9

Raphael A. J. Wüst, Brent R. Nassichuk, R. Marc Bustin, 2013, Porosity characterization

of various organic-rich shales from the Western Canadian Sedimentary Basin,
Alberta and British Columbia, Canada, in W. Camp, E. Diaz, and B. Wawak, eds.,
Electron microscopy of shale hydrocarbon reservoirs: AAPG Memoir 102, p. 81–100.

Porosity Characterization of Various

Organic-rich Shales from the Western
Canadian Sedimentary Basin, Alberta
and British Columbia, Canada
Raphael A. J. Wüst
Trican Geological Solutions Ltd., 621-37 NE Calgary, Canada T2E 2MI and James Cook University,
4811 Townsville, Australia (e-mail: rwust@trican.ca)

Brent R. Nassichuk
Trican Geological Solutions Ltd., 621-37 NE Calgary, Canada T2E 2MI (e-mail: bnassichuk@trican.ca)

R. Marc Bustin
University of British Columbia, 2329 W. Mall, Vancouver, British Columbia V6T 1Z4, Canada
(e-mail: bustin@mail.ubc.ca)

ABSTRACT
Scanning electron microscopy (SEM) of shales from three unconventional gas/liquid plays
­(Nordegg, Montney, Duvernay) of the Western Canadian Sedimentary Basin were combined with
routine analytical investigations (x-ray diffraction, source rock analyses, mercury ­porosimetry, and
petrography) to characterize mineral composition, mineral assemblage, morphology, organic con-
tent, and porosity. The investigations demonstrated that despite marked diagenetic differences be-
tween these shales, some common textural and pore characteristics occurred in all samples. The
study showed that SEM morphological investigations of unconventional shale reservoirs provided
important information about mineral aggregates, cementation, and clay mineral distribution, which
allows interpretations about diagenetic ­history. Combining petrographic analyses with SEM is criti-
cal in sediments with pronounced cementation and mineral overgrowth. Conventional secondary
electron SEM studies of untreated samples have several advantages over focused ion beam (FIB)
milled surfaces as morphological characteristics of larger areas can be evaluated. The study showed
that the organic material may not be fully identified, but organic fragments and some bituminous
material can still be recognized. In general, abundant large pores (200 to . 2000 nm) can be observed
under SEM, but other techniques, such as the mercury porosimetry, failed to identify these larger
pores in these shales. The main morphological porosity types in these shales were intergranular (be-
tween particles), dissolution, residual growth, fracture, and phyllosilicate pores. The high number
of 5- to 20-nm pores as determined by mercury porosimetry could not be confirmed by SEM, which
may have been because of resolution issues. Hence, future porosity characterization of gas shales
needs to investigate these problems of discrepancy between analytical and imaging technologies.

Copyright © 2013 by The American Association of Petroleum Geologists.

DOI: 10.1306/13391707M1023585

81

13835_ch09_ptg01_hr_p081-100.indd 81 22/05/13 10:38 AM

 82  Wüst et al

INTRODUCTION The focus of this research was to characterize both

the intrinsic and fracture-related pores of these rocks.
Characterizing textural parameters of shale hydrocar- Previous gas shale studies in the United States have
bon reservoirs, including porosity and permeability, is characterized four different pore types, which range
critical to understand the potential for flow and fluid from nano- to micrometer sizes (Loucks et al., 2009;
transport during production. Pore space influences the Milner et al., 2010; Schieber, 2010), although these ig-
amount of hydrocarbons stored within the rocks; per- nored the large micro- to submillimeter-size pores
meability governs the flow through the reservoir. In within some of the shale reservoirs. The four types
tight reservoir rocks, including shales and mudstones, previously defined are (1) phyllosilicate framework
matrix permeability is often small (in the range of 1023 (voids between clays and grain aggregates or ce-
to 1027 md), but artificial stimulation and hydraulic ments); (2) organic matter (pores associated with
fracturing increase fracture density and thus reduce organic matter, mostly kerogen and bitumen); (3) car-
the flow length through the tight rock to the fracture. bonate dissolution (pores resulting from diagenetic al-
In tight-rock resource plays, porosity characterization teration of certain grains, mainly carbonate minerals);
through empirical and visual means is instrumental and (4) intraparticle pores. According to Milner et al.
for any exploration strategy. (2010), the intraparticle porosity forms networks be-
Shale and mudstones are fine-grain (particle size tween nanofossil fragments coated with amorphous
,63 mm) sedimentary rocks that can contain marked kerogen and occurs primarily in fecal pellets.
amounts of organic matter (up to 30 wt.%) (Riediger
et al., 1990). Rocks with high organic content often rep-
resent ideal source rocks and have become common Geologic Settings
targets in current exploration trends. Unconventional
coal and shale gas has reshaped the energy future of An overview of the formations in western Alberta/
our planet, and reserves for at least the next 135 years eastern British Columbia and sample depth with geo-
are suggested (Conti et al., 2011). In ­Alberta alone, gov- physical logs is provided in Figure 2. Samples were se-
ernment estimates suggest that at least 850 tcf of shale lected from interpreted organic-rich intervals based on
gas may be present. Shale plays in Western Canada (see geophysical log characteristics (high gamma ray and
Figure 1) have become valuable and important sources resistivity; Figure 2).
for gas, condensate, and oil.
This study investigated three organic-rich shale
successions of the Western Canadian Sedimentary Nordegg
Basin (Figure 1) that have markedly different com-
positional characteristics. A total of six samples (two The Lower Jurassic Nordegg Member is composed of
of each formation/unit) have been investigated from 25–45 m (78.7–147.6 ft) of organic-rich marine sedi-
the Devonian Duvernay Formation, the Lower Triassic ments (quartz and/or carbonate rich) that were de-
Montney Formation, and the Lower Jurassic Nordegg posited in a restricted basinal environment adjacent to
Member of the Fernie Formation. reefs and carbonate platforms (Riediger et al., 1990).
The Lower Triassic Montney Formation in ­British The Nordegg Member often consists of three lithologi-
Columbia and Alberta is a thick (.200–300 m [656.1– cal facies: (1) a lower phosphatic organic-rich shale; (2)
984.2 ft] in areas), fine-grain clastic deposit that has a silty or carbonate-rich mudstone; and (3) an upper
been classified as a gas shale play, although the forma- phosphatic, organic-rich shale. In areas of low thermal
tion consists of tight sandstone and siltstone interbed- maturity, total organic content values of the Nordegg
ded with shale. Initial focus on the Montney Formation are up to 23 wt.% (Riediger and Bloch, 1995). The high
centered on the conventional targets; however, recent total organic carbon (TOC) contents reflect high algal
research and developments in horizontal drilling and productivity with reduced bottom-water mixing. Nor-
hydraulic fracture stimulation have allowed for the ex- degg adsorbed gas capacities range from 0.05 cm 3/g
ploitation of the unconventional source rock horizons. (1.6 scf/t) to over 2 cm3/g (64.1 scf/t) in organic-rich
On the other hand, some other thin (,40 m [131.2 ft]) zones (Ross and Bustin, 2007).
shale plays, such as the Devonian Duvernay Formation
and the Lower Jurassic Nordegg Member of the Fernie
Formation, have only recently attracted attention be- Montney
cause of their potential for wet gas and other hydrocar-
bon liquids as gas prices have plummeted to new lows The Montney Formation in northeastern British
over the last 10 years. ­C olumbia and adjacent parts of Alberta consists

13835_ch09_ptg01_hr_p081-100.indd 82 22/05/13 10:38 AM

 Porosity Characterization of Various Organic-rich Shales from the Western Canadian Sedimentary Basin, Alberta and British Columbia, Canada  83

Figure 1. Overview map of the Western Canadian Basin showing the deep shale extent of the Devonian Duvernay Formation
(stippled line) that is bounded in the east by the deformation front (black line) (after Edwards et al., 1994). Location of the
two wells from the Nordegg Member, one situated in the oil window (Upper Nordegg 6-9-66-24W5) and the other one
from the dry-gas window (Lower Nordegg 16-23-57-6W6), of the well of the Montney Formation (15-34-80-18W6; Upper
Montney) and the well of the Duvernay Formation (1-23-49-25W4) were investigated.

mostly of relatively deep-water, mid- to distal-shelf, in a variety of settings under arid, midlatitudinal
and slope shales, siltstones, and rare limestones, conditions. The sediments are dominantly siltstones
some of which are tempestites and/or turbidites that contain abundant immature feldspars and few
­( Edwards et al., 1994). The deposition occurred on clay minerals shed from a low-gradient continental
the western margin of the North American craton interior.

13835_ch09_ptg01_hr_p081-100.indd 83 22/05/13 10:38 AM

 84  Wüst et al

Duvernay protected Leduc embayments. Basinward from the

reefs, Duvernay basin-fill generally thins markedly.
The Duvernay interval is a unique depositional unit
within the Woodbend Group (Stoakes and Creany,
1985; Fowler and ­S tasiuk, 2001). The Duvernay METHODS
­Formation is characterized by extensive basinal depos-
its and synchronous with the middle stage of Leduc Samples of the Lower Triassic Montney Formation
reef growth (Figure 2). The Duvernay Formation were collected from the British Columbia core reposi-
­consists of dark brown bituminous shale with up to tory at Charlie Lake, British Columbia; samples from
17 wt.% TOC (Riediger et al., 1990; Stasiuk and Fowler, the Devonian Duvernay Formation and the Nordegg
2004) and limestone. The high organic carbon content Member were collected from the Energy Resources
reflects a major change in the stratification and oxy- Conservation Board in Calgary, Alberta. Rock sam-
genation of basinal waters during the maximum trans- ples were cut and prepared for petrographic mi-
gressive stage of the Woodbend. Adjacent to the Leduc croscopy and scanning electron microscopic (SEM)
reef complexes, the Duvernay is thick and intermixed analyses. Subsamples were also analyzed for miner-
with reef-derived detritus and occurs especially within alogy, ­thermal maturity, and TOC as well as mercury

Figure 2. Geophysical logs and sample depths (dots) of the three formations/member (four wells) investigated in this study.
A schematic overview of the formations in the Western Canadian Sedimentary Basin of eastern central Alberta and western
British Columbia is also given.

13835_ch09_ptg01_hr_p081-100.indd 84 22/05/13 10:38 AM

 Porosity Characterization of Various Organic-rich Shales from the Western Canadian Sedimentary Basin, Alberta and British Columbia, Canada  85

porosimetry (mercury injection capillary pressure, An Olympus BX51 research microscope with a DP72
MICP) analyses. Digital Camera System and an EXFO (x-Cite ­Series 120Q)
Geochemical analyses were conducted using a fluorescence illuminator was used for the petrographic
Source Rock Analyzer™ (SRA), a similar instrument to analysis of the samples. Photomicrographs were taken
Rock-Eval, to determine the TOC and degree of organic at 340, 3100, 3200, and 3400. All thin sections were
thermal maturity based on Tmax values. In the Western prepared using an ultraviolet (UV)-­fluorescent epoxy
Canadian Sedimentary Basin, the Tmax values of rocks (Rhodamine B) and polished to 30-mm thickness. Half of
are commonly used to infer thermal maturity because the thin section was then stained with a dual-carbonate
it is a rapid and inexpensive technique, and many stain (Alizarin Red S to distinguish between calcite and
organic-rich marine shales lack true vitrinite macerals dolomite; potassium ferricyanide to distinguish carbon-
required for vitrinite reflectance measurements. The ate minerals containing iron).
TOC content in shale reflects the potential for sorbed A JEOL JSM-6610 SEM was used to determine sur-
gas capacity. The SRA is equipped with a flame ioniza- face characteristics down to nanometer scale of all
tion detector and two infrared detectors. X-ray diffrac- samples. Two to four subsamples of each core sample
tion (XRD) analyses were carried out to determine and were picked and mounted on stubs and gold coated.
quantify the mineralogical composition. Micrometer-scale elemental analysis was completed
A Bruker D4 Endeavour XRD instrument with a using an Oxford Instruments INCAx-act EDS (energy
LynxEye Detector was used with a cobalt x-ray tube. dispersive spectroscopy) system containing a silicon
The samples were analyzed between 4 and 70° 2theta drift detector (SDD). Most images were taken under
at 40 kV and 30 mA. The XRD trace data were ana- high-vacuum mode using SE (secondary electron) at
lyzed using Rietveld methods (Jade 9) to quantify the 25 kV, spot size 25–40, and working distance between
mineralogy. Data quality assurance was achieved 11 and 13 mm.
­using internal standards (shale mixtures).
A Micromeritics AutoPore IV Mercury Porosimeter
9500 was used to measure mercury porosimetry by RESULTS
MICP, up to 60,000 psi. Mercury porosimetry routinely
is applied over a capillary diameter range from 0.003 Table 1 shows the following information about each
to 360 mm. formation: depth, mineralogical composition, mercury

Table 1. Well Identifier, Formation/Unit Names, General Maturity Level, Sample Depths, Mineralogical Composition (XRD),
Skeletal Density, Average Pore Diameter, Total Porosity, Average Permeability (All Data from Mercury Porosimetry), Thermal
Maturity, S1–S3, and TOC Values (SRA) of the Six Samples Analyzed Quartz (Qtz), albite (Ab, microcline (Mc), orthoclase (Or),
calcite (Cal), dolomite (Dol), ankerite (Ank), Illite/mica (Ill/Mca), chlorite (Chl), kaolinite (Kln), pyrite (Py), apatite (Ap).

13835_ch09_ptg01_hr_p081-100.indd 85 22/05/13 10:38 AM

 86  Wüst et al

Figure 3. Hand specimens of the six samples analyzed for this study. Top orientation toward top of page. Scale in centimeters.
See Figure 2 for formation details.

porosimetry (porosity and permeability), thermal ma- (lower Nordegg 16-23-57-6W6) (Table 1). Both samples
turity, and TOC analyses. Core photographs are shown have 50–60% carbonates and about 30–35% quartz with
in Figure 3. The data highlights the difference between small amounts of clays, pyrite, apatite, and feldspars.
the three formations. Duvernay samples ­often con- The samples contain abundant radiolarian and microfos-
tain slightly higher clay contents than the other for- sil debris (Appendix A9; ­Figures 4, 5), and TOC is high
mations; the Montney samples often have elevated (5.5–7.2 wt.%). The sample of the lower Nordegg con-
feldspar and dolomite contents. Petrographic descrip- tains some layers with abundant shell material, and two
tions and thin section photomicrographs are available XRD and SRA analyses were performed (Table 1) to cap-
in the Appendix of this chapter in the accompanying ture and determine the variability in carbonate content.
CD-ROM of this volume. Mercury porosimetry determined that mean pore diam-
eters were 19 nm (upper Nordegg) and 15 nm (lower
Nordegg); ­total porosities were 6% (upper) and 1.8%
Lower Jurassic Nordegg Member (Fernie Formation) (lower). The upper Nordegg Member sample (2026) had
a TOC of 5.5 wt.% and Tmax of about 434°C (mature),
Two wells from the Nordegg Member were selected; and the lower Nordegg sample (7838) had a TOC of 6.5–
one is situated in the oil window (upper Nordegg 6-9- 7.2 wt.% and Tmax of 595°C (overmature) (located in close
66-24W5), and the other one is from the dry-gas window proximity of the deformed belt line; Figure 1).

13835_ch09_ptg01_hr_p081-100.indd 86 22/05/13 10:39 AM

 Porosity Characterization of Various Organic-rich Shales from the Western Canadian Sedimentary Basin, Alberta and British Columbia, Canada  87

Figure 4. Scanning electron microscopic (SEM) photomicrograph of the Upper Nordegg (06-09-066-24W5; sample 2026.24 m
[6647.7690 ft]) showing an overview (A) of the fine-grain nature of this shale with some idiomorphic dolomite (B), (C) and
mineralized (Si-rich) plant fragments (D). All arrows refer to various pores (dissolution/residual growth, black; intergranular/
fracture, white; organic matter, white with round tip).

Scanning electron microscopic images showed caused by greater burial depth (current depth 363 m
that these rocks were very fine-grain (,15 mm) sedi- [1190.9 ft]) and tectonic stress near the deformed belt.
ments with high organic matter content (Figures 4–7). Clay minerals (and possibly organic matter) showed
The organics not only coat mineral matter but also draping features around grains, indicating differential
infill intercrystalline pore space (bituminous resi- compaction.
due). In addition to radiolaria, sample 2026 showed
other abundant microfossils (coccolith fragments; see
Appendix 1). These aggregates of microfossils formed Lower Triassic Montney Formation
a complex framework of abundant macropores. Neo-
formed diagenetic minerals were also abundant, The Montney Formation was investigated in well
including dolomite and quartz, and some of the car- 15-34-80-18W6 (upper Montney) at depths of 2073 and
bonates had dissolution pores. Sample 7838 showed a 2200 m (6801 and 7218 ft) (Table 1). The samples had
pronounced alignment of minerals and clays along the about 38–43% quartz, 15–17% feldspars, 20–28% car-
bedding planes, with associated fractures (Figures 6, bonates (about 10% calcite), 8–17% clays, and 2–8%
7), which is interpreted as a result of higher pressure pyrite. The samples were dominated by fine-sand to

13835_ch09_ptg01_hr_p081-100.indd 87 22/05/13 10:39 AM

 88  Wüst et al

Figure 5. Scanning electron microscopic (SEM) photomicrographs of the Upper Nordegg (06-09-066-24W5; sample 2026.24 m
[6647.7690 ft]) showing details of the mineralized plant fragment with some large pores (A), abundant nannofossils (B), (D)
and bituminous material coating some of the mineral matter (TOC was 5.5 wt.%).

silt-size detritus (feldspars, quartz, mica; see Appendix Primary minerals often showed clay mineral coatings,
A9-2 and A9-3) with carbonate cements and mod- some of which had grown during diagenesis as they
erate organic material (TOC 1.5–3 wt.%). Mercury- consist of fibrous structures. Dissolution, abandoned
porosimetry-determined average pore diameters were growth (interpreted as caused by pore-fluid changes
22 nm (2073) and 11 nm (2200); total porosities were or lattice poisoning; see Figures 9), and phyllosilicate
4% (2073) and 1.5% (2200). Sample 2073 had a TOC pores were commonly observed. The deeper sample
of 1.5 wt.% and Tmax of about 432°C (mature); sample (2200) contained more fracture pores than the upper
2200 had a TOC of 3 wt.% and Tmax of 467°C (over- sample and showed quartz overgrowth and clay coat-
mature). The lower porosity and pore sizes in sample ings (Figures 10, 11).
2200 appeared to be related to greater depth of maxi-
mum burial, as reflected by the Tmax measurements.
Scanning electron microscopic analysis of the upper Upper Devonian Duvernay Formation
sample (2073) showed the presence of abundant coarse
angular minerals (mainly secondary carbonates), with The Duvernay Formation was investigated in well
some quartz, feldspars, and mica sheets (Figures 8, 9). 1-23-49-25W4 at depths of 5886 ft (1794 m) and

13835_ch09_ptg01_hr_p081-100.indd 88 22/05/13 10:39 AM

 Porosity Characterization of Various Organic-rich Shales from the Western Canadian Sedimentary Basin, Alberta and British Columbia, Canada  89

Figure 6. Scanning electron microscopic (SEM) photomicrographs of the Lower Nordegg (16-23-057-06W6; sample 7838 ft
[2389.0224 m]) showing an overview of the laminated sample (A), details of the bedding-parallel fractures (B), and clay
fabric deformed during burial diagenesis (C). This well has been subject to higher-stress regimes as it is situated in close
proximity to the deformed belt. Higher levels of diagenesis may have led to recrystallization of the delicate nannofossils as
only possible fragments were observed in this sample (D).

5912 ft (1802 m) (Table 1). The samples had about cements forming smooth surfaces; other grains were
18–36% quartz, about 10% feldspars, 36–39% carbon- often coated with clay minerals (Figures 12–15).
ates (about 33–37% calcite), 14–21% clays, and pyrite ­Secondary cements had replaced and infilled micro-
(3–12%). They were dominated by dark, organic-rich fossils and pore space along the bedding planes. Some
shales ­( Appendix A9-4) with TOC up to 13 wt.% pores were the result of interrupted growth (residual
(Riediger et al., 1990; Stasiuk and Fowler, 2004) and growth pores), which was likely caused by chemi-
lime mudstones. Mercury-porosimetry-determined cal changes of the fluid composition during growth
average pore diameters were 12 nm (5886) and 11 nm of the minerals (Figures 12, 13, 15). Both dissolution
(5912); total porosities were 1.1% (5886) and 1.3% and abandoned growth pores are common and are up
(5912). Sample 5886 had a TOC of 9.1 wt.% and Tmax to 2 mm in diameter. In general, these rocks appeared
of about 434°C ­(mature); sample 5912 had a TOC of fairly tight with few interconnected pores. Pore spaces
10.3 wt.% and Tmax of 438°C (mature). were typically elongated and were up to 15–20 mm in
Scanning electron microscopic images showed length and 1–2 mm wide. The deeper sample contained
­recrystallization with some quartz and carbonate higher clay mineral contents, and phyllosilicate pore

13835_ch09_ptg01_hr_p081-100.indd 89 22/05/13 10:39 AM

 90  Wüst et al

Figure 7. Scanning electron microscopic (SEM) photomicrographs of the Lower Nordegg (16-23-057-06W6; sample 7838 ft
[2389.0224 m]) showing an overview of the sparry calcite-filled radiolarian (A), various fractures (B), (D), and the fine-grain
nature of the sediment (see quartz grain embedded in clay-organic matter).

space was common. The clays also showed differential and the surfaces were coated with gold. This sample
compaction around competent grains (Figure 15). preparation technique had the advantage over argon-
ion and focused ion beam (FIB) milling in that it pre-
served the rock fabric and allowed for investigations
DISCUSSION of a larger (more representative) sample area. Ion-
milled samples appeared to have an advantage, how-
The focus of this study was to investigate grain mor- ever, in imaging nanometer-scale pores in organic
phology and porosity of three different shale hy- matter (e.g., Ambrose et al., 2010; Milner et al., 2010),
drocarbon reservoirs from the Western Canadian and FIB-SEM studies provided valuable 3-D images
Sedimentary Basin. The formations investigated were to evaluate porosity and permeability distributions
the Lower Jurassic Nordegg Member (Fernie Forma- (Keller et al., 2011). However, care needs to be taken
tion), the Lower Triassic Montney Formation, and the with ion-milled samples as energy level and vacuum
Upper Devonian Duvernay Formation. Scanning elec- conditions can easily introduce new pore space in
tron microscopy was complemented by petrographic, organic material because of desiccation and shrink-
XRD, SRA, and mercury porosimetry. The SEM sam- age. In addition, sputter coating may also introduce
ples were prepared from fresh broken rock material, artificial porosity. Here, we focused on analyzing

13835_ch09_ptg01_hr_p081-100.indd 90 22/05/13 10:39 AM

 Porosity Characterization of Various Organic-rich Shales from the Western Canadian Sedimentary Basin, Alberta and British Columbia, Canada  91

Figure 8. Scanning electron microscopic (SEM) photomicrographs of the Upper Montney (15-34-080-18W6; 2073.65 m)
showing an overview of the sample (A), embedded mica sheets (B), and abundant carbonate (mainly dolomite) showing
overgrowth features on sample surfaces (C), (D). Note also some of the filamentous clays on the grain surfaces with high
pore space.

fresh broken surfaces to gather information about The diagenetic cements were confirmed using petro-
mineral and microfossil topography, relief and grain graphic analysis. Radiolaria in the Nordegg samples
shapes, organics, pore shapes and characteristics, as that had their original shell replaced by calcite were
well as mineral growth, overgrowth, and clay coating common together with abundant microfossils (coc-
features. coliths) ­(Figure 5; Figure A1), but these could not be
recognized in the deeper sample, likely because of re-
crystallization (Figures 6, 7). Mineral grains from the
Mineral Aggregates and Special Features Nordegg samples were often less than 10 mm with
only diagenetic minerals (carbonate, quartz, pyrite)
The SEM images of the three different formations forming larger crystals. Commonly, clay minerals
showed characteristic microstructures. Both the mask all other detrital mineral matter. Biogenic silica
­N ordegg Member and Duvernay Formation sam- and calcite clearly contributed much of the material
ples contained significant amounts of biogenic silica in these rocks (Figures 4–7), and authigenic mineral
and calcite with little detrital material. Overgrowth matter was a common constituent in these shales
cements were recognized by their smooth surfaces. ­( Figures 12–14). The Montney shale samples were

13835_ch09_ptg01_hr_p081-100.indd 91 22/05/13 10:39 AM

 92  Wüst et al

Figure 9. Scanning electron microscopic (SEM) photomicrographs of the Upper Montney (15-34-080-18W6; 2073.65 m
[6803.3137 ft]) showing details of filamentous clay minerals (A), (B), (D), which appear to have had unrestricted space to
grow. Large intergranular pores (arrows) and dissolution/residual growth pores (C) can also be observed.

markedly different and were dominated by detri- porosity characterizations (e.g., Curtis et al., 2010;
tal minerals (quartz, feldspars, mica) (Figures 8–11; ­Schieber, 2010; Sondergeld et al., 2010; Curtis et al.,
Figures A2, A3) and early diagenetic carbonates 2011). Methods included electron imaging (transmis-
(dolomite/calcite). Most of these detrital grains and sion electron microscopy [TEM], SEM, FIB-SEM),
early diagenetic carbonates showed late diagenetic ­o ptical petrography, and analytical methods (mer-
overgrowth features and mineral grains composed cury porosimetry, helium porosimetry, gas adsorption,
of quartz, carbonates, or clay minerals. The cements etc.). The most common pore types observed were in-
within the sandy siltstones of the Montney Formation terparticle, dissolution/growth, and fracture porosity.
(TOC 1.5–3 wt. %) had occluded much of the origi- Although rare, phyllosilicate framework pores, similar
nal porosity ­(Figures 8, 12; Figures A2, A3), ­resulting to those described by Schieber (2010), were observed
in the low-permeability nature of the Montney in all six samples of this study (e.g., Figures 6, 9–11,
Formation. 15) and were interpreted as the result of diagenetic
clay formation and transformation (Scotchman, 1989;
Burley and MacQuaker, 1992; Michalopoulos and
Pore Space and Pore Shapes Aller, 2004). Porous kerogen was not observed (except
for the mineralized plant fragment in Figures 4, 5)
The advent of shale as an unconventional reservoir despite the high TOC contents, but this may have been
play has resulted in numerous studies focusing on caused by the sample preparation technique.

13835_ch09_ptg01_hr_p081-100.indd 92 22/05/13 10:39 AM

 Porosity Characterization of Various Organic-rich Shales from the Western Canadian Sedimentary Basin, Alberta and British Columbia, Canada  93

Figure 10. Scanning electron microscopic (SEM) photomicrographs of the Upper Montney (15-34-080-18W6; 2200.05 m
[7218 ft]) showing an overview of this slightly finer-grain sample (A). Note that the surface was covered by secondary ce-
ments (overgrowth; see petrographic images) and some clay coatings, which make grain identification difficult. Large pores
(.1–2 mm) were common, and some short (~5–60 mm) thin fracture porosity was also observed (B)–(D). Clay coatings
(thin platelets) were common (D).

The pore sizes observed in SEM images ranged pores were common (e.g., Figure 16) and yet not
between 200 and 2000 nm, but mercury porosimetry recognized by mercury porosimetry. During mer-
data showed only a few pore sizes in that range (Fig- cury injection, only connected pores were measured;
ure 16). The most abundant pore sizes measured by thus, it may be possible that some of the larger pores
mercury porosimetry ranged between 4 and 100 nm. observed in SEM might not have been connected to
The pore size distribution difference between SEM be accessed by mercury during the measurements.
image analysis and mercury porosimetry is a com- In addition, the principle of mercury porosimetry is
mon problem that has been observed in many other based on the assumption of cylindrical pores (Webb
studies (Curtis et al., 2010; Sondergeld et al., 2010). It and Orr, 1997); thus, larger pore spaces with narrow
was suggested that the larger pores in conventional pore throats may result in incorrect pore size distri-
SEM analysis may be caused by grain plucking dur- bution data. In addition, the pore shapes observed
ing sample preparation (Sondergeld et al., 2010). The in SEM images were not cylindrical and were often
morphological analyses of this study indicated, how- long, narrow pores with limited aperture. With im-
ever, that grain plucking was relatively rare and could aging techniques, pores are typically measured by
easily be identified and excluded in pore size mea- the longest dimension, whereas mercury porosim-
surements. Abundant large (.500 nm) dissolution etry measures the entry (the ­s mallest) dimension.

13835_ch09_ptg01_hr_p081-100.indd 93 22/05/13 10:39 AM

 94  Wüst et al

Figure 11. Scanning electron microscopic (SEM) photomicrographs of the Upper Montney (15-34-080-18W6; 2200.05 m
[7218.0118 ft]) showing details of pores (arrows) between different grains (A), dissolution (A), residual growth (B), and
­fractures porosity (B), (C), framboidal pyrite embedded in carbonate cements and some clays (C), and quartz grain with quartz
overgrowth in cross section (D). Note that some of the carbonates in (A) and (B) appear to have thin clay sheet inclusions.

This  phenomenon requires further attention even dissolution of carbonates will lead to chemical buff-
though our average pore diameter data for this study ering, and carbonate cements may form. For disso-
determined by the mercury method ­(Table 1) were in lution of silicates such as biogenic silica, quartz, or
line with other shale studies from the United States feldspars, pH levels are often thought to be alkaline
(e.g., Devonian shales in the Appalachian basin) using (.8.5) or extremely acidic (Alexander et al. 1954;
similar techniques (Nelson, 2009). Lewin, 1961), conditions that can be achieved dur-
Dissolution and/or residual growth (i.e., inter- ing burial diagenesis. Extreme acid fluid conditions
rupted crystal growth) pore space in both carbonate would dissolve most of the carbonates, which then
and quartz grains imply changes in pore water chem- would lead to a buffering of the pore fluid system.
istry. Pore water chemistry changes during burial In sedimentary deposits, Schott and Oelkers (1995)
diagenesis of organic-rich sediments commonly oc- observed that the hydrolysis of mixed oxide silicates
cur because of decarboxylation of organic matter is a multistep process that involves breaking of sev-
(kerogen) (Hunt, 1996) and the formation of organic eral bonds and structure/­chemistry of the mineral’s
acids (including carboxylic). The presence of organic near-surface region. Thus, several minerals could con-
acids could contribute to carbonate dissolution (e.g., tribute to aqueous silica from the conversion of iso-
­F igures 9, 11, 15). In low-permeability mudstones, lated silica units through the hydrolysis of modifier

13835_ch09_ptg01_hr_p081-100.indd 94 22/05/13 10:39 AM

 Porosity Characterization of Various Organic-rich Shales from the Western Canadian Sedimentary Basin, Alberta and British Columbia, Canada  95

Figure 12. Scanning electron microscopic (SEM) photomicrographs of the Duvernay (1-23-049-25W4; 5886 ft [1794 m])
showing its fine-grain nature (A). Various pores (arrows) (A)–(D) and clay coatings (B)–(D) with some quartz cements (C),
(D) can be observed.

cation-oxygen bonds. Hence, surface or partial disso- calcite and quartz showed pore spaces that resembled
lution of silicate minerals, including clays, quartz, and residual (inhibition) growth (Figures 9, 11–13), and it
feldspars (Figures 9, 13), followed by precipitation of is likely that pore fluid changes contribute to these
silicates (quartz, clays) (Figures 12, 14) may be more shapes.
common in these shales than thought. Alternatively,
some of these pores represent residual growth forms
or inhibition of crystal growth. Folk (1974), Dowty Clay Minerals
(1976), Lahann (1978), and others discussed various
mechanisms. There, Mg2+ ions are suggested to induce The six samples from the three formations had clay
calcite growth inhibition because of accumulated lat- contents between 4 and 21% (Table 1) and were com-
tice strain; thus, these ions introduce a poisoning of posed of illite/mica and occasionally small amounts
the crystal growth perpendicular to the calcite c axis of chlorite and kaolinite. Scanning electron micro-
(Folk, 1974). Lahann (1978) suggested that both sur- scopic images showed that most clay minerals often
face charge and colloid chemical changes induce crys- formed fibrous or platelet aggregates on the surface
tal growth inhibition; Dowty (1976) suggested that the of other minerals (Figures 9, 10, 13). Under elevated
template fraction (fraction of total bond energy) dic- burial pressure, alignments of clay minerals were
tates growth of mineral surfaces. In our study, both also observed (Figures 6, 15). The morphology of the

13835_ch09_ptg01_hr_p081-100.indd 95 22/05/13 10:39 AM

 96  Wüst et al

Figure 13. Scanning electron microscopic (SEM) photomicrographs of the Duvernay (1-23-049-25W4; 5886 ft [1794 m])
showing different pore types (arrows) (A)–(D) in these samples. Fracture (A), (C), dissolution and growth (B), and intergranu-
lar porosity (C), (D) were common. Note also the clay platelets that covered the surface of minerals.

clays suggests that the clay minerals (apart from the CONCLUSIONS
mica) were mostly diagenetic and formed likely dur-
ing a later stage of diagenesis. However, we suggest Conventional SEM analyses of shale hydrocarbon res-
that this is the result of rearrangement of existing ervoirs from three different formations of the Western
clay aggregates instead of precipitation and neofor- ­C anadian Sedimentary Basin were combined with
mation from fluids. Such diagenetic clay aggregates routine analytical investigations (XRD, SRA, mercury
are common in sedimentary rocks (e.g., Scotchman, porosimetry, petrography) to characterize mineral
1989). The scope of this study could not determine if composition and assemblage, morphology, organic
the growth of the clays was related to transformation contents, and porosity. The investigations demon-
schemes (smectite-illite, kaolinite-illite) or simply the strated that despite marked diagenetic differences be-
rearrangement of existing illite/mica and kaolinite/ tween these shales, some common texture and pore
chlorite clay minerals. However, in the Montney characteristics were observed. The main findings of
Formation samples, fibrous illite clay minerals (Fig- this study are as follows:
ure  9) appeared to be authigenic, and large mica
sheets (elongated length up to 100 mm) were detrital 1. Scanning electron microscopic morphologi-
­(Figures 8, 10, 11). cal studies of unconventional shales provide

13835_ch09_ptg01_hr_p081-100.indd 96 22/05/13 10:39 AM

 Porosity Characterization of Various Organic-rich Shales from the Western Canadian Sedimentary Basin, Alberta and British Columbia, Canada  97

Figure 14. Scanning electron microscopic (SEM) photomicrographs of the Duvernay (1-23-049-25W4; 5912 ft [1802 m])
showing an overview of this shale (A). Note the deformed textures (clays) and pore space associated (B), (C); some grains
showed competent habits (formed prior to late compaction). Along one former fracture, quartz and calcite cements were ob-
served (bedding parallel) with some fracture porosity (D). In thin section, organic matter is often enriched along such features
(former fractures; see Appendix A9-4).

important information about mineral aggregates, 4. Scanning electron microscopic investigations

cementation, and clay mineral distribution, which of three shale types from the Western Cana-
allows interpretations of diagenetic history. dian Sedimentary Basin showed that large pores
2. Scanning electron microscopic combined with pet- (.2000 nm) were present and had various shapes,
rographic analyses provides critical complementary which may be the primary reason why mercury
information in particular for sediments in which ce- porosimetry fails to identify the larger pores in
mentation and mineral overgrowth are pronounced. these shales as MICP methods identify pore throat
3. Scanning electron microscopic analysis of organic- sizes.
rich shales may not be able to fully characterize 5. The main types of pores in these shales were inter-
the organic material, but organic fragments and granular, dissolution, residual growth (growth
some bituminous material can still be recognized. inhibition), fracture, and phyllosilicate poros-
­Organic matter porosity has been modeled in other ity. Mercury porosimetry identified abundant 5- to
SEM studies using argon- and gallium-ion (FIB) 20-nm pores (i.e., pore throats) that were not identi-
ion-milled samples. fied in the samples using SEM.

13835_ch09_ptg01_hr_p081-100.indd 97 22/05/13 10:39 AM

 98  Wüst et al

Figure 15. Scanning electron microscopic (SEM) photomicrograph of the Duvernay (1-23-049-25W4; 5912 ft [1802 m])
showing details of an apatite grain (A), some dissolution and residual growth features (B). Note that some of these pores (B)
were infilled with secondary authigenic crystals (here quartz, but also calcite and dolomite), illustrating pore fluid chemistry
changes during diagenesis. Most clay minerals show an elongate texture parallel to the bedding planes (C), (D).

ACKNOWLEDGMENTS Ambrose, R. J., R. C. Hartman, M. Diaz-Campos, Y. Akkutlu,

and C. H. Sondergeld, 2010, New pore-scale consid-
We would like to thank N. Minions, R. Monzon, erations for shale gas in place calculations: Society of
F. Wang, and C. Twemlow for lab support and Wayne Petroleum Engineers Unconventional Gas Conference,
Camp, Patrick Monahan, and Craig Hall for their com- February 23–25, Pittsburgh, Pennsylvania, SPE Paper
ments and suggestions that improved this manuscript. 131772, 17 p., doi: 10.2118/131772-MS.
Financial support for Bustin was received from Geo- Burley, S. D., and J. H. S. MacQuaker, 1992, Authigenic
clays, diagenentic sequences and conceptual diage-
science British Columbia, Canada.
netic models in contrasting basin-margin and basin-
center North-Sea Jurassic sandstones and mudstones, in
D. W. Houseknecht and E. D. Pittman, eds., Origin, dia-
REFERENCES CITED genesis, and petrophysics of clay minerals in sandstones:
SEPM Special Publication 47, p. 81–110.
Alexander, G. B., W. M. Heston, and R. K. Iler, 1954, The Conti, J. J., P. D. Holtberg, J. A. Beamon, A. M. Schaal,
solubility of amorphous silica in water: Journal of J. C. Ayoub, and J. T. Turnure, 2011, Annual energy
Physical Chemistry, v. 58, no. 6, p. 453–455, doi: 10.1021/ outlook 2011 with projections to 2035: U.S. Energy
j150516a002. Information Administration, Washington D.C., 235 p.

13835_ch09_ptg01_hr_p081-100.indd 98 22/05/13 10:39 AM

 Porosity Characterization of Various Organic-rich Shales from the Western Canadian Sedimentary Basin, Alberta and British Columbia, Canada  99

Figure 16. (A, top): Incremental pore volume versus pore diameter derived from the mercury porosimetry showing all, and the
four samples with lower incremental pore volume (B, bottom). Note that small pore diameters between 200 and 2000 nm
were documented in the results, which is in contrast to the scanning electron microscopic (SEM) observations.

Curtis, M. E., R. J. Ambrose, C. H. Sondergeld, and Petroleum Conference, October 19–21, Calgary, Alberta,
C. S. Rai, 2010, Structural characterization of gas shales SPE Paper 137693, 15 p., doi: 10.2118/137693-MS.
on the micro- and nano-scales: Canadian Society for Curtis, M. E., R. J. Ambrose, C. H. Sondergeld, and
Unconventional Gas/Society of Petroleum Engineers C. S. Rai, 2011, Transmission and scanning electron
Canadian Unconventional Resources and International microscopy investigation of pore connectivity of gas

13835_ch09_ptg01_hr_p081-100.indd 99 22/05/13 10:39 AM

 100  Wüst et al

shales on the nano-scale: Society of Petroleum Engineers Resources and International Petroleum Conference,
North American Unconventional Gas Conference, June October 19–21, Calgary, Alberta, SPE Paper 138975, 5 p.,
14–16, The Woodlands, Texas, SPE Paper 144391, 10 p., doi: 10.2118/138975-MS.
doi: 10.2118/144391-MS. Nelson, P. H., 2009, Pore-throat sizes in sandstones, tight
Dowty, E., 1976, Crystal structure and crystal growth: 1. The sandstones, and shales: AAPG Bulletin, v. 93, no. 3,
influence of internal structure on morphology: American p. 329–340.
Mineralogist, v. 61, p. 448–459. Riediger, C. L., M. G. Fowler, L. R. Snowdon, F. Goodarzi,
Edwards, D. E., J. E. Barclay, D. W. Gibson, G. E. Kvill, and and P. W. Brooks, 1990, Source rock analysis of the Lower
E. Halton, 1994, Triassic strata of the Western Canadian Jurassic “Nordegg Member” and oil-source rock cor-
Sedimentary Basin, in G. Mossop and I. Shetsen, eds., relations, northwestern British Columbia: Bulletin of
Geological atlas of the Western Canada Sedimentary Canadian Petroleum Geology, v. 38A, p. 236–249.
Basin: Canadian Society of Petroleum Geologists and Riediger, C. L., and J. D. Bloch, 1995, Depositional and diage-
Alberta Research Council, Special Report 4, p. 259–275 netic controls on source-rock characteristics of the Lower
http://www.ags.go.ab.ca/publications/wcsb_atlas/ Jurassic “Nordegg Member,” Western Canada: Journal of
a_ch16/ch_16.html (accessed January 2, 2011). Sedimentary Research, v. A65, no. 1, p. 112–126.
Folk, R. L., 1974, The natural history of crystalline calcium Ross, D. J. K., and R. M. Bustin, 2007, Shale gas potential of
carbonate: Effect of magnesium content and salinity: the Lower Jurassic Gordondale Member, Northeastern
Journal of Sedimentary Petrology, v. 44, p. 40–53. B.C., Canada: Bulletin of Canadian Petroleum Geology,
Fowler, M. G., and L. D. Stasiuk, 2001, Devonian hydrocar- v. 55/1, p. 51–75.
bon source rocks and their derived oils in the Western Schieber, J., 2010, Common themes in the formation and pres-
Canada Sedimentary Basin: Bulletin of Canadian ervation of intrinsic porosity in shales and mudstones—
Petroleum Geology, v. 49, no. 1, p. 117–148. illustrated with examples across the Phanerozoic: Society
Hunt, J. M., 1996, Petroleum geochemistry and geology: of Petroleum Engineers Unconventional Gas Conference,
New York, Freeman, 743 p. February 23–25, Pittsburgh, Pennsylvania, USA, SPE
Keller, L. M., L. Holzer, R. Wepf, P. Gasser, B. Munch, and Paper 132370, 10 p., doi: 10.2118/132370-MS.
P. Marschall, 2011, On the application of focused ion Schott, J., and E. H. Oelkers, 1995, Dissolution and crystal-
beam nanotomography in characterizing the 3D pore lization rates of silicate minerals as a function of chemical
space geometry of Opalinus clay: Physics and Chemistry affinity: Pure and Applied Chemistry, v. 67, p. 903–910.
of the Earth, Parts A/B/C, v. 36, no. 17–18, p. 1539–1544. Scotchman, I. C., 1989, Diagenesis of the Kimmeridge Clay
Lahann, R. W., 1978, A chemical model for calcite crystal Formation, onshore UK: Journal of the Geological Society,
growth and morphology control: Journal of Sedimentary v. 146, no. 2, p. 285–303.
Petrology, v. 48, no. 1, p. 337–344. Sondergeld, C. H., R. J. Ambrose, and C. S. Rai, 2010, Micro-
Lewin, J., 1961, The dissolution of silica from diatom structural studies of gas shales: Society of Petroleum
walls: Geochimica et Cosmochimica Acta, v. 21, no. 3–4, Engineers Unconventional Gas Conference, February
p. 182–198. 23–25, Pittsburgh, Pennsylvania, SPE Paper 131771, 17 p.,
Loucks, R. G., R. M. Reed, S. C. Ruppel, and D. M. Jarvie, doi: 10.2118/131771-MS.
2009, Morphology, genesis, and distribution of Stasiuk, L. D., and M. G. Fowler, 2004, Organic facies in
­n anometer-scale pores in silicious mudstones of the Devonian and Mississippian strata of Western Canada
Mississippian Barnett Shale: Journal of Sedimentary Sedimentary Basin: Relation to kerogen type, paleoen-
Research, v. 79, p. 848–861. vironment, and paleogeography: Bulletin of Canadian
Michalopoulos, P., and R. C. Aller, 2004, Early diagen- Petroleum Geology, v. 52, no. 3, p. 234–255.
esis of biogenic silica in the Amazon delta: Alteration, Stoakes, F. A., and S. Creaney, 1985, Sedimentology of a
authigenic clay formation, and storage: Geochimica et carbonate source rock: The Duvernay Formation of
Cosmochimica Acta, v. 68, no. 5, p. 1061–1085. Alberta, Canada, in M. W. Longman, K. W. Shanley, R. F.
Milner, M., R. McLin, and J. Petriello, 2010, Imaging tex- Lindsay, and D. E. Eby, eds., Rocky Mountain Carbonate
ture and porosity in mudstones and shales: Comparison Reservoirs: SEPM Special Publication 7, p. 343–375.
of secondary and ion-milled backscatter SEM meth- Webb, P. A., and C. Orr Jr., 1997, Analytical methods in fine
ods: Canadian Society for Unconventional Gas/Society particle technology: Norcross, Georgia, Micrometrics
of Petroleum Engineers Canadian Unconventional Instrument Corporation Publishers, 325 p.

13835_ch09_ptg01_hr_p081-100.indd 100 22/05/13 10:39 AM