lable at ScienceDirect

Marine and Petroleum Geology 75 (2016) 181e191

Contents lists avai

Marine and Petroleum Geology

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

Research paper

The lithofacies and reservoir characteristics of the Upper Ordovicianand Lower Silurian black shale in the Southern Sichuan Basin and itsperiphery, China

Chao Han a, b, Zaixing Jiang a, *, Mei Han b, Minghao Wu a, Wen Lin c

a School of Energy Resources, China University of Geosciences, Beijing 100083, Chinab College of Earth Science and Engineering, Shandong University of Science and Technology, Qingdao 266510, Chinac Langfang Branch, Research Institute of Petroleum Exploration and Development, PetroChina, Langfang, Hebei, 065007, China

a r t i c l e i n f o

Article history:Received 11 November 2015Received in revised form2 April 2016Accepted 17 April 2016Available online 20 April 2016

Keywords:LithofaciesReservoir characteristicsBlack shaleShale gasPore structureSouthern Sichuan Basin and its periphery

* Corresponding author.E-mail address: jiangzx@cugb.edu.cn (Z. Jiang).

http://dx.doi.org/10.1016/j.marpetgeo.2016.04.0140264-8172/© 2016 Elsevier Ltd. All rights reserved.

a b s t r a c t

The black shale of the Upper Ordovician and Lower Silurian is a significant target for shale gas explo-ration in the Southern Sichuan Basin. In this study, we introduced a lithofacies classification for shalebased on rock mineral composition. Because the pore structure of gas shale reservoirs are complex andgreatly affect the gas storage and transport in shale, four different methods of low-pressure nitrogen gasadsorption, high-pressure mercury intrusion, scanning electron microscopy (SEM) and gas expansionmethods were used to investigate the reservoir pore structure and storage space. Combined with X-raydiffraction, total organic matter content (TOC), gas content, methane adsorption, porosity and perme-ability and wireline data, the key factors that affect shale gas content and storage of shale gas wereanalyzed.

According to our shale classification, three main lithofacies, i.e., calcareous mudstone (CM), highcalcareous mixed mudstone (HCMM) and low calcareous mixed mudstone (LCMM), were identified inour study area. The experiment indicates that the shale has a high TOC, thermal maturity and gascontent. The methane adsorption isotherms show that the sorbed gas content has a positive correlationwith TOC. Difference of TOC between the three is small. Nitrogen adsorption indicates that mesoporesdominate the shale pore composition and organic matter is the main source of mesopores. Mercuryinjection shows that the macropore volume accounts for approximately 14e22% of the total pore volume.No obvious pore structure differences were found for different lithofacies. However, taking the porosityinto consideration, LCMM has the largest macropore volume. The SEM observations revealed that organicmatter pores, interparticle pores and intraparticle pores are the main pore types. Interparticle poresmainly develop in LCMM which has a relatively high quartz and low carbonate content, it may be themajor reason for high porosity, high macropore volume, and high free gas content of LCMM. Free gas andabsorbed gas are the major storage form of shale gas. Given similar temperature and pressure conditions,the total organic matter content is the major factor that affects the adsorbed gas content. Therefore, withhigh free gas and adsorbed gas content, LCMMs with high total organic matter contents would be themost favorable type of lithofacies in this region.

© 2016 Elsevier Ltd. All rights reserved.

1. Introduction

Due to the technological innovations of horizontal drilling andhydraulic fracturing, gas production from shale has dramaticallyimproved (Wang and Carr, 2012a, b; Curtis, 2002, 2012; Clarkson

et al., 2012a). The recent frontier research on gas shale hasfocused on lithofacies classification (Hickey and Henk, 2007;Loucks and Ruppel, 2007; Dong et al., 2015), lithofacies identifica-tion (Wang et al., 2013, 2014a; Chang et al., 2000, 2002), deposi-tional environments (Stow et al., 2001; Gross et al., 2015) andreservoir characteristics (Loucks et al., 2009, 2012; Curtis et al.,2012; Cao et al., 2015). Lithofacies and reservoir characteristicsare two important focuses in the study of gas shale.

Lithofacies, the basic properties of rocks, has been used in

C. Han et al. / Marine and Petroleum Geology 75 (2016) 181e191182

geology, stratigraphy and sedimentology for more than seventyyears (Wang and Carr, 2012b). It has a close connection with min-eral components, organic matter (Wang and Carr, 2012a, b) andreservoir properties (Chang et al., 2002; Jungmann et al., 2011). Aconsiderable amount of lithofacies research has been conducted incarbonate and siliciclastic reservoirs, in which wireline logs havebeen widely employed for carbonate and siliciclastic lithofaciesidentification (Chang et al., 2000; Saggaf and Nebrija, 2003). Asound understanding of the connection between shale lithofaciesand reservoir ability could provide guidelines for making modestassessments of undiscovered hydrocarbon resources and designingproper exploration programs. A limited number of shale reservoirlithofacies studies have focused on the Barnett Shale, MarcellusShale, New Albany Shale, and little concerns on the Upper Ordo-vician and Lower Silurian shale in the Southern Sichuan Basin.

Shale reservoirs are characterized as a self-contained sourcereservoir system. Abundant gas can be stored in pores and fracturesin the form of free gas, absorbed gas and solution gas (Curtis, 2002;Ross and Marc Bustin, 2007), free gas and sorbed gas are two majorforms (Labani et al., 2013). Shale reservoirs usually have extremelylow permeability, relatively low porosity and contain extensivenanopores, the pore characterization and pore structure is of greatsignificance for gas storage and flow capacity (Ross and MarcBustin, 2009; Slatt and O'Brien, 2011; Wang et al., 2014a,b; Yanget al., 2014). To reduce exploration risks and achieve economicalproduction, it is important to understand the pore structure,sorption and potential gas capacities of shale reservoirs (Ross andMarc Bustin, 2007, 2009; Yang et al., 2014). The InternationalUnion of Pure and Applied Chemistry (IUPAC) made a pore classi-fication based on pore diameter: micropores (<2 nm), mesopores(2e50 nm) and macropores (>50 nm). Shale has a wide pore sizedistribution ranging from micropores to mesopores and macro-pores (Loucks et al., 2009). Loucks et al. (2012) created a descriptiveclassification for matrix-related pores, interparticle pores foundbetween particles and crystals, intraparticle pores located withinparticles and organic matter pores located within organic matter.Assessments of pore features are typically made by quantitativeand visual qualitative analyses (Clarkson et al., 2012a, 2012b; Wanget al., 2014b). FE-SEM images are the most common way to quali-tatively characterize pore systems (Curtis et al., 2012). Quantitativeanalyses include high-pressure mercury intrusion (HPMI), low-pressure nitrogen and CO2 gas adsorption (Clarkson et al., 2012b).Low-pressure nitrogen and CO2 gas adsorption were used formicro-mesopore analysis, and high-pressure mercury was used formacropore characterization (Rouquerolt et al., 1994).

In China, the potential shale gas resources could reach approx-imately 26 � 1012 m3 (Chen et al., 2011), and increasing shale gasexploration has been conducted in Sichuan Basin, from which theupper OrdovicianWufeng shale and Lower Silurian Longmaxi shalehave been identified as major exploration targets because of theirhigh organic matter abundance, considerable thickness and highmaturity (Chen et al., 2011, 2014; Cao et al., 2015). However, to date,a lithofacies classification has not been performed, which coulddirectly reflect reservoir and pore structure. The major goals of ourwork are to construct a systematic classification for the upperOrdovician Wufeng shale and Lower Silurian Longmaxi shale,which is useful for reservoir comparisons, and to provide furthertheoretical support for future shale gas exploration in similarbasins.

2. Geologic setting and stratigraphy

The Sichuan Basin is a tectonically stable sedimentary basin,with an area of 180 � 103 km2. It is located in southwestern Chinabordered by the Micang-Daba Mountains to the north, the

Songpan-Ganzi Terrane (SGT) and Longmen Mountains to the westand the Xuefeng Mountains to the east. The study area is located inthe Southern Sichuan Basin (Fig. 1). The Sichuan Basin underwenttwo tectonic deposition stages, the Sinian-Middle Triassic passivecontinental margin stage and the Late Triassic-Eocene forelandbasin stage (Liang et al., 2014). A series of tectonic movementsformed the current structural framework. The Sichuan Basin iscovered by the strata from the Neoproterozoic to the Quaternary(Liang et al., 2014; Guo et al., 2014). The Sinian-Middle Triassicdeposits mainly consist of thick marine carbonates. In the lateTriassic, influenced by the closure of the Palaeo-Tethys Ocean andthe subduction of the oceanic crust of the Yangtze Plate, theSichuan Basin evolved from marine-terrigenous transitionaldeposition to terrigenous clastic rocks. In the late Ordovician andearly Silurian, two third-scale global transgressions occurred,resulting in sedimentation of theWufeng and Longmaxi shale (Guoet al., 2014).

The Wufeng and Longmaxi shales are widely distributed in thestudy area. The Wufeng shale has a thickness ranging from 5 to11 m. The lowerWufeng shale is composed of a series of graptolite-rich black shale. The Guanyinqiao Member, which is the upper partof the Wufeng shale, has a thickness of approximately 0.2e0.6 mand higher calcium content. The Longmaxi shale also has a largedistribution, with a considerable thickness of approximately 200m.The lower Longmaxi shale also contains graptolite-rich black shaleand high total organic matter content, whereas the upper Longmaxishale consists of gray and sandy shale, with low total organicmatter content. Our study interval includes the Wufeng and lowerLongmaxi black shale, with a thickness of approximately 40 m andabundant organic matter.

3. Samples and methods

A total of 138 core samples collected in the study areawere usedto analyze the geochemical and mineral characteristics. The gascontent determination of 35 samples was performed when a freshcore was taken out. Eighteen core plugs were used for heliumporosimetry and nitrogen permeability. Then, they were preparedas small blocks, with the remaining offcuts and other samplescrushed to powder with a diameter of less than 250 mm for the totalorganic carbon (TOC) content measurement and low temperaturenitrogen adsorption. Mercury intrusion was performed using a9510-Ⅳ porosimeter. The porosity was tested by a ULTRAPORE-200A using the helium expansion method. Permeability wasmeasured using an ULTRA-PERMTM200. The TOC of all sampleswere measured using a LECO CS-230 carbon analyzer. A RINT-TTR3was used for X-ray diffraction at a voltage of 45 kV and a current of100 mA to test the mineral content of all samples. Ar-ion-beammilling could produce a relatively flat surfaces which are suitablefor highmagnification imaging. In our study, eight samples pro-cessed by an Ar-ion-beam were observed using FEI Quata200Fequipped with an energy-dispersive spectrometer (EDS). Anothereight unmilled samples were observed using a HITACHI S-4800 todetermine the microstructure and morphology of the minerals andorganic matter.

The thermal maturity was tested using optical microscopy. Themost favored way of assessing the thermal maturity is vitrinitereflectance (VRo), however, Longmaxi shale contain no vitrinite(Liang et al., 2014) for VRo measurement. Because VRo has a strongrelationship with bitumen reflectance (BRo) (Jacob, 1989;Schoenherr et al., 2007), we were able to obtain VRo usingSchoenherr's equation in cases in which solid bitumen was presentand vitrinite was absent. The equation is as follows:

Fig. 1. Geographic location, lithological column and mineral composition profile of the Longmaxi andWufeng formation. (a) Location of Sichuan Basin; (b) Location of the study area(modified according to Wang et al., 2013); (c) Lithological column and mineral composition profile of Longmaxi and Wufeng formation, Well YS106.

C. Han et al. / Marine and Petroleum Geology 75 (2016) 181e191 183

VRo ¼ ðBRo þ 0:2443Þ=1:0495 (1)

4. Results

4.1. Mineralogy and lithofacies

XRD and thin-section analysis show that the shale contains largeamounts of quartz, carbonate, clay and small amounts of pyrite andfeldspar (Fig. 2a). The quartz content ranges from 3.8 to 43.5 wt%,with an average of 21 wt%, and most of the silica shows a crypto-crystalline texture, suggesting an authigenic origin. The feldsparranges from 0.8 to 14.1 wt%, with an average of 7 wt%. The car-bonate ranges from 8.9 to 65 wt%, with an average of 32 wt%. Clay,mainly including illite, chlorite and an illite-smectite mixed layer,ranges from 16 to 48.9 wt%, with an average of 34 wt%.

Modified based on Lazar's nomenclature guidelines for fine-grained sedimentary rocks (2015), the lithofacies were classified

into four main groups (Fig. 2b), including argillaceous mudstone(AM), siliceous mudstone (SM), calcareous mudstone (CM) andmixedmudstone (MM). The first three lithofacies (AM, SM, and CM)contain a single component that accounts for more than 50%. In themixed mudstone, no single component is greater than 50%. In ourclassification, we divide MM into high calcareous mixed mudstone(HCMM) and low calcareous mudstone (LCMM) according to thecarbonate content. The carbonate content of HCMM is greater than33%, whereas LCMMhas a carbonate content of less than 33%. In thestudy area, CM, HCMM and LCMM were identified (Fig. 3a). CMmainly shows massive structure with high carbonate contents,silica, calcite and the other minerals are distributed homoge-neously (Fig. 3b). HCMM mainly developed horizontal bedding(Fig. 3c), the laminas are always composed of bright and darklaminas, with the bright laminas mainly containing calcite, and thedark laminas mainly containing clay minerals. LCMM is mainlylaminated (Fig. 3d), with silica showing an obvious lamellardistribution.

Fig. 2. Whole rock mineral composition (a) and lithofacies classification (b).

Fig. 3. The main lithofacies and thin-section characteristics. (a) the lithofacies mainly include CM, HCMM and LCMM. (b) CM with massive structure and high carbonate content. (c)HCMM with horizontal bedding, contain bright calcite laminas and dark clay laminas. (d) LCMM with silica lamina.

C. Han et al. / Marine and Petroleum Geology 75 (2016) 181e191184

4.2. Organic geochemical characteristics

4.2.1. Organic matter and TOC contentOrganic matter particles have multiform size and shape, always

extend parallel to the bedding and are squeezed along rigid grains.The smallest measured diameter of organic matter particles is100 nm. We also observed organic matter particles up to 200 mm atlower magnification. The shale has a TOC content ranging from 0.24to 7.9 wt%, with an average of 2.4 wt%. LCMM showed the highestTOC content (average: 2.6%), followed by HCMM (average: 2.5%),with CM showing the lowest value (average: 2.2%).

4.2.2. Thermal maturityThermal maturity indicates that the organic matter trans-

formation rate of hydrocarbons and the maximum temperatureexperienced during the basin history. Under the microscope, novitrinite were found. We measured the BRo (table) and calculated

the VRo using Equation (1). The equivalent VRo ranges from 2.47% to2.7%, with an average of 2.54%. The equivalent VRo suggests thatshales are very high in maturity and have reached the dry gaswindow.

4.2.3. Gas contentThe gas content analyses of cores are in the range of 0.16e3.48

scc/g with an average of 1.96 scc/g. It is common to observe positivecorrelations between TOC and gas content in shale reservoirs(Wang and Carr, 2012a, b; Liang et al., 2014). In our study, the gascontent is also positively correlated with the TOC content (Fig. 4a).The adsorption isotherm represents the total sorbed gas content asa function of pressure and suggests that the sorbed gas content ofthe six measured samples has a positive correlationwith TOC in thestudy area (Fig. 4b). At the same pressure and temperature, the highTOC samples usually have a high methane adsorption capacity.

Fig. 4. Relationship between gas content and TOC (a) and methane adsorption isotherms (b).

C. Han et al. / Marine and Petroleum Geology 75 (2016) 181e191 185

4.3. Helium porosity and permeability

Compared with conventional reservoirs, shale has an extremelylow porosity and low permeability (Nelson, 2009). The heliumporosity of samples averaged 2.37% and ranged from 0.7% to 4.12%(Fig. 5a). Their permeability was between 0.0064 mD and 0.0111mD with an average of 0.0068 mD (Fig. 5a). Among the threelithofacies, LCMM showed the highest porosity (average: 2.9%),followed by HCMM (average: 2%), with CM showing the lowestporosity (average: 0.9%) (Fig. 5b).

4.4. Nitrogen adsorption isotherm analysis

Nitrogen adsorption can be applied to study mesopores(2e50 nm) (Mastalerz et al., 2013). The nitrogen adsorption iso-therms (at �196 �C) of the shale samples are shown in Fig. 5a.According to the IUPAC classification of adsorption isotherms (Sing,1985), the shapes of the isotherms (Fig. 6a) are type IV and arecharacterized by hysteresis loops, which are caused by capillarycondensation occurring in the mesopores. Labani et al. (2013)asserted that pore filling at a low relative pressures is related tomicropores and fine mesopores, and that pore filling at high pres-sures is correlated with large mesopores and macropores.

De Boer (1958) devised a classification including five types ofhysteresis loops and correlated them with different types of poreshapes. Based on this classification, all of the sample isothermsbelong to hysteresis type B and the pores can be considered as slit-type pores. IUPAC also developed a classification system (1985) foradsorption hysteresis, and according to this system, the sample

Fig. 5. The porosity and permeability values of the Longmaxi shal

isotherms belong to Type H4 loops associated with narrow slit-likepores. The pore volume distribution, derived from the nitrogenadsorption branch for the isotherms, is interpreted using BJH the-ory. The distribution is unimodal with a large peak at approxi-mately 4 nm in the diameter range of 1e100 nm (Fig. 6b). Thecumulative pore volume curve shows that the cumulative porevolume of the shale ranges between 0.0081 and 0.0234 ml/g withthe predominant portion of the pore volume from mesopores witha diameter ranging from 2 to 5 nm (Fig. 7a), which increases withincreasing TOC (R2 ¼ 0.86) (Fig. 7b).

The nitrogen isotherms of kerogen are shown in Fig. 8a. Theisotherms are also type IV and contain hysteresis.

loops. The hysteresis loops are more similar with Type H3 loopsbecause they do not show any limiting adsorption at high p/po. Thepore volume distribution, interpreted by BJH theory, exhibitsbimodal characteristics. The peak at a diameter of approximately4 nm (Fig. 8b) is correlated with the pore volume distribution of thesamples.

4.5. Mercury intrusion and macropore porosity

Mercury intrusion can be used to detect meso-macropores(2 nme100 mm) (Kaufmann et al., 2009). Mercury intrusion isbased on the relationship between the amount of mercury intru-sion and pressure. The pressure can be converted to the pore radiususing the following equation:

Pc ¼ 0.735/rc.,

e (a) and the porosity comparison of different lithofacies (b).

Fig. 6. Nitrogen adsorption isotherms collected for the shale samples (a) and pore volume distribution (b).

Fig. 7. Shale cumulative pore volume and its relationships with TOC.

Fig. 8. Nitrogen isotherms collected for the kerogen (a) and pore volume distribution with pore size (b).

C. Han et al. / Marine and Petroleum Geology 75 (2016) 181e191186

where rc is the pore radius (mm) at the pressure Pc (MPa). Thisequation is simplified from the Washburn Equation. By measuringthe amount of mercury intruded into a pore system at constantpressure, we obtained the pore volume. Before this experiment, wemeasured the Helium porosity, volume and density of the sample.Combined with the mercury intrusion volume, the mercury injec-tion saturation was calculated (Fig. 9a). The mercury injectionsaturation (Fig. 9a) shows that macropores (pore radius>25 nm)account for 14.4e22% (average: 17.4%) of the total pore volume, andthe remaining volume was occupied by micro-mesopores. Thismercury injection saturationwas thenmultiplied by the porosity toobtain the percentage of macropores in the shale samples, which istermed the macropore porosity. The LCMM showed the highest

macropore porosity, followed by HCMM, with CM showing thelowest value (see Fig. 9b).

4.6. Reservoir storage space

The pore type are abundant in the study area. According to theclassification of Loucks et al. (2012), we could identify organicmatter pores, intraparticle pores and interparticle pores.

4.6.1. Organic matter poresOM pores are created by the generation and expulsion of hy-

drocarbons during organic matter thermal evolution (e.g. Jarvieet al., 2007; Loucks et al., 2009; Lu et al., 2015). It is commonly

Fig. 9. Mercury intrusion curve (a) and macropore porosity comparison (b).

C. Han et al. / Marine and Petroleum Geology 75 (2016) 181e191 187

suggested that organic matter pores are related to the organicmatter type. Marine sourced type II kerogen appears to have morepotential to form organic matter than terrestrial type III kerogen(Loucks et al., 2012). The relationship between organicmatter poresand thermal maturity is still controversial. Valenza et al. (2013)asserted that a higher thermal maturity is accompanied by an in-crease in surface area and pore volume and the decrease in averagepore size. Bernard et al. (2012a, 2012b) suggested that only samplesin the gas window contain organic matter pores. However, Reedet al. (2014) found nanopores in oil-window maturity kerogenand Loucks et al. (2014) suggest OM pores form not only in kergon,but also in solid bitumen and pyrobitumen.

In two dimensions, the organic matter pores were usually iso-lated with spherical to ellipsoidal shapes, and angular morphol-ogies were also observed. The long dimensions of visible organic

Fig. 10. Organic matter pore features. (a, b) Organic matter with diameters of less than 5 mmgood connectivity in the unmilled samples; (d) Organic matter pores were not found in or

matter pores range from a few nanometers to one hundred nano-meters (Fig. 10a, b). Through observations of unmilled samples, thepores could be connected to form a three-dimensional porenetwork (Fig. 10c). Based on the observations, we found organicmatter pores in the small organic matter particles, which have di-ameters of less than 5 mm; however, it is difficult to find organicmatter pores in the large organic matter particles (Fig. 10d).

4.6.2. Interparticle poresInterparticle pores are commonly observed between grains and

crystals such as quartz, feldspar, calcite and clay flocculates. Inyoung shallow-buried mud, interparticle pores are abundant,connected and can form an effective pore network (Velde, 1996;Milliken and Reed, 2010). However, they decrease sharply withincreasing overburden and diagenesis (Loucks et al., 2012). Using

containing abundant organic pores in the milled samples; (c) Organic matter pores haveganic matter particles with larger diameters.

C. Han et al. / Marine and Petroleum Geology 75 (2016) 181e191188

the scanning electron microscope, we could observe interparticlepores between calcite, quartz and clay minerals. Interparticle poresbetween calcite are always distributed along the sharp edges ofcalcite, with pore sizes ranging from 50 nm to 1 mm (Fig. 11a), andmainly develop in CM and HCMM. Interparticle pores betweenquartz have a smooth edge, with pore sizes reaching up to 200 nm(Fig. 11b), they mainly develop in LCMM. Interparticle pores be-tween clay minerals are difficult to observe on samples processedby an Ar-ion-beam. Most interparticle pores are primary pores andare seriously affected by diagenesis and mineral composition.

4.6.3. Intraparticle poresMost intraparticle pores are found within particles related to

authigenic crystals in the study area. However, primary intra-particle pores are difficult to find. Examples mainly include (1)intraparticle pores within framboidal pyrite (Fig. 12a), (2) intra-particle pores of calcite formed via partial dissolution or fluid in-clusions left during crystal formation (Fig. 12b,c), (3) cleavage-sheetintraparticle pores within clay and mica-mineral grains (Fig. 12d).

Pyrite framboids, composed of many small pyrite crystals, areformed in anoxic sea environments. Pore diameters between pyritecrystals range from 25 nm to 325 nm. Intraparticle pores withincalcite formed by dissolution, and we identified a maximum porediameter of 3 mm. Clay and mica-mineral grains develop cleavage-sheet intraparticle pores, with pore sizes of usually less than100 nm.

Fig. 11. Interparticle pore features: (A, B) Interparticle pores betwee

4.7. The analysis of shale gas composition calculated from wirelinedata

Shale gas is stored as free gas in the pores and fractures of thematrix. Adsorbed gas is found on the organic material and clayminerals, and a small amount of gas is dissolved in water or oil(Curtis, 2002; Gasparik et al., 2012, 2013). According to themethane adsorption isotherms and the formation temperature andpressure from the wireline data, the adsorbed gas content could becalculated. Using formation porosity, temperature, pressure andclay minerals, the free gas content was determined using a reser-voir saturation model. In our study, we focused on free andadsorbed gas. The gas content wireline data were provided bySchlumberger. Through the comparison of gas composition ofdifferent lithofacies, we determined that LCMM had the highestproportion of free gas, with an average of 29% inwell 107 and 49% inwell 106, whereas the proportion of free gas in CM was only 15% inWell 107 and 5% in Well 106 (Fig. 13).

5. Discussion

5.1. Shale lithofacies

The original definition of lithofacies, defined by Teicher in 1958,is an integration of all lithologic features in sedimentary rocks, suchas color, mineral composition, texture and structure. Lithofacieshave been widely used in sandstone and carbonate studies (e.g.

n calcite are irregular. (C, D) Interparticle pores around quartz.

Fig. 12. Intraparticle pore features. (A) Intraparticle pores exist between pyrite crystals. (B, C) Intraparticle pores in calcite. (D) cleavage-sheet intraparticle pores in mica.

Fig. 13. The gas composition comparison of Well 106 and Well 107.

C. Han et al. / Marine and Petroleum Geology 75 (2016) 181e191 189

Sahay et al., 2015; Chang et al., 2000; Saggaf and Nebrija, 2003). Thequalitative features of sedimentary rocks, including color, compo-nents, texture and grain-size distribution, are significant for theinterpretation of hydrodynamic conditions and sedimentary envi-ronment. However, shales are formed from fine-grained sedimentsof which the average grain size is less than 63 mm. They aredeposited in much more uniform hydrodynamic conditions andlow-action sedimentary environments; thus, the qualitative char-acteristics tend to be similar (Wang and Carr, 2012a, b). Therefore,the application of qualitative characteristics is limited. A practicalshale lithofacies classification must consider the specificity anddevelopment of the shale-gas reservoir. There are substantial dif-ferences between shale-gas reservoirs and conventional sandstoneand carbonate reservoirs. Horizontal drilling and hydraulic frac-turing play key roles in the development of shale gas (Curtis, 2002,

2012). The total organic carbon and the amount of hydrocarbonthat can be extracted are two main factors in well planning andhydraulic fracturing of unconventional resources. Brittleness couldbe directly measured to determine the ability of a formation tocreate avenues for hydrocarbons (Zhang et al., 2015).

Rock mineralogy has a significant influence on the efficiency ofhydraulic fractures (Jarvie et al., 2007), and high silica, feldspar andcarbonate content are known to correspond to a lower Poissonratios, higher Young's modulus values and greater brittleness (Dinget al., 2012; Soua, 2014). In our study area, the brittle mineralcontent is greater than 60% (Fig. 3a), and the total organic carbon orgas content is a key factor. In our shale lithofacies classification,shale is classified into five main groups: AM, SM, CM, LCMM andHCMM. Three lithofacies, including CM, LCMM and HCMM, areidentified. The gas content and reservoir characteristics of these

C. Han et al. / Marine and Petroleum Geology 75 (2016) 181e191190

three lithofacies has a clear distinction. LCMM is the best shalereservoir in the study area. With the help of conventional logs, it iseasy to identify different lithofacies in our study area. Our lith-ofacies classification has strong practicality in terms of reservoircomparisons.

5.2. Nanopore structure

Shale gas was stored in three forms (Zhou et al., 2014), and freegas and adsorbed gas are the two main forms (Labani et al., 2013;Jarvie et al., 2005). A better understanding of pore structure couldhelp to understand the gas storage mechanism. The combined useof nitrogen adsorption, high-pressure mercury injection and SEM isan effectivemethod for the precise description of the pore structureof shales. The nitrogen adsorption hysteresis loops show that shalepores are mainly composed of mesopores (IUPAC, 1985). Thecombination of mercury intrusion and helium porosity indicatesthat the macropore volume accounts for approximately 14e22% ofthe total pore volume, there are no obvious difference betweendifferent lithofacies. However, LCMM has the highest porosity,which leads to the highest macropore volume of the three lith-ofacies. Using SEM, we observed numerous interparticle macro-pores around quartz in the LCMM samples and relatively fewinterparticle macropores in the CM and HCMM samples. Due tosome depositional carbonate minerals (high-Mg calcite andaragonite) is metastable, the dissolution of parts of host rocks andre-precipitation of cements are routine during burial diagenesis(Jiang et al., 2015). The chemical compaction and associatedcementation responding to increasing thermal exposure andeffective stress during burial always lead to progressive destructionof porosity (Ehrenberg et al., 2012). With more carbonate content,the CM and HCMM will lost more primary pores than LCMM. Thehigher proportion of free gas in LCMM according to the wirelinedata provides additional evidence.

The shale sample pore volume distribution calculated from ni-trogen adsorption shows a peak at approximately 4 nm, which isconsistent with the pore volume distribution of kerogen. In addi-tion, high organic matter samples usually have higher peak valuesthan low organic matter samples and pore volume increases withincreasing TOC. Such evidence indicates that organic matter poresare the main sources mesopores in shale. Nitrogen adsorptionisotherm shows shale pores aremainly composed of silt-type pores.However, a significant portion of the organic matter pore are small,their average pore size is less than 10 nm, and it is not accessiblewith the resolution of SEM in our in our experiments. It is the mainreason why we could not observe silt-type organic matter pores.

5.3. The main control factor for shale gas reservoirs

Using SEM observations, we found that organic matter pores,intraparticle pores and interparticle pores act as the main gasstorage space. It is commonly suggested that the organic mattercontent, thermal maturity and organic matter type are the majorfactors influencing the development of organic matter pores (e.g.,Curtis, 2002, 2012; Loucks et al., 2012). The organic matter contentalways has a strong positive correlationwith the gas content (Wangand Carr, 2012a, 2012b; Liang et al., 2014; Ji et al., 2014). The organicmatter type also affects the organic matter pore generation. Type Ikerogen, with more “live carbon” (Pepper, 1991), can form moreorganic matter pores (Jarvie et al., 2007). Even the relationshipbetween organic matter pores and thermal maturity is controver-sial, the thermal evolution degree (Ro) of black shale is 2.54% onaverage in the study area, it is similar to the Jiaoshiba area which isknown for its great success in shale gas exploration and has anaverage Ro of 2.65% (Guo et al., 2014). Thermal maturity is not a key

factor in our area. The Longmaxi shale is mainly Type I kerogen andis high to very high inmaturity (Liang et al., 2014). Therefore, TOC isthe key factor that affects organic matter pores.

Interparticle pores mainly develop in LCMM which has a rela-tively high quartz and low carbonate content, it may be the majorreason for high porosity, high macropore volume, and high free gascontent of LCMM. CM and HCMM, which have higher carbonatecontents, have relatively low porosity and macropore volume. Freegas and absorbed gas are the major storage form of shale gas. Givensimilar temperature and pressure conditions, the total organicmatter content is themajor factor that affects adsorbed gas content.Therefore, LCMM with high total organic matter content tend tohave a higher free gas and adsorbed gas content, it is the mostfavorable type of lithofacies for exploration in this region.

6. Conclusions

Sedimentary lithofacies are classified into four main groupsbased on rock mineral composition: AM, CM, SM and MM. In thestudy area, CM and MM were identified, and MM was divided intoHCMM and LCMM according to the carbonate content. Shale poresare mainly composed of mesopores (in pore volume) and organicmatter is the main source of mesopores. Macropores account forapproximately 14e22% of total pore volume. The SEM observationsrevealed that organic matter pores, interparticle pores and intra-particle pores are the main pore types. Interparticle pores mainlydevelop in LCMM, it may be themajor reason for high porosity, highmacropore volume, and high free gas content of LCMM. Free gasand absorbed gas are the major storage forms of shale gas. In thesame region, temperature and pressure conditions of strata issimilar, the total organic matter content is the major controllingfactors that affect adsorbed gas content. Therefore, LCMM withhigh TOC would be the most favorable type of lithofacies in thisregion.

Acknowledgements

The work presented in this paper was supported by the NationalScience and Technology Special (Grant No. 2011ZX05009-002),SDUST Research Fund (Grant No. 2015TDJH101) and Chinese Na-tional Natural Science Foundation (Grant No. 41372108).

References

Bernard, S., Horsfield, B., Schulz, H.-M., Wirth, R., Schreiber, A., Sherwood, N., 2012a.Geochemical evolution of organic-rich shales with increasing maturity: a STXMand TEM study of the Posidonia Shale (Lower Toarcian, northern Germany).Mar. Pet. Geol. 31, 70e89.

Bernard, S., Wirth, R., Schreiber, A., Schulz, H.-M., Horsfield, B., 2012b. Formation ofnanoporous pyrobitumen residues during maturation of the Barnett shale (fortWorth basin). Int. J. Coal Geol. 103, 3e11.

Cao, T., Song, Z., Wang, S., Cao, X., Li, Y., Xia, J., 2015. Characterizing the porestructure in the Silurian and Permian shales of the Sichuan Basin, China. Mar.Pet. Geol. 61, 140e150.

Chang, H.C., Kopaska-Merkel, D.C., Chen, H.C., Durrans, S.R., 2000. Lithofaciesidentification using multiple adaptive resonance theory neural networks andgroup decision expert syste. Comput. Geosci. 26, 591e601.

Chang, H.C., Kopaska-Merkel, D.C., Chen, H.C., 2002. Identification of lithofaciesusing Kohonen self-organizing maps. Comput. Geosci. 28, 223e229.

Chen, S., Zhu, Y., Wang, H., Liu, H., Wei, W., Fang, J., 2011. Shale gas reservoircharacterisation: a typical case in the southern Sichuan Basin of China. Energy36, 6609e6616.

Chen, S., Zhu, Y., Qin, Y., Wang, H., Liu, H., Fang, J., 2014. Reservoir evaluation of thelower Silurian Longmaxi formation shale gas in the southern Sichuan Basin ofChina. Mar. Pet. Geol. 57, 619e630.

Clarkson, C.R., Jensen, J.L., Chipperfield, S., 2012a. Unconventional gas reservoirevaluation: what do we have to consider? J. Nat. Gas Sci. Eng. 8, 9e33.

Clarkson, C.R., Jensen, J.L., Pedersen, P.K., Freeman, M., 2012b. Innovative methodsfor flow-unit and pore-structure analyses in a tight siltstone and shale gasreservoir. AAPG Bull. 96, 355e374.

Curtis, J.B., 2002. Fractured shale gas systems. AAPG Bull. 86, 1921e1938.

C. Han et al. / Marine and Petroleum Geology 75 (2016) 181e191 191

Curtis, M.E., Sondergeld, C.H., Ambrose, R.J., Rai, C.S., 2012. Microstructural inves-tigation of gas shales in two and three dimensions using nanometer-scaleresolution imaging. AAPG Bull. 96, 665e677.

De Boer, J.H., 1958. The structure and properties of porous materials. In: Proceed-Ings of the Tenth Symposium of the Colston Research Society Held in theUniversity of Bristol. Butterworths, London, pp. 68e94.

Ding, W., Li, C., Li, C., Xu, C., Jiu, K., Zeng, W., Wu, L., 2012. Fracture development inshale and its relationship to gas accumulation. Geosci. Front. 3, 97e105.

Dong, T., Harris, N.B., Ayranci, K., Twemlow, C.E., Nassichuk, B.R., 2015. Porositycharacteristics of the Devonian Horn River shale, Canada: insights from lith-ofacies classification and shale composition. Int. J. Coal Geol. 141e142, 74e90.

Ehrenberg, S.N., Walderhaug, O., Bjerlykke, K., 2012. Carbonate porosity creation bymesogenetic dissolution: reality or illusion? AAPG Bull. 96, 217e225.

Gasparik, M., Ghanizadeh, A., Bertier, P., Gensterblum, Y., Bouw, S., Krooss, B.M.,2012. High-pressure methane sorption isotherms of black shales from theNetherlands. Energy Fuel 26, 4995e5004.

Gasparik, M., Bertier, P., Gensterblum, Y., Ghanizadeh, A., Krooss, B.M., Littke, R.,2013. Geological controls on the methane storage capacity in organic-richshales. Int. J. Coal Geol. 123, 34e51.

Gross, D., Sachsenhofer, R.F., Bechtel, A., Pytlak, L., Rupprecht, B., Wegerer, E., 2015.Organic geochemistry of Mississippian shales (Bowland Shale Formation) incentral Britain: implications for depositional environment, source rock and gasshale potential. Mar. Pet. Geol. 59, 1e21.

Guo, X., Li, Y., Liu, R., Wang, Q., 2014. Characteristics and controlling factors ofmicropore structures of the Longmaxi shale in the Jiaoshiba area, Sichuan Basin.Nat. Gas. Ind. B 1, 165e171.

Hickey, J.J., Henk, B., 2007. Lithofacies summary of the Mississippian Barnett shale,mitchell 2 T.P. Sims well, Wise county, Texas. AAPG Bull. 91, 437e443.

Jacob, H., 1989. Classification, structure, genesis and practical importance of naturalsolid oil bitumen (“migrabitumen”). Int. J. Coal Geol. 11, 65e79.

Jarvie, D.M., Hill, R.J., Pollastro, R.M., 2005. Assessment of the gas potential andyields from shales: the Barnett shale model. Unconv. Energy Resour. South.Midcont. 37e50.

Jarvie, D.M., Hill, R.J., Ruble, T.E., Pollastro, R.M., 2007. Unconventional shale-gassystems: the Mississippian Barnett Shale of north-central Texas as one modelfor thermogenic shale-gas assessment. AAPG Bull. 91, 475e499.

Jiang, L., Cai, C., Worden, R.H., Li, K., Xiang, L., Chu, X., Shen, A., Li, W., 2015. Rareearth element and yttrium (REY) geochemistry in carbonate reservoirs duringdeep burial diagenesis: implications for REY mobility during thermochemicalsulfate reduction. Chem. Geol. 415, 87e101.

Ji, W., Song, Y., Jiang, Z., Wang, X., Bai, Y., Xing, J., 2014. Geological controls andestimation algorithms of lacustrine shale gas adsorption capacity: a case studyof the Triassic strata in the southeastern Ordos Basin, China. Int. J. Coal Geol.134e135, 61e73.

Jungmann, M., Kopal, M., Clauser, C., Berlage, T., 2011. Multi-class supervised clas-sification of electrical borehole wall images using texture features. Comput.Geosci. 37, 541e553.

Kaufmann, J., Loser, R., Leemann, A., 2009. Analysis of cement-bonded materials bymulti-cycle mercury intrusion and nitrogen sorption. J. Colloid Interface Sci. 336(2), 730e737.

Labani, M.M., Rezaee, R., Saeedi, A., Hinai, A.A., 2013. Evaluation of pore sizespectrum of gas shale reservoirs using low pressure nitrogen adsorption, gasexpansion and mercury porosimetry: a case study from the Perth and CanningBasins, Western Australia. J. Pet. Sci. Eng. 112, 7e16.

Lazar, O.R., Bohacs, K.M., Macquaker, J.H.S., Schieber, J., Demko, T.M., 2015.Capturing key attributes of fine-grained sedimentary rocks in outcrops, cores,and thin sections: nomenclature and description guidelines. J. Sediment. Res. 85(3), 230e246.

Liang, C., Jiang, Z., Zhang, C., Guo, L., Yang, Y., Li, J., 2014. The shale characteristicsand shale gas exploration prospects of the Lower Silurian Longmaxi shale,Sichuan Basin, South China. J. Nat. Gas Sci. Eng. 21, 636e648.

Loucks, R.G., Reed, R.M., Ruppel, S.C., Hammes, U., 2012. Spectrum of pore types andnetworks in mudrocks and a descriptive classification for matrix-relatedmudrock pores. AAPG Bull. 96, 1071e1098.

Loucks, R.G., Reed, R.M., Ruppel, S.C., Jarvie, D.M., 2009. Morphology, genesis, anddistribution of nanometer-scale pores in siliceous mudstones of the Missis-sippian Barnett shale. J. Sediment. Res. 79, 848e861.

Loucks, R.G., Ruppel, S.C., 2007. Mississippian Barnett shale: lithofacies and depo-sitional setting of a deep-water shale-gas succession in the Fort Worth basin,Texas. AAPG Bull. 91, 579e601.

Lu, J., Ruppel, S.C., Rowe, H.D., 2015. Organic matter pores and oil generation in the

Tuscaloosa marine shale. Am. Assoc. Pet. Geol. Bull. 99, 333e357.Mastalerz, M., Schimmelmann, A., Drobniak, A., Chen, Y., 2013. Porosity of Devonian

and Mississippian New Albany Shale across a maturation gradient: insightsfrom organic petrology, gas adsorption, and mercury intrusion. AAPG Bull. 9,1621e1643.

Milliken, K.L., Reed, R.M., 2010. Multiple causes of diagenetic fabric anisotropy inweakly consolidated mud, Nankai accretionary prism, IODP Expedition 316.J. Struct. Geol. 32 (12), 1887e1898.

Nelson, P.H., 2009. Pore-throat sizes in sandstones, tight sandstones, and shales.AAPG Bull. 93, 329e340.

Pepper, A.S., 1991. Estimating the petroleum expulsion behaviour of source rocks: anovel quantitative approach. Geol. Soc. 59 (1), 9e31.

Reed, M.R., Loucks, R.G., Ruppel, S.C., 2014. Comment on “Formation of nanoporouspyrobitumen residues during maturation of the Barnett Shale (Fort Worth Ba-sin)” by Bernard et al. (2012). Int. J. Coal Geol. 127, 111e113.

Ross, D.J.K., Marc Bustin, R., 2007. Impact of mass balance calculations on adsorptioncapacities in microporous shale gas reservoirs. Fuel 86 (17e18), 2696e2706.

Ross, D.J.K., Marc Bustin, R., 2009. The importance of shale composition and porestructure upon gas storage potential of shale gas reservoirs. Mar. Pet. Geol. 26,916e927.

Rouquerolt, J., Avnir, D., Fairbridge, C.W., Everett, D.H., Haynes, J.H., Pernicone, N.,Ramsay, J.D.F., Sing, K.S.W., Unger, K.K., 1994. Recommendations for the char-acterization of porous solids. Pure Appl. Chem. 66, 1739e1758.

Saggaf, M.M., Nebrija, E.L., 2003. A fuzzy logic approach for the estimation of faciesfrom wire-line logs. AAPG Bull. 87, 1223e1240.

Sahay, V.K., Mude, S.N., Samant, B., 2015. Lithofacies and depositional environmentof the ferruginous claystone member sediments of Naredi, Kutch, Gujarat, India.Geology 5, 268e270.

Schoenherr, J., Littke, R., Urai, J.L., Kukla, P. a., Rawahi, Z., 2007. Polyphase thermalevolution in the Infra-Cambrian Ara Group (South Oman Salt Basin) as deducedby maturity of solid reservoir bitumen. Org. Geochem. 38, 1293e1318.

Slatt, R.M., O'Brien, N.R., 2011. Pore types in the Barnett and Woodford gas shales:contribution to understanding gas storage and migration pathways in fine-grained rocks. AAPG Bull. 95, 2017e2030.

Sing, K.S., 1985. Reporting physisorption data for gas/solid systems with specialreference to the determination of surface area and porosity. Pure Appl. Chem.57, 603e619.

Soua, M., 2014. Paleozoic oil/gas shale reservoirs in southern Tunisia: an overview.J. Afr. Earth Sci. 100, 450e492.

Stow, D.A.V., Huc, A.-Y., Bertrand, P., 2001. Depositional processes of black shales indeep water. Mar. Petroleum Geol. 18, 491e498.

Teichert, C., 1958. Concepts of facies. AAPG Bull. 42, 2718e2744.Valenza, J., Drenzek, N., Marques, F., Mastalerz, M., 2013. Geochemical controls on

shale microstructure. Geology 41, 611e614.Velde, B., 1996. Compaction trends of clay-rich deep sea sediments. Mar. Geol. 133,

193e201.Wang, G., Carr, T.R., 2012a. Marcellus shale lithofacies prediction by multiclass

neural network classification in the Appalachian basin. Math. Geosci. 44,975e1004.

Wang, G., Carr, T.R., 2012b. Methodology of organic-rich shale lithofacies identifi-cation and prediction: a case study from Marcellus Shale in the Appalachianbasin. Comput. Geosci. 49, 151e163.

Wang, G., Carr, T.R., Ju, Y., Li, C., 2014a. Identifying organic-rich Marcellus Shalelithofacies by support vector machine classifier in the Appalachian basin.Comput. Geosci. 64, 52e60.

Wang, Y., Dong, D., Yang, H., He, L., Wang, S., Huang, J., Pu, B., Wang, S., 2013.Quantitative characterization of reservoir space in the lower Silurian Longmaxishale, southern Sichuan, China. Sci. China Earth Sci. 57 (2), 313e322.

Wang, Y., Zhu, Y., Chen, S., Li, W., 2014b. Characteristics of the nanoscale porestructure in northwestern Hunan shale gas reservoirs using field emissionscanning electron microscopy, high-pressure mercury intrusion, and gasadsorption. Energy & Fuels 28 (2), 945e955.

Yang, F., Ning, Z., Liu, H., 2014. Fractal characteristics of shales from a shale gasreservoir in the Sichuan Basin, China. Fuel 115, 378e384.

Zhang, B., Zhao, T., Jin, X., Marfurt, K.J., 2015. Brittleness evaluation of resource playsby integrating petrophysical and seismic data analysis. Interpretation 3 (2),T81eT92.

Zhou, Q., Xiao, X., Tian, H., Pan, L., 2014. Modeling free gas content of the lowerPaleozoic shales in the Weiyuan area of the Sichuan Basin, China. Mar. Petr.Geol. 56, 87e96.