in coals with carboncontents less than 75 wt. % (i.e., subbituminousrank and below) (Gan et al., 1972). Although someof this coal-specific knowledge can be applied toshales, some processes likely exerting unique in-fluences on the porosity of shales exist, e.g., hydro-carbon generation, cracking, and formation of solidbitumen (Jarvie et al., 2007; Loucks et al., 2009).The difficulty of quantifying or even documentingthese influences was exemplified by Curtis et al.(2011a), who used FIB and SEM imaging in a com-parison of two Devonian Marcellus Shale samples(Appalachian Basin) having Ro approximately 1.1%and more than 3.0% and observed no conclusivedifference in porosity. Pores with diameters from10 to 140 nm were found in organic matter in thesample with Ro approximately 1.1%. The organicmatter had estimated porosities of 12.1% and about18.5% in two different parts of the sample. TheMarcellus Shale sample with Ro more than 3.0%featured 5- to 20-nm-size pores in organic matterwith estimated porosities of 6.1 and 15.4% in twodifferent areas. The authors acknowledged that thesedifferences could be rooted in contrasting organicmatter types as well as in differences in maturity,

Mastalerz et al. 1623

Page 4: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

but their techniques could not distinguish betweentypes of organic matter. Using the same FIB-SEMtechniques on a sequence of Devonian WoodfordShale samples from Oklahoma having maturitiesfrom Ro 0.51 to 6.36%, Curtis et al. (2012) ob-served no secondary porosity (unrelated to primaryporosity in the original immature organic matter) inorganic matter below Ro 0.90%. In contrast, sec-ondary porosity was observed in samples havingRo 1.23%, 1.67%, 3.60%, and 6.36%, whereas noporosity was seen at Ro 2.00%. Based on the latter,it was concluded that (1) maturity alone is not areliable predictor of porosity in organic matter, and(2) other factors such as organic matter compositioncomplicate a prediction. Bernard et al. (2012) usedorganic geochemistry and spectromicroscopic tech-niques on a suite of Mississippian Barnett Shalesamples from Texas ranging from immature (Tmax =420°C, Ro ~ 0.40%) to postmature (Tmax > 500°C,Ro > 1.85%) to document the chemical evolution ofshale, suggesting formation of nanoporous pyro-bitumen as a result of secondary cracking at highmaturation. Another example is based on SEMobservations of ion-milled surfaces of JurassicKimmeridge Clay Formation shales from the NorthSea varying in maturity from immature (Tmax =407°C, Ro ~ 0.25%) to late mature (Tmax = 460°C,Ro ~ 1.10%) (Fishman et al., 2012). Based onsimilar size and shape of pores in immature andmature samples, the authors suggest that porespreserved in these shales are not related to ther-mogenic hydrocarbon generation, and that most ofthe observable porosity is interparticle porosity.They further suggest that any pores that devel-oped via hydrocarbon generation during matura-tion failed to be preserved. Although SEM-basedand other high-resolution techniques aid in thevisualization of pores, their main drawback is thelack of representative quantification of pores interms of pore volumes and wider scale pore-sizedistributions.

The obvious need for additional systematicstudies on shales spanning a wide range of thermalmaturities prompted our effort toward a betterunderstanding of the evolution of porosity in re-sponse to hydrocarbon generation, migration, andcracking. For this study, we selected a suite of

1624 Geohorizon

New Albany Shale samples spanning a maturityrange from immature (Ro 0.35%) to postmature(Ro 1.41%).

METHODOLOGY

Sample Material

The New Albany Shale samples were collectedfrom core material archived at the Indiana Geo-logical Survey and the Illinois State GeologicalSurvey. The sampling sites were carefully selected tocover a range of maturities from immature topostmature (Figure 1). Some samples come fromthe same locations and depths as the samplesthat were characterized extensively in our pre-vious work (Lis et al., 2005, 2006, 2008; Strąpoćet al., 2010), and their characteristics helped sam-ple selection in such a way that they represent notonly a progressive maturity suite but also a rangeof mineralogical composition and organic mattercontents. This way, although it is a limited set ofsamples, it represents the New Albany shales rel-atively well. The samples contain marine type IIkerogen having total organic carbon (TOC) con-tents ranging from 1.2 to 13.0 wt. % (Table 1).Each sample was split into several aliquots to beused for analyses of the mineralogical composi-tion, organic petrography, and various porosimetrytechniques. Rock material used for analysis wasmassive, well preserved, and with no sign of weath-ering or oxidation. Weathered parts, if present, wereremoved during sample collection.

Mineralogical Composition

Shale samples were first ground in a micronizerwith water to less than 5 mm in size, and then ovendried at 110°C overnight. A Bruker D8 Advancex-ray diffractometer was used for x-ray diffrac-tion analyses of the pulverized samples. Thediffractometer was equipped with a Sol-X solid-state detector and a Cu x-ray tube operated at40 kV and 30 mA. The powdered samples werescanned from 2° to 70° with a count time of 2 s per0.02° step. Multicomponent quantifications were

Page 5: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

achieved by Rietveld refinements with TOPASsoftware.

Gas Adsorption

Low-pressure gas adsorption measurements withnitrogen (N2) and carbon dioxide (CO2) wereconducted on a Micromeritics ASAP-2020 appa-ratus. Shale sample aliquots weighing 1 to 2 g wereanalyzed with N2 to obtain information aboutmesopores (2–50 nm in diameter), whereas CO2

was used to characterize micropores (<2 nm indiameter) that are more accessible for CO2 thanfor N2. The classification of pore sizes used in thisarticle follows the classification system of the In-ternational Union of Pure and Applied Chem-istry (Orr, 1977). This classification of pore sizeshas proven to be very convenient in coal studies

(Clarkson and Bustin, 1996; Mastalerz et al., 2008),and because the shales studied are organic matterrich and have similar pore sizes to coals, it has beencommonly used for shales as well (e.g., Clarksonet al., 2012).

Samples were automatically degassed at about110°C under vacuum for about 14 hr to removeadsorbed moisture and volatile matter prior to

Table 1. Vitrinite Reflectance (Ro), Total Organic Carbon (TOC)

Content, Depth, and Maturity of Shale Samples

Sample

Ro (%) TOC (wt. %)

Ma

Depth (m)

stalerz et al.

Maturity Stage

472–1

0.35 1.2 61 immature MM4 0.55 13.0 764 early mature NA2 0.65 5.3 853 early mature IL–5 1.15 4.3 1607 late mature IL–1 1.41 6.3 78 postmature

Figure 1. Map showing a location of the sampling sites in the framework of the extent of New Albany Shale. Isolines represent vitrinitereflectance (Ro) values in %.

1625

Page 6: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

analyses with either N2 or CO2. Nitrogen gas ad-sorption was quantified after exposure to N2 at apressure up to 97.3 kPa (730 mm Hg) at the tem-perature of liquid nitrogen (77.35 K). CO2 ad-sorption occurred at a pressure of up to 104.5 kPa(784 mm Hg) at 0°C (273.1 K). The equilibriuminterval (i.e., the time during which the pressuremust remain stable within a small range) was 30 s,and the pressure tolerance was set at 0.6666 kPa(5 mm Hg). The instrument’s computer softwareautomatically generated adsorption isotherms andcalculated surface areas, pore volumes, and pore dis-tributions based on multiple adsorption theories, i.e.,Langmuir, Brunauer-Emmett-Teller (BET), Barrett-Joyner-Halenda (BJH), Dubinin-Radushkevich(D-R), and Dubinin-Astakhov (D-A), among oth-ers. Mesopore-size distribution was determined bythe BJH method, whereas micropore-size distribu-tion was determined by the density functional the-ory (DFT)method (Webb andOrr, 1997). TheDFTfit to isotherms of all samples was very good, withthe standard deviation of fit ranging from 0.00063to 0.00101 cm3/g. A detailed description of thesetheories and techniques can be found in Gregg andSing (1982).

Total Porosity

Porosity of the samples was determined at CoreLaboratories in Houston using a technique based onBoyle’s Law, by measuring grain volume at ambientconditions and bulk volume by Archimedes’ princi-ple and coupling ofmercury immersionwith heliumpycnometry. Samples were oven dried at 115°C toconstant weight (±1 mg). Porosity was calculatedfrom the difference between bulk volume and grainvolume.

Mercury Intrusion Capillary Pressure

Mercury intrusion data were collected on a Micro-meritics Autopore 9320 instrument along an incre-mentalmercury pressure increase from0 to379MPa.Pore-size distributions (PSD) were determined usingthe Washburn equation (Washburn, 1921). Perme-ability was determined by the Swanson technique(Swanson, 1981). This technique was also used to

1626 Geohorizon

determine pore-throat distribution. A pore throat isthe largest entrance toward a pore, but not the actualinner size of a pore (Giesche, 2006), and conse-quently, it is commonly smaller than a pore diameteror pore width.

Other Analyses

TOC was quantified using a Leco analyzer. Petro-graphic analyses including random reflectance mea-surements of the shales were conducted in reflectedlight under a Photoscope II microscope, followingstandard coal petrography procedures (Taylor et al.,1998).

RESULTS

Mineralogy and Organic Petrography

The mineralogical compositions of shale sam-ples were determined via x-ray diffraction (XRD)(Table 2; Figure 2). Carbonates (calcite, dolomite,and ankerite) are dominant in the immature sam-ple 472-1. The early mature samples MM4 andNA2 are relatively similar in composition and con-tain small amounts of carbonates. Clays (illite inparticular) are dominant, whereas quartz accountsfor 35 and 36 wt. % inMM4 andNA2, respectively.The late mature sample IL-5 contains 26 wt. %

Table 2.Mineralogical Composition (inWeight%) of Shale Samples

Constituent

472–1(wt. %)

MM4(wt. %)

NA2(wt. %)

IL–5(wt. %)

IL–1(wt. %)

Quartz

19 35 36 31 25 Calcite 27 0 0 11 0 Dolomite 13 0 1 15 10 Ankerite 10 0 0 0 0 Albite 8 9 13 6 27 Orthoclase 2 0 1 4 17 Pyrite 2 1 3 2 9 Illite/Muscovite 12 34 36 26 3 Chlorite 1 3 3 0 1 Kaolinite 3 4 3 0 1 TOC 1 13 5 4 6
Page 7: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

carbonates, 31 wt. % quartz, and 26 wt. % clayminerals. The most mature sample IL-1 has abun-dant feldspar (44 wt. %), 25 wt. % of quartz, andonly 5 wt. % of clays (Table 2; Figure 2).

The organic matter content (expressed asTOC, Table 1) ranges from 1.2 wt. % in the leastmature sample 472-1 to 13.0 wt. % in the earlymature sample MM4. Amorphous organic matter(amorphinite) and alginite are the dominant com-ponents of immature and early mature samples472-1, MM4, and NA2 (Table 3; Figure 3A–D),which renders these samples fluorescent. Solidbitumen is present but rare in MM4 and NA2 andabsent in the least mature sample 472-1 (Table 3).The late mature and postmature samples IL-5 andIL-1 dominantly feature amorphinite and solid bi-tumen, with solid bitumen being most abundantin the postmature sample IL-1 (Table 3; Figure 3F–H). Fluorescence has been extinguished in thetwo most mature samples. Vitrinite and inertinitefrom terrestrial input are rare in all studied shalesamples.

In addition to the identification of organic mat-ter, microscopic observations in reflected light pro-vide information about the spatial distribution oforganic matter and also offer insights into size anddistribution of pores more than 1 mm,whereas poresless than 1 mm cannot be reliably detected with alight microscope. Observations about the distribu-tion of organic matter and porosity in samples withincreasing maturity are summarized below.

1. The organic matter in the least mature sample472-1 occurs as elongated algal bodies (alginite)and small wisps of liptodetrinite along beddingplanes (Figure 3A). Thesemacerals are distributedirregularly throughout the sample with little di-rect contact between them. Amorphous organicmatter (amorphinite) is admixed in the matrixand may provide limited connectivity betweenorganic matter. No distinct pores more than 1 mmare detected.

2. The early mature sample MM4 (Figure 3B, C) isrich in organic matter that is commonly inter-connected. Alginite bodies are frequently in con-tact with abundant amorphinite and form an or-ganic interconnected system. Some alginites havea porous spongelike structure with pore sizes al-most reaching 1 mm (Figure 3C). We interpretthese pores to be a result of early oil generationand expulsion from this hydrogen-rich algal ma-terial. Larger sporadic poresmay occur in inertinitemacerals or are associated with solid bitumen(Figure 3B).

3. Organic matter in the early mature sample NA2is distributed similarly as in the previous sample(Figure 3D). Micrometer-size pores are presentbut sporadic and can occur in inertinite or pyrite

0

10

20

30

40

50

60

70

80

90

100

472-1 MM4 NA2 IL-5 IL-1

ClaysPyriteFeldsparCarbonatesQuartzW

eigh

t %

Ro 0.35% Ro 0.55% Ro 0.65% Ro 1.15% Ro 1.41%

Figure 2. Mineralogical composition of shale samples (seedetails for specific minerals in Table 2).

Table 3. Organic Petrographic Composition (Volume %, on Mineral Matter Free Basis) of Shale Samples*

Sample

Ro (%) AOM (wt. %) ALG (wt. %) LPD (wt. %) BIT (wt. %) V+I (wt. %)

Mastalerz et al.

Fluorescence

472–1

0.35 85 10 5 0 traces very strong MM4 0.55 52 35 10 2 1 strong NA2 0.65 52 27 10 10 1 strong IL–5 1.15 54 0 0 45 1 nonfluorescent IL–1 1.41 24 0 0 75 1 nonfluorescent

*Ro = vitrinite reflectance; AOM = amorphous organic matter; ALG = alginite; LPD = liptodetrinite; BIT = solid bitumen; V+I = vitrinite plus inertinite.

1627

Page 8: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

(Figure 3E). However, the insular occurrence ofthese pores in inertinite and pyrite in the shalematrix likely causes isolation of these pores andlimits access to fluids.

4. The distribution of organic matter in the latemature sample IL-5 is very different from that inless mature shales. Amorphous organic matterand solid bitumen are evenly distributed through-out the sample, fill intergranular spaces, and wrap

1628 Geohorizon

around mineral grains. At the resolution of re-flected light microscopy, solid bitumen can ap-pear nonporous (Figure 3F) or show high poros-ity with a sievelike structure (Figure 3G). Such adistribution of pores in organic matter suggestshigh conductivity and good pore accessibility.

5. The most mature sample IL-1 has a similarorganic matter distribution as sample IL-5. Itfeatures an excellently interconnected organic

Figure 3. Photomicrographs taken un-der reflected light and oil immersion ex-cept in panel C, which is in fluorescentlight. (A) Alginite (brown) in immaturesample 472–1; (B) randomly distributedorganic matter in early mature sampleMM4; (C) Leiosphaeridia alginite withpores of about 1 mm; (D) vitrinite andsolid bitumen in early mature sampleNA2; (E) pores of approximately 1 mmin pyrite in early mature sample NA2;(F) nonporous solid bitumen in latemature sample IL–5; (G) nonporousand porous solid bitumen in late maturesample IL–5; (H) solid bitumen fillingsmall and larger pores in postmaturesample IL–1.

Page 9: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

matter network and few pores more than 1 mm(Figure 3H).

Total Porosity and Total Pore Volume

Both helium porosimetry and MICP porosimetrywere employed to provide comparative values forgrain density, bulk sample density, total pore volume,and total porosity (Table 4). Total pore volumes asdetermined by helium porosimetry are larger thanthose obtained by MICP, whereas grain density andbulk density data are generally comparable betweenthe two methods. The smallest difference betweenthe two techniques is noted for sampleNA2, and thelargest difference is observed for the most maturesample IL-1, where the He-determined total porevolume is more than two times larger than the valuedetermined byMICP. Similarly, helium porosimetryalso measures larger values for total porosity thatrange from 9.1 vol. % for the least mature sample472-1 to 1.5 vol. % for the late mature sample IL-5.In comparison, the same samples yield MICP-derived values from 5.6 vol. % for 472-1 to 0.8 vol. %for IL-5. Smaller pore volumes measured via MICPcan be explained by the inability of mercury to effi-ciently enter pores less than 3 nm in diameter,makingit necessary to rely on helium porosimetry to ac-count for small pores.

Despite some differences between the twoporosimetric techniques, both show the sameoveralltrends for total porosity and the total pore volumewith increasing maturation (Table 4; Figure 4A).Total porosity and total pore volume are largest inthe least mature sample 472-1, decline with in-

creasing maturity to intermittent minima in the latemature sample IL-5, and later increase toward themost mature sample IL-1. No significant correla-tions exist between TOC and either total porosity ortotal pore volume for this suite of shale samples(Figure 4B).

Table 4. Porosity Based on Helium and MICP Porosimetry of Shale Samples*

He Porosimetry

Hg Intrusion Porosimetry

Sample

GrainDensity(g/cm3)

BulkDensity(g/cm3)

Total PoreVolume(cm3/g)

TotalPorosity(%)

GrainDensity(g/cm3)

BulkDensity(g/cm3)

Mast

Total PoreVolume(cm3/g)

alerz et al.

TotalPorosity(%)

472–1

2.74 2.49 0.0365 9.1 2.79 2.63 0.0213 5.6 MM4 2.30 2.21 0.0186 4.1 2.26 2.2 0.0122 2.7 NA2 2.56 2.43 0.0210 5.1 2.53 2.43 0.0168 4.1 IL–5 2.58 2.55 0.0059 1.5 2.58 2.56 0.0032 0.8 IL–1 2.58 2.49 0.0141 3.5 2.53 2.5 0.0055 1.4

*MICP = mercury intrusion capillary pressure.

R=0

.35%

o

R=1

.15%

o

R=0

.65%

o

R=1

.41%

o

R=0

.55%

o

A

B

3To

tal p

ore

volu

me

(cm

/g)

3To

tal p

ore

volu

me

(cm

/g)

MICP

MICP

TOC (weight %)

1.0 1.50.50

00

0.010

0.020

0.030

0.040

0

0.010

0.020

0.030

0.040

5 1510

Figure 4. Relationships between total pore volume (as de-termined by mercury intrusion capillary pressure [MICP] andhelium porosimetry) and (A) vitrinite reflectance, and (B) totalorganic carbon (TOC).

1629

Page 10: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

1630 Geohorizon

Pore-Size Distribution

Micropores (Ø < 2 nm)Micropore characteristics were quantified by CO2

adsorption and surprisingly document strongest ad-sorption in both the immature shale sample 472-1and in the postmature sample IL-1 (Figure 5A).The late mature IL-5 sample adsorbs the leastamount of CO2, whereas mature samples MM4and NA2 adsorb intermediate volumes. In otherwords, volumes of adsorbed CO2 in this samplesuite intermittently decline with increasing maturityto reach a minimum in the late mature sample IL-5,and then increase toward the most mature sampleIL-1. This pattern is corroborated by DFT cumula-tive micropore-size distributions (Figure 5B) wheresamples 472-1 and IL-1 have the largest volumes ofmicropores less than 1 nm in width, and the latemature sample IL-5 has the smallest microporevolume. The latter sample IL-5 also has the smallestD-A micropore volume (Table 5). This bimodaldistribution of micropore characteristics over shalematurity indicates that shales initially lose micro-pores until the late mature stage, but greatly re-generate micropores at higher maturity. The DFTincremental micropore-size distribution demon-strates significant changes throughout maturation(Figure 6). The overall changes can be summarizedas follows. (1) Several volumetric maxima are pre-sent in all analyzed shale samples within the poresize range from 0.4 to 1.1 nm; (2) the immaturesample 472-1 has four volumetric maxima at porewidths of 0.52, 0.58, 0.78, and 0.83 nm (Figure 6A);(3) the two early mature samples MM4 and NA2

Figure 5. (A) Low–pressure CO2 adsorption isotherms of shalesamples. Note that the largest volumes of adsorbed CO2 occur inthe least mature sample 472–1 and in the most mature sampleIL–1. P = actual gas pressure; Po = vapor pressure of the ad-sorbing gas. (B) Cumulative micropore volumes of shale samplesobtained by low–pressure CO2 adsorption technique. Note thatthe largest cumulative volumes occur in the least mature sample472–1 and in the most mature sample IL–1.

Table 5. Micropore, Mesopore, and Macropore Volumes Determined Via Gas Adsorption and MICP Techniques*

N2 and CO2 Adsorption**

Hg Intrusion

Sample

D–A MicroporeVolume (cm3/g)

BJH MesoporeVolume (cm3/g)

Macropore VolumeCalc. (cm3/g)

Micropore VolumeCalc. (cm3/g)

MesoporeVolume (cm3/g)

MacroporeVolume (cm3/g)

472–1

0.0098 0.0243 0.0025 0.0151 0.0199 0.0014 MM4 0.0167 0.0008 0.0011 0.0064 0.0117 0.0005 NA2 0.0138 0.0052 0.0199 0.0042 0.0169 0 IL–5 0.0055 (0.0177) 0.0004 (0.0004) 0.0000 0.0026 0.0033 0 IL–1 0.0062 (0.0074) 0.0036 (0.0062) 0.0042 0.0085 0.0056 0

*For IL–5 and IL–1 samples, values in brackets indicate analyses on the same sample but after extraction with dichloromethane.**D–A = Dubin in Astakhov; BSH = Barrett-Joyner-Halenda.

Page 11: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

express similar pore-width maxima at 0.47, 0.52,0.60, 0.68, 0.74, and 0.81 nm (Figure 6B, C); (4) thelate mature sample IL-5 features only three poreranges with pore-width maxima at 0.54, 0.68, and0.77 nm (Figure 6D); (5) the most mature sam-ple IL-1 has several volumetric pore-size maxima(Figure 6E) similar to those of immature sample472-1 (Figure 6A); (6) very small volumes are re-corded for all pores in late mature sample IL-5,for example, the most prominent micropore rangehaving a maximum size at 0.54 nm measures only0.000006 cm3/g (i.e., 0.006 mm3/g; Figure 6D).This stands in stark contrast to postmature sampleIL-1 where the volume of the most prominent mi-cropore range having a maximum size of 0.5 nmexceeds the corresponding pore volume in late ma-ture sample IL-5 by a factor of 20 (Figure 6D, E).

TheMICP technique’s inability to forcemercuryinto pores with pore-throat values less than 3 nmmakes it necessary to calculate micropore volumesvia subtraction of the MICP-determined sum ofmesopore and macropore volumes from the totalpore volume as determined by helium porosimetry.This approach assumes that measurements of thetotal pore volume by helium porosimetry aremore reliable than MICP data because the formercovers all pore sizes. Micropore volumes calcu-lated in this way range from 0.0026 cm3/g in sampleIL-5 to 0.0151 cm3/g in sample 472-1 (Table 5).Comparing these values to the micropore vol-umes determined by CO2 adsorption (0.0055–0.0167 cm3/g; Table 5) and considering analyticaldifferences between these two techniques, weconclude that micropore volumes measured byHe adsorption and MICP and CO2 adsorptionare qualitatively comparable.

Mesopores (Ø 20–50 nm)Nitrogen low-pressure adsorption isotherms wereused to quantifymesopore characteristics (Figure 7).The pattern of nitrogen adsorption volumes withincreasing maturity of shale samples follows thatof CO2 adsorption (Figure 5A). Strongest N2

adsorption is observed in the immature sample472-1, followed by declining values with increas-ing maturity to a minimum value in the late ma-ture sample IL-5, and finally increasing adsorption

A

B

C

D

E

472-1

MM4

NA2

IL-5

IL-1

R =0.35%o

R =0.55%o

R =0.65%o

R =1.15%o

R =1.41%o

3M

icro

pore

vol

ume

(cm

/g)

Mic

ropo

re v

olum

e (c

m/g

)3

3M

icro

pore

vol

ume

(cm

/g)

3M

icro

pore

vol

ume

(cm

/g)

3M

icr o

pore

vol

ume

(cm

/g)

Pore width (nm)

Pore width (nm)

Pore width (nm)

Pore width (nm)

Pore width (nm)

Figure 6. Incremental micropore volumes of shale samplesobtained by the low–pressure CO2 adsorption technique.

Mastalerz et al. 1631

Page 12: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

toward the most mature sample IL-1. The N2

adsorption isotherms are of type IV with hyster-esis being noticeable in all samples (Figure 7).The hysteresis loop is closed in the case of theimmature sample 472-1 (Figure 7A) but open forthe postmature sample IL-1 (Figure 7B). Thepresence of a pronounced hysteresis indicates thatevaporation from pores is a distinctly different pro-cess than condensation within the pores and sug-gests that capillary condensation occurred withinthe mesopores (Gregg and Sing, 1982; Bustin andClarkson, 1999). The absence of total closure ofthe low-pressure hysteresis loop was interpreted asbeing due to swelling or adsorption of nitrogen inmicropores (Gregg and Sing, 1982).

1632 Geohorizon

Cumulative nitrogen adsorption data indicatemesopore-size distributions over increasing matu-rity. The largest mesopore volume occurs in theimmature shale sample 472-1 (Figure 8A; Table 5),followed by the second-largest mesopore volumein the postmature sample IL-1. The late maturesample IL-5 has the smallest mesopore volume(and also the smallest micropore volume; Table 5)of all shale samples (Figure 8B). Interestingly, thisis closely followed in mesopore volume by the earlymature sampleMM4,which also contains the largestconcentration of organic matter (TOC 13 wt. %,Table 1). Barrett-Joyner-Halenda (BJH) mesopore-size distributions are similar for all measured shalesamples, with the pore-size class of about 27–91 nmaccounting for most of the pore volumes (Figure 9A–E). Pores having diameters between 2 and 10 nm are

Figure 7. (A) Low–pressure N2 adsorption isotherms of shalesamples. Note that the largest volumes of adsorbed N2 occur inthe least mature sample 472–1, followed by the most maturesample IL–1. (B) The same isotherms as in (A), but omitting thestrongest adsorbing sample 472–1 to better see the absorbanceof the remaining samples. P = actual gas pressure; Po = vaporpressure of the adsorbing gas.

Figure 8. (A) Cumulative mesopore volumes of shale samplesobtained by the low-pressure N2 adsorption technique. (B) Thesame as in (A), but omitting the strongest adsorbing sample 472–1to better see mesopore volumes of the remaining samples.

Page 13: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

rare, with the exception of the immature sample472-1 (Figure 9A). The latter observations con-trast sharply against coals where most of the totalmesopore volume is provided by the 2–10-nmpore-size range (Mastalerz et al., 2012).

The MICP–based trend and range of mesoporevolumes from 0.0033 cm3/g in the late maturesample IL-5 to 0.0199 cm3/g in the immaturesample 472-1 (Table 5) are similar to the trend andrange obtained fromN2 adsorption. The cumulativeand incremental distributions of pore-throat radiifor all shale samples demonstrate that essentially allpore throats fall within the mesopore range from 2to 50 nm (i.e., 1–25 nm radius; Figure 10).

Macropores (Ø > 50 nm)Using gas adsorption data, the macropore volumewas calculated in this studyby subtractingmicroporeand mesopore volumes (measured by CO2 and N2

gas adsorption, respectively) from the total porevolumes obtained by helium porosimetry, becausethe N2 adsorption technique alone cannot ade-quately quantify the distribution of pores more than100 nm. The calculated macropore volumes rangefrom 0 cm3/g in the late mature sample IL-5 to0.0042 cm3/g in the most mature sample IL-1(Table 5).

The MICP technique could not detect anymacropores in the three most mature samples(Table 5; Figure 10) and only small contributionsof macropores to the total porosity in the im-mature sample 472-1 and in the early maturesample MM4. A major discrepancy in macroporequantification between MICP and gas adsorptiontechniques is evident for the most mature sampleIL-1 where the MICP technique detects no mac-ropores, yet N2 adsorption yields a macroporevolume of 0.00412 cm3/g (Table 5).

Surface Area and Pore-Throat-Size ValuesGas adsorption-derived data on surface areas andaverage mesopore and micropore sizes, as well as

A

B

C

D

E

472-1R =0.35%o

MM4R =0.55%o

NA2R =0.65%o

IL-5R =1.15%o

IL-1R =1.41%o

3Po

re v

olum

e (c

m/g

)3

Por e

vol

ume

(cm

/g)

3Po

re v

olum

e (c

m/g

)3

Pore

vol

ume

(cm

/g)

3Po

re v

olum

e (c

m/g

)

Pore width range (nm)

Pore width range (nm)

Pore width range (nm)

Pore width range (nm)

Pore width range (nm)

Figure 9. Incremental mesopore volumes (A) from the leastmature shale sample 472–1 and proceeding sequentially to (E) themost mature sample IL–1, all obtained by low–pressure N2 ad-sorption technique.

Mastalerz et al. 1633

Page 14: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

1634 Geohorizon

MICP-derived total pore areas,median pore-throatdiameters, and largest pore throats (Table 6) pro-vide the following observations: (1) The BETsurface area as determined by N2 adsorption islargest in immature sample 472-1 (14.73 m2/g),reaches a minimum in late mature sample IL-5(0.12 m2/g), and increases in the most maturesample IL-1 (0.94 m2/g). Values for the BET equiv-alent surface area measured by CO2 adsorptionare larger relative to those measured by N2 ad-sorption, except for the immature sample 472-1.This suggests thatmicropores (not accessible toN2,but accessible to CO2) are important contributorsto total porosity in all shale samples except 472-1.(2) The average mesopore size does not seem toexpress any trend with increasing maturity, butaverage micropore sizes are largest in the least ma-ture (472-1) and the most mature (IL-1) samplesand reach a minimum in late mature sample IL-5.(3) The total pore areas and median pore-throatdiameters determined by MICP reach their mini-mum in late mature sample IL-5 and express max-ima in 472-1 and IL-1. (4) The MICP techniquesuggests pore-throat ranges for the largest pores ineach sample being 500 to 1000 nm in the immaturesample 472-1, 100 to 150 nm in the early maturesample MM4, and between 50 and 100 nm indiameter in the remaining, more mature samples.

Data from the employed MICP technique dem-onstrate that only the two least mature samples havepore-throat-size values more than 50 nm, and evenin those samples, such pore sizes are rare (Table 6;Figure 10). Other samples have maximum porethroats less than 50 nm, which is relatively small inlight of the fact that petrographic observations inreflected light document the rare presence of about1-mm pores in the same samples (Figure 3). Thisdiscrepancy suggests that either the approximately1-mm pores are closed and not accessible to mer-cury, or their pore throats are much smaller thanthe pore diameter.

Figure 10. Pore-throat-radius distributions [mm] from (A) theleast mature shale sample 472–1 and proceeding sequentially to(E) the most mature sample IL–1, all obtained by MICP technique.Swanson permeability = permeability calculated from capillarypressure data using correlations of Swanson (1981).

Page 15: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

DISCUSSION

Our results demonstrate significant changes in po-rosity along the maturation sequence from the im-mature sample 472-1 (Ro 0.35%) to the postmaturesample IL-1 (Ro 1.41%). It is, however, difficult toseparate the influences of maturity from other in-fluencing factors. Although all samples representkerogen type II, a relatively wide range of organicmatter content exist (TOC 1.2 to 13.0 wt. %, orabout 2.5% to 26% by volume) that can influenceporosity distribution. Generally, we expect to findpositive relationships between organic matter con-tent and total porosity, or between organic mattercontent and micropore or mesopore volumes(Chalmers and Bustin, 2006; Ross and Bustin, 2009;Strąpoć et al., 2010; Clarkson et al., 2013;Millikenet al., 2013). In our set of samples, however, norelationships were detected between pore volumeand TOC (Figure 4B) or between micro- andmacropore volumes and TOC (Figure 11A–C). Wesuggest that significant relationships may be appar-ent among shales of comparable maturity, whereasin the case of our samples, the relationships are ob-structed by strong differences inmaturity. In addition,this lack of correlation between TOC and pore vol-ume may also suggest that both interparticle and in-traparticle pores are significant contributors to thepore system of these shales.

The best direct evaluation of the influence of theamount of organic matter on porosity is based onsamples MM4 and NA2, which are both early ma-ture (0.55% and 0.65% Ro, respectively) and havesimilar mineralogical compositions (Figure 2) but

differ drastically in their TOC contents of 13.0versus 5.3 wt. %. The contrasting TOC contentsdo not translate into large differences in total po-rosity (4.1 vol. % inMM4 versus 5.1 vol. % inNA2;Table 4) or in total pore volume (Figure 12A).However, sample MM4 having elevated TOC fea-tures a relatively larger micropore volume as mea-sured by gas adsorption (Figure 12A, B) and similarobservations using the mercury intrusion tech-nique (Figure 12C, D).We conclude that the largermicropore volume in sample MM4 can be attri-buted to a higher organic matter content.

Mineralogical composition is another potentialfactor that may influence porosity and complicatethe interpretation of porosity evolution with in-creasing maturity. Our shale samples vary in theirmineralogical compositions (Table 2; Figure 2), butno correlations were found between total porosityand the contents of clays, carbonates, quartz, orfeldspar (Figure 13A). We suggest that any influ-ence on porosity by mineralogical composition ismasked by the stronger influence of changes inmaturity. In fact, total porosity increases with clayand quartz contents and decreases with carbonatecontent in three shale samples when we ignore theleast and most mature samples (Figure 13B). Gasadsorption-derived microporosity increases withclay content, shows a weak positive correlation withquartz content, and decreases with carbonate con-tent.Mesopore andmacropore volumes do not showstrong correlations with mineral contents, except fornegative correlations between quartz content andmesopore volume, and possibly between clay con-tent and macropore volume (Figure 14B, C). No

Table 6. Selected Porosity Characteristics Determined Via Gas Adsorption and MICP Techniques

N2 and CO2 Adsorption

Hg Intrusion

Sample

BET N2 Sa*

(m2/g)

AverageMesoporeSize (nm)

BET CO2 Sa(cm3/g)

AverageMicroporeSize (nm)

Total PoreArea (m2/g)

MedianPore-Throat

Diameter (nm)

Mastalerz et al

LargestPore-Throat

Diameters (nm)

472–1

14.7 7.4 5.9 1.35 4.36 23.40 >200, <500 MM4 0.2 16.0 16.3 0.19 4.30 7.50 >50, <100 NA2 0.6 14.9 5.9 0.53 3.85 5.80 >20, <50 IL–5 0.1 13.6 5.0 0.25 0.03 5.20 >20, <50 IL–1 0.9 16.1 7.4 1.17 2.10 11.40 >20, <50

*BET = Brunauer-Emmett-Teller; Sa = surface area.

. 1635

Page 16: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

significant correlations have been observed betweenabundances of minerals and micro-, meso-, andmacroporosities measured via MICP. Our overallobservations suggest that organic matter togetherwith clay minerals (dominantly illite in this sampleset) jointly contribute to the micropore volume inshales, which is in agreement with previous studies(Javadpour, 2009; Loucks et al., 2009; Kuila andPrasad, 2011; Chalmers and Bustin, 2012). Deci-phering the controls of clays, organic matter, and

1636 Geohorizon

maturity on pore-size distributions is critical to un-derstand the porosity network in shales.

We acknowledge that other factors are notaddressed in this article that can influence poros-ity of organic matter. For example, burial depthand depth-related pressure and temperatureconditions have been suggested to influence me-soporosity (but not microporosity) of coal in theIllinois Basin (Mastalerz et al., 2008), and the in-fluence of these factors on organic matter in shalesneeds to be evaluated. However, among the fac-tors considered in this article, maturity appears toexert dominant control upon porosity develop-ment in shales and typically trumps the effects ofcompositional differences. The following maturity-related observations result from our data:

1. Porosity in shale plummets during the transitionfrommature to latemature (Figures 15, 16). Oursample set witnesses the transition betweenRo 0.65% and 1.15% (Figures 15A; 16A), butadditional samples having intermediate matu-rities need to be examined to constrain the ma-turity level and the rapid or gradual dynamicsacross the change. The reduction in porosityis associated with a rearrangement in pore-size proportions where micropores become rel-atively more abundant at the late mature stage(Figures 15B; 16B) at the apparent preferentialexpense of mesopores.

2. Significant formation of new pores in organicmatter between Ro 1.15% and 1.41% results in alarge increase in total pore volumes (Figure 15A).Gas adsorption techniques indicate the appear-ance of pores greater than 50 nm within thismaturity range (Figure 15B), although the MICPtechnique fails to detect pores having such largepore throats (Figure 16B).

3. In response to increasing maturity, the averagemicropore width increases greatly from a mini-mum value of 0.25 nm in the late mature sampleIL-5 to 1.17 nm in the most mature sample IL-1(Table 6). Surface areas measured by both N2

and CO2 gas adsorption are smallest in sampleIL-5 as well (Table 6).

4. Increasing maturity causes a decrease in medianpore-throat diameter until the latemature stage.

Figure 11. Relationships between total organic carbon (TOC)content and (A) micropore, (B) mesopore, and (C) macropore vol-umes of shale samples. TOC = total organic carbon; MICP = mercuryintrusion capillary pressure.

Page 17: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

The median pore-throat diameter reaches itsminimum of 5.2 nm in the late mature sampleIL-5 and increases greatly to 11.4 nm in themostmature sample IL-1 (Table 6), although no

pores having pore-throat-size values more than50 nm are present.

These observations demonstrate a complex evo-lution of the pore system in shale with maturation,especially in shales that are rich in organic matter,such as the New Albany Shale studied here. We ex-pect that changes of the pore system will be lesspronounced, and perhaps not as complex, in shaleshaving low organic matter contents, where the pro-cesses of hydrocarbon generation, cracking, and mi-gration have a lesser impact on bulk sample proper-ties. Therefore, we agree with some recent studies(Milliken et al., 2013; Valenza et al., 2013) in sug-gesting that the changes in porosity of shale withmaturation, to a large extent, reflect changes in theporosity of organicmatter.We do acknowledge that adecrease in matrix pore volume at low maturity maybe related to burial and compaction (Figure 17).Previous studies suggested that organic porosity in-creased with maturity as a result of organic mattertransformation (Chalmers and Bustin, 2008; Louckset al., 2009; Modica and Lapierre, 2012). Our studydemonstrates that the evolution of porosity does notfollow a monotonous trend, but instead traverses afew minima in response to hydrocarbon generation

Figure 12. Comparison betweenorganic–rich shale sample MM4(13.0 wt. % TOC) and leaner sampleNA2 (5.3 wt. % TOC) with regard tototal pore volumes (A and C) andpercentages of micro-, meso-, andmacroporosities (B and D) using gasadsorption and mercury intrusioncapillary pressure (MICP) techniques.

Figure 13. Relationships between total porosity and mineralcontents. (A) All shale samples; (B) same as (A), but omitting thetwo samples with the lowest and highest maturities.

Mastalerz et al. 1637

Page 18: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

(Figure 17). This observation is in conflict with arecent study that uses SEM visualization of thesimilarity of sizes and shapes of organic pores inshales of different maturity from the Kimmeridge

1638 Geohorizon

Clay Formation to suggest that preserved organicpores are not related to hydrocarbon generationprocesses (Fishman et al., 2012). The authors claimthat “if any organic pores developed during the gen-eration of oil or thermogenic gas, they were appar-ently not preserved” (Fishman et al., 2012, page 49)and further conclude that inorganic porosity (i.e.,interparticle and intraparticle pores) is significant inthe storage of hydrocarbons. We maintain that theirconclusion about the lack of a relationship betweenporosity and hydrocarbon generation is prematurebecause their study relied exclusively on visualizationof very small areas and did not assess or quantifyporosity via volumetric methods.

Scanning electron microscopy of Barnett Shalesuggested that at low-maturity (<0.7% Ro) organicpores are sporadic or absent, and that the few na-nopores present are elongate and oriented parallel toboundaries of organic particles, whereas at highermaturity more than 0.8% Ro, an abundance of na-nopores is present in organic matter (Loucks et al.,2009). Our study supports the ideas that (1) po-rosity is altered duringmaturation, and (2) organicmatter is largely responsible for porosity changesbecause of the transformation of its convertible car-bon (Jarvie, 1991) into liquids and hydrocarbon,

Figure 14. Relationships between mineral contents and (A) mi-cropore, (B) mesopore, and (C)macropore volumes of shale samples.

B

A

IL-1

472-1

MM4

NA2

IL-5

R =0.35%o

R =0.55%o

R =0.65%o

R =1.15%o

R =1.41%o

IL-1

472-1

MM4

NA2

IL-5

R =0.35%o

R =0.55%o

R =0.65%o

R =1.15%o

R =1.41%o

Figure 15. (A) Absolute micro-, meso-, and macropore volumesand (B) percentages of micro-, meso-, and macroporosities withincreasing maturity based on gas adsorption data.

Page 19: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

nitrogen, and carbon dioxide gases. Conversion ofkerogen to hydrocarbons can reach 10% by weightas early as Ro 0.6% (Peters, 1986), and therefore,

pores seen in alginite of sample MM4 at Ro 0.55%(Figure 2C) could be an example of such earlytransformation, which would extend further athigher maturity and generate additional nanopores(Loucks et al., 2009).

Our study, however, demonstrates also thatporosity can be reduced at some levels during mat-uration. Within our set of samples, this is best ex-emplified by sample IL-5 having Ro 1.15% duringthe early part of the late mature stage (Figures 15–17). It is important to understand that porosimetricgas adsorption and mercury intrusion methods candetect only pores that are not occupied by fluidsthat are relatively incompressible like water, oil, andbitumen. In other words, pores that are prefilledwithan incompressible fluid cannot accept significantamounts of new gas or fluids. We interpret the ob-served decline in porosity during the late maturestage as the result of prior pore filling by oil or for-mation of solid bitumen that reduced the availableopen pore space and restricted gas flow (Figure 17).Indeed, sample IL-5 has a significantly lower per-meability than any other sample (Figure 10). This

B

A

MICPIL-1

472-1

MM4

NA2

IL-5

R =0.35%o

R =0.55%o

R =0.65%o

R =1.15%o

R =1.41%o

IL-1

472-1

MM4

NA2

IL-5

R =0.35%o

R =0.55%o

R =0.65%o

R =1.15%o

R =1.41%o

Figure 16. (A) Absolute micro-, meso-, and macropore volumesand (B) percentages of micro-, meso-, and macroporosities withincreasing maturity based on the mercury intrusion capillary pres-sure (MICP) technique.

3(cm /g)

Figure 17. Schematic diagram relating the observed changes in porosity to maturation and hydrocarbon generation. Open circles on theporosity curves indicate the positions of the samples studied.

Mastalerz et al. 1639

Page 20: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

retention of hydrocarbons could be caused by block-ing of pore throats by bitumen, asphaltenes, etc.(Jarvie et al., 2007). To test if this reduction of porevolume is related to pore filling by bitumen, weextracted two most mature samples (IL-5 and IL-1)with dichloromethane and reanalyzed for mesoporeand micropore volumes. Late mature sample IL-5yielded significantly more extract, and this sampledramatically increased the micropore volume afterextraction (Table 5), as calculated from significantlylarger volumes of absorbed CO2 during low-pressureadsorption analysis, with mesopore volume remain-ing comparable to that before extraction. These ob-servations document that bitumen present in smallpores contributes to the reduction of pore volumes atthis maturity level.

With further thermal maturation, the sub-sequent increase in porosity would be caused bysecondary cracking of oil and bitumen to gas, un-blocking of previously filled pores, and expulsion offluid products. In the postmature sample IL-1,some pores are still blocked by bitumen, as indicatedby increased micropore and mesopore volumes inthe dichloromethane-extracted sample (Table 5). Itfollows that the evolution in porosity of shales withincreasing maturity may be shifted between variousbasins because of differences in organic matter typeand contrasting abilities of “convertible” carbonpools to be transformed to fluids.

This study directly compares porosity andpore-size distribution data from gas adsorptionand MIPC techniques. As a rule, MICP providesinformation about specific pore-throat sizes thatallow the access to pore volumes. Pore-throat sizeis an important parameter because it correlateswell with rock permeability and porosity (Nelsonand Batzle, 2006; Nelson, 2009). Several reasonsto expect method-related differences in pore-sizedistribution and pore-volume data from gas ad-sorption and MICP techniques when measuringshales exist. First of all, in contrast to He and CO2,mercury cannot access micropores. In general,high pressure (379 MPa in this study) is requiredfor mercury to access the pore network in low-permeability rocks. The use of high pressure en-tails the possibility of introducing artifacts bybreaking particles, opening new pores, and com-

1640 Geohorizon

pressing the existing pore structure (Bustin et al.,2008; Kuila and Prasad, 2011). In addition, MICPmeasures pore throats (i.e., the largest entrance to-ward a pore, but not the actual inner size of a pore),whereas PSD calculations from gas adsorption tech-niques considering removal of gas from the poresof a specific inner diameter (e.g., Clarkson et al.,2012). Therefore, theMICP technique is expectedto show smaller pore sizes within themesopore andmacropore ranges compared to adsorption techni-ques or optical techniques. Our comparative resultsfrom gas adsorption and MICP techniques demon-strate that, regardless of some discrepancies betweenthese techniques, the respective pore volume andpore-size-distribution data are of the same magni-tude and show similar trends with maturation.

CONCLUSIONS

1. Shales that are rich in organic matter show sig-nificant changes in porosity characteristics withincreasing maturity. A suite of immature (Ro

0.35%, sample 472-1) to postmature (Ro 1.41%,sample IL-1) New Albany Shale samples ex-presses large differences in total porosity, micro-porosity, mesoporosity, and macroporosity. Thesedifferences are caused by several factors, includingthermal maturity, mineralogical composition,and organic matter content. The role of organicmatter type should be minimal in our sampleset, because all the shale samples contain kerogentype II where marine alginite and amorphiniteare the dominant contributors to total organicmatter. Shales containing kerogen type I or kero-gen type III may express a different pattern ofporosity changes through maturation.

2. The influence of organic matter content on po-rosity characteristics is demonstrated by two earlymature shale samples, MM4 (Ro 0.55%) andNA2 (Ro 0.65%) which have drastically differentTOC contents of 13.0 and 5.3 wt. %, respec-tively. At this level of maturity, the higher TOCcontent does not influence the total pore volume,although it increases the micropore contributionto total porosity. This influence of organic mat-ter content on porosity may possibly vary with

Page 21: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

maturity. When considering the entire set ofshale samples having different maturities andTOC contents, no significant relationships areapparent between TOC contents and total porevolume, micropore volume, or macropore vol-ume. We suggest that relationships may be rec-ognizable among shales of similar maturity, asdocumented in some previous studies, but largedifferences in maturity obscure the relativelyminor role of TOC contents affecting porosity inour sample set.

3. The mineralogical composition of shale is a sig-nificant factor influencing porosity characteristics.Total porosity appears to increase with clay andquartz contents and decreases with carbonatecontent in our suite of samples.Microporous claystogether with organic matter are the main con-tributors to micropore volumes in shales. How-ever, the influence of the mineralogical compo-sition on shale porosity is not dominant and doesnot allow the use ofmineral parameters to predictpore characteristics and pore abundance in theseshales. We acknowledge that our sample suite islimited, and additional samples reflecting variousmineralogical compositions could help to betterunderstand the influence of minerals on shaleporosity.

4. We suggest that the large, and perhaps evendominant control of maturity on the develop-ment of porosity in our suite of shale samplesmasked effects from compositional differences.Increasingmaturity is paralleledby several changesin porosity-related characteristics: (1) total po-rosity is reduced greatly during the transitionfrom mature to late mature shale; (2) the largeincrease in total pore volumes between Ro 1.15%and 1.41%witnesses the generation of new pores,and (3) changes in pore volumes are accompaniedby a rearrangement in pore-size proportions. Thetransition from early mature to late mature shalemakes micropores relatively more abundant andpreferentially eliminates mesopores, whereasfurther maturation to postmature shale createsmesopores. These changes are attributed dom-inantly to structural changes in organic matterin response to hydrocarbon generation and mi-gration. The evolution of organic porosity in our

suite of shale samples does not follow a mono-tonous trend with increasing maturity, but in-stead follows a nonlinear, trough-shaped paththat seems to respond to events related to hy-drocarbon generation (Figure 17). The genera-tion of new pores is related to the transforma-tion of organic matter into hydrocarbons in theearly mature stage and to secondary cracking inthe late mature stage, whereas the intermittentdecline in porosity is interpreted as a result of porefilling by oil and bitumen that reduces availableopen pore space and restricts gas flow.

5. Finally, our study directly compares porosity andpore-size-distribution data from gas-adsorption(He, CO2, N2) and MICP techniques. Despite afew discrepancies, the pore volume and pore-size-distribution data of the porosimetric meth-ods are of the same magnitude and show similartrends with maturation. In combination, the var-ious techniques provide superior synoptic infor-mation about pore characteristics in shales. Gasadsorption techniques in particular excel in pro-viding information about pore volumes, whereasMICP offers valuable quantification of pore-throatdiameters.

REFERENCES CITED

Bernard, S., R. Wirth, A. Schreiber, H. M. Schulz, and B.Horsfield, 2012, Formation of nanoporous pyrobitumenresidues during maturation of the Barnett Shale (FortWorth Basin): International Journal of Coal Geology,v. 103, p. 3–11, doi:10.1016/j.coal.2012.04.010.

Bruant, R. G., A. J. Guswa,M. A. Celia, and C. A. Peters, 2002,Safe storage of CO2 in deep saline aquifers: EnvironmentalScience and Technology, v. 36, no. 11, p. 240A–245A,doi:10.1021/es0223325.

Bustin, R. M., and C. R. Clarkson, 1999, Free gas storage inmatrix porosity: A potentially significant coalbeds re-source in low rank coals: Proceedings of the InternationalCoalbed Methane Symposium, Tuscaloosa, Alabama,p. 197–214.

Bustin, R.M., A.M.M. Bustin, X. Cui, D. J. K. Ross, andV. S.Murthy Pathi, 2008, Impact of shale properties on porestructure and storage characteristics: SPE Shale Gas Pro-duction Conference, Society of Petroleum Engineers, FortWorth, Texas,November 16–18, 2008, SPEPaper 119892,28 p., doi:10.2118/119892-MS.

Chalmers, G. R. L., and R.M. Bustin, 2006, The organic matterdistribution andmethane capacity of theLowerCretaceous

Mastalerz et al. 1641

Page 22: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

strata of northeastern British Columbia, Canada: Interna-tional Journal of Coal Geology, v. 70, p. 223–239, doi:10.1016/j.coal.2006.05.001.

Chalmers, G. R. L., and R. M. Bustin, 2008, Lower Creta-ceous gas shales in northeastern British Columbia: Part I.Geological controls on methane sorption capacity: Bulle-tin of Canadian Petroleum Geology, v. 56, p. 1–21,doi:10.2113/gscpgbull.56.1.1.

Chalmers,G. R. L., andR.M. Bustin, 2012,Geological evalua-tion of Halfway-Doig-Montney hybrid gas shale-tightgas reservoir, northeastern British Columbia: Marineand Petroleum Geology, v. 38, p. 53–72, doi:10.1016/j.marpetgeo.2012.08.004.

Chalmers, G. R., R. M. Bustin, and I. M. Power, 2012, Char-acterization of gas shale pore systems by porosimetry,pycnometry, surface area, and field emission scanningelectron microscopy/transmission electron microscopyimage analyses: Examples from the Barnett, Woodford,Haynesville, Marcellus, and Doig units: AAPG Bulletin,v. 96, p. 1099–1119, doi:10.1306/10171111052.

Clarkson, C. R., and R. M. Bustin, 1996, Variation in micro-pore capacity and size distribution with composition inbituminous coal of the Western Canadian sedimentarybasin: Implications for coalbed methane potential: Fuel,v. 75, p. 1483–1498, doi:10.1016/0016-2361(96)00142-1.

Clarkson, C. R., M. Freeman, L. He, M. Agamalian, Y. B.Melnichenko,M.Mastalerz, R.M. Bustin, A. P. Radliñski,and T. P. Blach, 2012, Characterization of tight gas reser-voir pore structure using USANS/SANS and gas adsorp-tion analysis: Fuel, v. 95, p. 371–385, doi:10.1016/j.fuel.2011.12.010.

Clarkson,C.R.,N. Solano,R.M.Bustin,A.M.M.Bustin,G.R.L.Chalmers, L. He, Y. B. Melnichenko, A. P. Radliñski,and T. P. Blach, 2013, Pore structure characterization ofNorth American shale gas reservoirs using USANS/SANS,gas adsorption, and mercury intrusion: Fuel, v. 103, p. 606–616, doi:10.1016/j.fuel.2012.06.119.

Curtis, M. E., R. J. Ambrose, C. H. Sondergeld, and C. S. Rai,2011a, Investigation of the relationship between organicporosity and thermal maturity in the Marcellus Shale:North American Unconventional Gas Conference and Ex-hibition, The Woodlands, Texas, June 14–16, 2011: SPEPaper 144370, 4 p., doi:10.2118/144370-MS.

Curtis, M. E., R. J. Ambrose, C. H. Sondergeld, and C. S. Rai,2011b, Transmission and scanning electron microscopy in-vestigation of pore connectivity of gas shales on the nano-scale: North American Unconventional Gas Conferenceand Exhibition, The Woodlands, Texas, June 14–16, 2011,SPE Paper 144391, 10 p., doi:10.2118/144391-MS.

Curtis, M. E., B. J. Cardott, C. H. Sondergeld, and C. S. Rai,2012, Development of organic porosity in theWoodfordShale with increasing thermal maturity: InternationalJournal of Coal Geology, v. 103, p. 26–31, doi:10.1016/j.coal.2012.08.004.

Fishman, N. S., P. C. Hackley, H. A. Lowers, R. J. Hill, S. O.Egenhoff, D. D. Eberl, and A. E. Blum, 2012, The natureof porosity in organic-richmudstones of the Upper JurassicKimmeridge Clay Formation, North Sea, offshore UnitedKingdom: International Journal of Coal Geology, v. 103,p. 32–50, doi:10.1016/j.coal.2012.07.012.

1642 Geohorizon

Gan, H., S. P. Nandi, and P. L. Walker Jr., 1972, Nature of theporosity in American coals: Fuel, v. 51, p. 272–277, doi:10.1016/0016-2361(72)90003-8.

Giesche, H., 2006, Mercury porosimetry: A general (practical)overview: Particle and Particle Systems Characterization,v. 23, p. 9–19, doi:10.1002/ppsc.200601009.

Gregg, S. J., and K. S. W. Sing, 1982, Adsorption, surfacearea, and porosity: New York, Academic Press, 303 p.

Jarvie, D. M., 1991, Total organic carbon (TOC) analysis, inR. K. Merrill, ed., Source and migration processes andevaluation techniques: AAPG Treatise of PetroleumGeology, Handbook of Petroleum Geology, p. 113–118.

Jarvie, D. M., R. J. Hill, T. E. Ruble, and E. M. Pollastro,2007, Unconventional shale-gas systems: The Mississip-pian Barnett Shale of north-central Texas as one modelfor thermogenic shale-gas assessment: AAPG Bulletin,v. 91, p. 475–499, doi:10.1306/12190606068.

Javadpour, F., 2009, Nanopores and apparent permeabilityof gas flow in mudrocks (shales and siltstones): Journalof Canadian Petroleum Technology, v. 48, p. 16–21,doi:10.2118/09-08-16-DA.

Kuila, U., andM. Prasad, 2011, Surface area and pore-size dis-tribution in clays and shales: SPE Annual Technical Con-ference and Exhibition, Denver, October 30–November2, 2011, SPE Paper 146869, 13 p., accessed November10, 2012, http://www.spe.org/atce/2011/pages/schedule/tech_program/documents/spe146869%201.pdf.

Levy, J. H., S. J. Day, and J. S. Killingley, 1997, Methanecapacities of Bowen Basin coals related to coal proper-ties: Fuel, v. 76, p. 813–819, doi:10.1016/S0016-2361(97)00078-1.

Lis, G. P., M. Mastalerz, A. Schimmelmann, M. Lewan, andA. B. Stankiewicz, 2005, FTIR absorption indices forthermal maturity in comparison with vitrinite reflec-tance Ro in type-II kerogens from Devonian black shales:Organic Geochemistry, v. 36, p. 1533–1552, doi:10.1016/j.orggeochem.2005.07.001.

Lis, G. P., A. Schimmelmann, and M. Mastalerz, 2006, D/Hratios and hydrogen exchangeability of type-II kerogenwith increasing thermal maturity: Organic Geochemis-try, v. 37, p. 342–353, doi:10.1016/j.orggeochem.2005.10.006.

Lis, G. P., M. Mastalerz, and A. Schimmelmann, 2008, In-creasing maturity of kerogen type II reflected by alkylben-zene distribution from pyrolysis–gas chromatography–mass spectroscopy: Organic Geochemistry, v. 39,p. 440–449, doi:10.1016/j.orggeochem.2008.01.007.

Loucks, R.G., R.M.Reed, S.C.Ruppel, andD.M. Jarvie, 2009,Morphology, genesis, and distribution of nanometer-scale pores in siliceous mudstones of the MississippianBarnett Shale: Journal of Sedimentary Research, v. 79,p. 848–861, doi:10.2110/jsr.2009.092.

Loucks, R. G., R. M. Reed, S. C. Ruppel, and U. Hammes,2012, Spectrum of pore types and networks in mud-rocks and a descriptive classification for matrix-relatedmudrock pores: AAPG Bulletin, v. 96, no. 6, p. 1071–1098, doi:10.1306/08171111061.

Mastalerz, M., A. Drobniak, D. Strąpoć, W. Solano Acosta,and J. Rupp, 2008, Variations in pore characteristicsin high volatile bituminous coals: Implications for

Page 23: Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion

coalbed gas content: International Journal of Coal Geol-ogy, v. 76, p. 205–216, doi:10.1016/j.coal.2008.07.006.

Mastalerz, M., L. He, B. Y. Melnichenko, and J. A. Rupp,2012, Porosity of coal and shale: Insights from gas ad-sorption and SANS/USANS techniques: Energy andFuels, v. 26, p. 5109–5120, doi:10.1021/ef300735t.

McAuliffe, C. D., 1979, Oil and gas migration—Chemical andphysical constraints: AAPG Bulletin, v. 63, p. 761–781.

Milliken, K. L., M. Rudnicki, D. A. Awwiller, and T. Zhang,2013, Organic matter-hosted pore system, MarcellusFormation (Devonian), Pennsylvania: AAPG Bulletin,v. 97, p. 177–200, doi:10.1306/07231212048.

Modica, C. J., and S. G. Lapierre, 2012, Estimation of kerogenporosity in source rocks as a functionof thermal transforma-tion: Example from the Mowry Shale in the Powder RiverBasin of Wyoming: AAPG Bulletin, v. 96, p. 87–108,doi:10.1306/04111110201.

Nelson, P. H., 2009, Pore–throat sizes in sandstones, tightsandstones, and shales: AAPG Bulletin, v. 93, p. 329–340, doi:10.1306/10240808059.

Nelson, P. H., and M. L. Batzle, 2006, Single–phase permeabil-ity, in J. Fanchi, ed., Petroleum engineering handbook:Volume I.General engineering: Richardson, Texas, Societyof Petroleum Engineers, v. 1, p. 687–726.

Orr, C., 1977, Surface area measurement, in I. M. Kolthoff,P. J. Elving, and F. H. Stross, eds., Treatise on analyticalchemistry: Part III. Analytical chemistry in industry:New York, John Wiley and Sons, v. 4, p. 321–358.

Passey, Q. R., K. M. Bohacs, W. L. Esch, R. Klimentidis, and S.Sinha, 2010, From oil-prone source rock to gas-producingshale reservoir—Geologic and petrophysical characteriza-tion of unconventional shale–gas reservoirs: InternationalOil and Gas Conference and Exhibition, Beijing, China,June 8–10, 2010, SPE Paper 131350, 29 p., doi:10.2118/131350-MS.

Peters, K. E., 1986, Guidelines for evaluating petroleum

source rock using programmed pyrolysis: AAPGBulletin,v. 70, p. 318–329.

Ross,D. J. K., andR.M. Bustin, 2009, The importance of shalecomposition and pore structure upon gas storage poten-tial of shale gas reservoirs: Marine and Petroleum Geol-ogy, v. 26, p. 916–927, doi:10.1016/j.marpetgeo.2008.06.004.

Schieber, J., 2010, Common themes in the formation and pre-servation of intrinsic porosity in shales and mudstones—Illustrated with examples across the Phanerozoic: Societyof PetroleumEngineers Unconventional Gas Conference,Pittsburgh, Pennsylvania, February 23–25, 2010, SPE Pa-per 132370, 10 p., doi:10.2118/132370-MS.

Strąpoć, D., M. Mastalerz, A. Schimmelmann, A. Drobniak,and N. R. Hasenmueller, 2010, Geochemical constraintson the origin and volume of gas in the New Albany Shale(Devonian–Mississippian), eastern Illinois Basin: AAPGBulletin, v. 94, p. 1713–1740, doi:10.1306/06301009197.

Swanson, B. F., 1981, A simple correlation between permeabil-ities and mercury capillary pressures: Journal of PetroleumTechnology, v. 33, p. 2498–2504, doi:10.2118/8234-PA.

Taylor, G. H., M. Teichmüller, A. Davis, C. F. K. Diessel, R.Littke, and R. Robert, 1998, Organic petrology: Berlin,Stuttgart, Gebrüder Borntraeger, 704 p.

Vandenbroucke, M., 1993, Migration of hydrocarbons, inM. L.Bordenave, ed., Applied petroleum geochemistry: Paris,Editions Technip, p. 122–148.

Valenza, J. J., N. Drenzek, F. Marques, M. Pagels, and M.Mastalerz, 2013, Geochemical controls on shale micro-structure: Geology, v. 41, p. 611–614, doi:10.1130/G33639.1.

Washburn, E. W., 1921, Note on a method of determining thedistribution of pore sizes in a porousmaterial: Proceedingsof the National Academy of Science, U.S.A., v. 7, p. 115–116.

Webb, P. A., and C. Orr, 1997, Analytical methods in fine parti-cle technology: Norcross, Micromeritics Instrument, 301 p.

Mastalerz et al. 1643