fractal and fractional

Article
Pore Microstructure and Multifractal Characterization of
Lacustrine Oil-Prone Shale Using High-Resolution SEM:
A Case Sample from Natural Qingshankou Shale
Shansi Tian 1,2,3 , Yuanling Guo 1,2 , Zhentao Dong 4, * and Zhaolong Li 5

1 State Key Laboratory of Shale Oil and Gas Enrichment Mechanisms and Effective Development,
Beijing 100101, China
2 Sinopec Key Laboratory of Shale Oil/Gas Exploration and Production Technology, Beijing 100101, China
3 Institute of Unconventional Oil & Gas, Northeast Petroleum University, Daqing 163318, China
4 School of Geosciences, China University of Petroleum (East China), Qingdao 266580, China
5 No. 1 Geological Logging Company of Daqing Drilling Engineering Company, Daqing 163318, China
* Correspondence: [email protected]

Abstract: Pore structure is one of the important parameters for evaluating reservoirs, critical in
controlling the storage capacity and transportation properties of hydrocarbons. The conventional
pore characterization method cannot fully reflect the pore network morphology. The edge-threshold
automatic processing method is applied to extract and quantify pore structures in shale scanning
electron microscope (SEM) images. In this manuscript, a natural lacustrine oil-prone shale in the
Qingshankou Formation of Songliao Basin is used as the research object. Based on FE-SEM, a high-
resolution cross-section of shale was obtained to analyze the microstructure of pores and characterize
the heterogeneity of pores by multifractal theory. The stringent representative elementary area (REA)
Citation: Tian, S.; Guo, Y.; Dong, Z.;
of the SEM cross-section was determined to be 35 × 35. Four pore types were found and analyzed
Li, Z. Pore Microstructure and in the stringent REA: organic pores, organic cracks, inorganic pores, inorganic cracks. The results
Multifractal Characterization of showed that inorganic pores and cracks were the main pore types and accounted for 87.8% of the total
Lacustrine Oil-Prone Shale Using pore area, and organic cracks were of the least importance in the Qingshankou shale. Inorganic pores
High-Resolution SEM: A Case were characterized as the simplest pore morphologies, with the largest average MinFeret diameter,
Sample from Natural Qingshankou and the least heterogeneity. Moreover, the inorganic cracks had a long extension distance and stronger
Shale. Fractal Fract. 2022, 6, 675. homogeneity, which could effectively connect the inorganic pores. Organic pores were found to
https://doi.org/10.3390/ be the most complex for pore structure, with the least average MinFeret diameter, but the largest
fractalfract6110675
heterogeneity. In addition, the extension distance of the organic cracks was short and could not
Academic Editors: Gianina Dobrescu, effectively connect the organic pore. We concluded that inorganic pores and cracks are a key factor
Florica Papa and Razvan State in the storage and seepage capacity of the Qingshankou shale. Organic pores and cracks provide
limited storage space.
Received: 30 September 2022
Accepted: 9 November 2022
Keywords: lacustrine shale; automatic image processing; organic pores; scanning electron microscope;
Published: 15 November 2022
multifractal
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations. 1. Introduction
Shale oil is stored in organic-rich shale strata with predominantly nano-scale pores,
and is an important future alternative energy source for conventional oil and gas in the
Copyright: © 2022 by the authors.
world [1,2]. The economic exploitation of shale oil is related to its reservoir characteristics,
Licensee MDPI, Basel, Switzerland. viz. the size and distribution of pore-throats [3], the characteristics of mineral and organic
This article is an open access article matter composition [4], and the reservoir space characteristics of different material compo-
distributed under the terms and sitions [5]. The complication of lacustrine shale oil exploitation in China is related to the
conditions of the Creative Commons insufficient understanding of the shale pore structure [6–8]. The multi-scale pore-throat
Attribution (CC BY) license (https:// system developed in oil-prone lacustrine shales is characterized by diverse pore types [9,10],
creativecommons.org/licenses/by/ development of nanoscale pores and strong heterogeneity [11–14]. Therefore, multi-scale
4.0/). and heterogeneous shale pore characterization is a challenging and critical study.

Fractal Fract. 2022, 6, 675. https://doi.org/10.3390/fractalfract6110675 https://www.mdpi.com/journal/fractalfract

Fractal Fract. 2022, 6, 675 2 of 12

Currently, the most widespread conventional pore characterization methods are per-
formed by mercury injection capillary porosimetry (MICP) [15,16], gas adsorption porosime-
try [17,18], and nuclear magnetic resonance (NMR) [19–21]. These methods can only detect
connected pores and cannot fully reflect the pore network [22–24]. Moreover, since these
methods simplify the pore morphology, the pore heterogeneous information about types,
morphologies, mineralogy, and microstructures are not accessible [25,26]. Nevertheless,
SEM technique enables in-depth analysis of the microscopic pore structure of shale [27–29].
The heterogeneity of pore types, pore size distribution, and morphological microstructure
can be revealed by high-resolution SEM images [22,23,29]. The reliability of pore network
characterization depends on the pore extraction accuracy of SEM images. Nowadays,
the mainstream pore extraction methods include manual drawing methods, thresholding
methods [30], edge detection methods [31], and watershed methods [32]. However, these
methods have obvious defects: reliance on geological experience, huge workload, and
being ineffective for organic pore identification. The edge-threshold automatic processing
(ETAP) method reported here is used to address these above identified issues [29].
The pore size distribution, morphology, and fracture extension in shale have self-
similar characteristics, thus fractal theory, especially fractal dimension, is widely used
to quantitatively characterize the heterogeneity of pore structure [33–35]. Based on the
power law of fractal theory, Wang [36] used nitrogen adsorption experiments to discover
the effect of fractal characteristics on the pore structure of shales. In addition, the fractal
dimension of shale with different pore structure types has been studied based on MICP [37],
NMR or SEM [38], with values between two and three [20,21]. However, some of the
complex shale pore networks are not mathematically but statistically similar, and their
fractal dimensions change with the observation scale, thus requiring description by multiple
fractals or continuous spectra of dimensions, termed multiple fractal spectra [38–40]. At
present, there are not many studies that apply multifractal theory to describe the results of
shale pore networks.
In this paper, the lacustrine oil-prone shale in the Qingshankou, Songliao Basin was
used as the research object. High-resolution SEM images of shale were obtained to analyze
the microscopic pore structure heterogeneity. The multifractal methods were used to
assess the heterogeneity of shale pore networks. Eventually, the effect of heterogeneity of
microscopic pore networks on hydrocarbon storage and transport capacity was discussed.

Geological Setting
Songliao Basin [41,42] is located in Northeast China and surrounded by hills and
mountains (Figure 1a), which is the most abundant non-marine sedimentary basin with
oil and gas resources in the world [41,43]. The oil and gas in the basin generally have a
distribution characteristic of upper oil and lower gas [44]. Natural gas is mainly developed
in the Huoshiling formation, Shahezi Formation (shale gas) and Yingcheng Formation,
during fault depression stage, which are continental coal-bearing pyroclastic rock forma-
tions with a maximum thickness of 8000 m [45] (Figure 1b). The oil is mainly developed in
the Denglouku formation, Quantou Formation, Qingshankou Formation (shale oil), Yaojia
Formation, Nenjiang Formation, Sifangtai Formation and Mingshui Formation, during
the depression stage, which are continental clastic oil shale formations with a maximum
thickness of about 5000 m [46].
Qingshankou Formation is the main oil source layer (Figure 1b), which can be divided
into qn1 member and qn23 member from bottom to top. Organic rich shale in qn1 member
is the main oil source rock, with high organic matter abundance. Ro in the main depression
region is 1.0–1.6% [42]. It is a key area for shale oil exploration, and it also provided the
shale sample for this study.
Fractal Fract. 2022, 6, x FOR PEER REVIEW 3 of 12

depression region is 1.0%–1.6% [42]. It is a key area for shale oil exploration, and it also
Fractal Fract. 2022, 6, 675
provided the shale sample for this study. 3 of 12

Figure 1. (a) Sampling position (well X1) and the Songliao Basin. (b) Lithological sketch in the study
Figure 1. (a) Sampling position (well X1) and the Songliao Basin. (b) Lithological sketch in the
area.
study area.

2. Materials
Materials and
and Methods
Investigated Samples
2.1. Investigated Samples
thispaper,
In this paper,lamellar
lamellaroil-prone
oil-prone shale
shale waswas taken
taken to study
to study the different
the different structure
structure char-
characteristics
acteristics among among
porespores and cracks.
and cracks. The lamellar
The lamellar organic-rich
organic-rich felsic felsic
shale shale was chosen
was chosen from
from K1qn1 formation at 2080.46 m in well X1. The mineral composition
K1qn1 formation at 2080.46 m in well X1. The mineral composition of shale is as follows: of shale is as
follows: clay minerals (39.1%), quartz (37.8%), feldspar (10.4%), calcite (6.9%),
clay minerals (39.1%), quartz (37.8%), feldspar (10.4%), calcite (6.9%), siderite (0.4%), and siderite
(0.4%),(4.9%).
pyrite and pyrite (4.9%).of
The results The
theresults
organicofgeochemical
the organic geochemical
experimentsexperiments are as listed
are as listed below: TOC
below:
(2.01 TOCS(2.01
wt.%), wt.%),
1 (1.77 S1S(1.77
mg/g), mg/g),
2 (7.54 mg/g),S2R(7.54 mg/g), Ro (1.13%).
o (1.13%).

2.2. Ar-ion
2.2. Ar-ion Milled
Milled FESEM
FESEM
Sample surfaces were polished by using the Ar-ion milling (AIM) method in a GATAN
Sample surfaces were polished by using the Ar-ion milling (AIM) method in a GA-
Inc, PIPS (precision ion polishing system) [47]. Polished surfaces were oriented perpendicu-
TAN Inc, PIPS (precision ion polishing system) [47]. Polished surfaces were oriented per-
lar to bedding. The treated samples not only had a smooth surface without damage, but also
pendicular to bedding. The treated samples not only had a smooth surface without dam-
restored the real structure inside the rock, and clear mineral boundaries and micro/nano
age, but also restored the real structure inside the rock, and clear mineral boundaries and
pores could be observed.
micro/nano pores could be observed.
Field emission scanning electron microscopy (FE-SEM) used ZEISS GEMINI 500 field
Field emission scanning electron microscopy (FE-SEM) used ZEISS GEMINI 500 field
emission scanning electron microscopy [47]. The instrument had ultra-high resolution.
emission scanning electron microscopy [47]. The instrument had ultra-high resolution. It
It could observe and process secondary electron images and reflected electron images of
could observe and process secondary electron images and reflected electron images of
various solid sample surface morphologies. Images for pore identification were taken at a
various solid sample surface morphologies. Images for pore identification were taken at a
working distance of 4 mm, voltage of 2.0 kV, contrast 70, brightness 55, magnification 1500×,
working distance of 4 mm, voltage of 2.0 kV, contrast 70, brightness 55, magnification
resolution 8192 × 6144, and swelling time of 8 µs. In addition, 5 × 5 groups of images with
1500X,
the sameresolution 8192 × were
magnification 6144, captured
and swelling timethe
to form of 8large
μs. In addition,
field 5 × 5 groups of
of high-resolution images
images.
with the same magnification were captured to form the large field of high-resolution im-
ages.
2.3. Automatic Porosity Qquantification and Determination of REA
A method for automatic shale porosity quantification using an edge-threshold auto-
matic processing (ETAP) technique was applied on all SEM images [29]. The first stage
entailed converting the SEM image of the shale into an 8-bit greyscale image, counting the
number of pixels in each greyscale range from 0–255, and drawing a relationship curve
Fractal Fract. 2022, 6, 675 4 of 12

between the number of pixels and greyscale. The relationship curve gave the grey cut-off
values for the pore (Pcutoff ) and the kerogen (Kcutoff ). The initial pore image and the ini-
tial kerogen-pore image were then obtained by segmenting the original shale greyscale
images using Pcutoff and Kcutoff , respectively. The initial pore image and initial kerogen-
pore image were superimposed, and the total of the peripheral length, area, and inner
circumference of each isolated connected area in the kerogen-pore image was counted
as Lki ; area Ski , long-axis Lli , and short-axis Lsi were also counted. Additionally, Spij was
calculated using the combined areas of all the pores in the superimposed initial pore.
When determining whether an isolated connected area in the initial kerogen-pore image
was a kerogen or mineral pore, the maximum area Spijmax of each isolated connected
area in the kerogen-pore image was found. The kerogen region discriminant function
Qsti = (Spijmax /Ski )/[Lki /Ski /(Lli /Lsi )] was then used to determine whether the isolated
connected area was a kerogen or mineral pore. The isolated connecting area in the original
kerogen-pore image was kerogen when Qsti was less than or equal to 1, and it was an
inorganic pore when Qsti was larger than 1. To determine the edge of the kerogen pores, the
edge detection method was applied to the kerogen region image. An error function, Qerror ,
was set as follows: Qerror = (Aoutside +Ainside )/Ainside , where the value Qerror showed the
effectiveness of the analysis. Values were iteratively rebalanced to achieve the maximum
Ainside and, as a result, to reduce Qerror to achieve the best performance. The discriminant
function Qt was set to Qt = (Ainside /Qerror )/Aboundary , where Aboundary was the total
pixels of pores; it was applied to normalize the value of Ainside /Qerror . When Qt reaches its
maximum value, the optimal threshold was obtained. At this step, the pore image from
the best threshold and the pore image from the edge detection approach were combined.
Finally, Lli /Lsi was used as a discriminant parameter to characterize the pore and crack
(pore: Lli /Lsi < 6; crack: Lli /Lsi ≥ 6) based on the extracted organic pores and inorganic
pores. Subsequently, we extracted five binarized images from the SEM images: region of
OM (organic matter), Ops (organic pores), Ips (inorganic pores), OCs (organic cracks) and
ICs (inorganic cracks).
The quantitative parameters for each pore were derived from ImageJ software (Figure 2).
Then, the Kraver method [22] was used to calculate the circularity, convexity, and elongation.

Figure 2. Illustration of the pore space parameters (Feret angle, Feret diameter, and MinFeret
Figure2. Illustration of the pore space
diameter) [48]. parameters ( feret angle, feret diameter, and minferet diameter)[48].

We used the box-counting method [48] to determine the representative elementary

area (REA). Five parameters (the total surface porosity, surface porosity of Ops and OCs,
48. Zhang, P.; Lu, S.; Li, J.; Chang,
surface X.;of
porosity Li,Ips
J.; Li,
and W.;ICs,
Chen, G.; Wang,
surface S.; Feng,
porosity of OM,W.,percentage
Broad ion of beam-scanning
surface porosity of
electron microscopy
Ops and poreOCs/surface
microstructure and multifractal
porosity of Ips andcharacterization of shale
ICs) were calculated oil reservoir:
within A case
increasing box sizes
ranging from 1 × 1 to 76 ×
sample from Dongying Sag, Bohai Bay Basin, China. Energy Exploration & Exploitation 2020, 38, (3),
76 in a single image.
613-628.
Fractal Fract. 2022, 6, 675 5 of 12

2.4. Multifractal Method

The pore binarized image is considered as data set I. The data set I is segmented into
N(r) partition with scale r. N(r) is equal to 2n , and n is a positive integer.

I
N(r) = = 2n (1)
r
The probability measure of the i-th partition set at scale r can be expressed as equation

Si ( r )
Pi = N(r)
(2)
∑i=1 Si (r)

where Si (r) is the surface porosity of i-th partition.

When the dataset I has multiple fractal characteristics, the probability measure Pi (r)
and the scale r are in a power exponential relationship.

Pi (r) ∝ rai (3)

where ai is the coarse Holder index or singular intensity, which represents the density of
the data distribution of the i-th partition set.
Therefore, different partition sets may have the same singular intensity. Using Na (r) to
denote the cumulative proportion of the partition sets with singular intensity distribution
between a and a ± da.
Na ( r ) ∝ r − f ( a ) (4)
where, f(a) is the singular spectrum.
The partition function is defined as

N(r)
Z(q, r) = ∑ pi (r)q ∝ rτ(q) (5)
i=1

where, q is the weight factor, which ranges from [−∞, +∞], τ(q) is the mass function.

N (r)
log Z(q, r) log ∑i=1 pi (r)q
τ(q) = −lim = −lim (6)
r→0 log r r→0 log r

We can define the generalized dimension D(q) as Equation (7) using Pi (r) and q.

N(r) q
 τ(q) = 1 log ∑i=1 pi (r) , q 6= 1

1−q 1−q log r
D(q) = N(r) (7)
∑i=1 pi (r)logpi (r)
,q = 1


logr

The relationship between α(q), τ(q) and f(a) can be obtained from the Legendre trans-
formation.
dτ(q)
α (q) = (8)
dq
f(a) = qα(q) − τ (q) (9)

3. Results and Discussion

3.1. REA Determination
The pore characterization reliability depends on the appropriate REA. In general, the
pore parameters vary as the box area increases. When the box area is larger than the REA,
the pore parameters do not vary significantly. The REA of shale has been studied based on
the point counting method, with values around 0.01 mm2 [48–50]. However, these studies
focus on the surface porosity representability and ignore the pore type. In this study, we
considered both the pore parameter and pore type to determine the REA. Figure 3a shows
Fractal Fract. 2022, 6, 675 6 of 12

the different sizes of boxes selected for determining the REA, with values varying from
1 × 1 to 76 × 76. When the REA is greater than 15 × 15, total surface porosity becomes
stable, yet surface porosity of OM remains unstable. However, the surface porosity no
longer varies for all pore types when the REA exceeds 35 × 35 (Figure 3b). At this point,
the REA can represent the pore type and porosity characteristics of the shale.

Figure 3. (a) Principle of REA determination, blue box is regular REA, red box is stringent REA;
Figure3: (a) Principle of REA determination, blue box is regular REA, red box is stringent REA; (b) the
(b) the variation of surface porosity for the four pore types versus box area.
variation of surface porosity for the four pore types versus box area.
3.2. Porosity from FE-SEM
According to the relationship between pores and mineral and organic matrix particles,
shale reservoir space can be divided into four categories: OPs, IPs, OPs, ICs (Figure 4a).

Figure
Figure4. (a) The result of REA (a) The result
with4.resolution of REA with
of 9.3nm, resolution
(b) organic of 9.3
pore; (c)nm, (b) organic
organic crack,pores; (c) organic cracks,
(d) inorganic
(d) inorganic pores, (e) inorganic cracks. (Red areas are organic pores, pink areas are organic cracks,
pore, (e) inorganic crack (Red areas
deep are organic
blue areas pores,
are inorganic pink
pores, areas
light are organic
blue areas cracks,
are inorganic deep blue areas
cracks.)
are inorganic pores, light blue areas are inorganic cracks ).
Fractal Fract. 2022, 6, 675 7 of 12

A total of 35,948 OPs were identified for this study, which is about 11.59% of the total
visible surface porosity (Table 1). This reveals that OPs are not the main pore type of the
Qingshankou Shale. OPs are the minimum in size and possess the smallest pore area and
radius (Table 1 and Figure 4b).

Table 1. Pore characteristic results for four pore types.

Pore Area Perimeter Feret MinFeret Area

Pore Type Quantity
Range (nm2 ) Range (nm) Range (nm) Range (nm) (%)

346 1 –1,371,731 2 53–18,944 26–3082 9–1610

Organic pores 11.59 35,948
2906 3 193 73 38
346–6,525,671 53–65,881 26–12,983 9–3684
Inorganic pores 63.56 36,248
22,951 780 275 101
519–407,195 79–4961 34–2315 9–237
Organic cracks 0.61 498
15,874 603 282 51
865–5,250,808 188–70,587 93–22,088 9–2512
Inorganic cracks 24.24 13,453
24,155 1321 562 81
Note: Superscripts 1, 2, and 3 indicate the minimum, maximum and average values, respectively.

IPs (Table 1 and Figure 4d) are mainly intergranular pores of clay minerals and
intergranular pores of quartz particles. IPs account for 63.56% of the overall pore space and
are an important part of the Qingshankou Shale (36,248 pores were identified in this time).
The pore area of IPs varies from 346 to 6,525,761 nm2 (mean 22,951 nm2 ), and the range of
perimeter is from 53 to 65,881 nm (mean 780 nm). The average values of Feret diameter
and MinFeret diameter are 275 and 101, respectively.
The development of OCs in the shale is the lowest, accounting for only 0.61% of
the total pore space (Table 1). It can be seen that OCs at the organic particle boundary
are parallel to the edge (Figure 4c). The Feret diameter of OCs is small with a range of
34–2315 nm, which means that the pores cannot be connected by organic crack effectively.
ICs are the most widely distributed fractures in the shale matrix. A total of 13,453 ICs
was detected, contributing 24.24% to the total pore space (Table 1). It is demonstrated that
ICs are distributed between mineral grains with surprising extension length; the Feret
diameter is up to 22088 nm (Figure 4e). It can be concluded that ICs are an important factor
in the flowability of shale oil.
Four morphological parameters were calculated according to the Klaver method [22]
for characterizing the morphological information of the pores: circularity, convexity, elon-
gation, and Feret angle. Circularity values range between 0 and 1, where a circularity
value of 1 represents a perfect circle. Convexity (0–1) indicates the roughness of the pore
surface, a value of 0 means absolutely smooth. The elongation (0–1) represents the relation-
ship between length and width of the pores, the closer to 1, the narrower and longer the
pores are [22].
The distribution of OPs circularity is obviously skewed to the right, with a frequency
peak of between 0.7 and 1.0, which implies that OPs are closest to a circle (Figure 5a). IPs
have poor circularity with a frequency peak of between 0.4 and 0.5, reflecting the irregular
shape of the particles constituting the IPs. The cracks are supposed to have a minimum
circularity. However, the OCs have a higher circularity with a frequency peak of between
0.3 and 0.5. It is demonstrated that the pores have a smoother surface than the cracks, and
the OPs have the smoothest surface (convexity up to 0.9) (Figure 5b). It is indicated that the
ICs and IPs elongation (peak at 0.9/0.7) is greater than OCs and OPs elongation (peak at
0.8/0.5) (Figure 5c), which is consistent with observation in the SEM images. Finally, the
directionality (Feret Angle) of the four pore types is discussed (Figure 5d). The curves for
the four pore types show a similar two-peak pattern, appearing at 40◦ and 160◦ .
Fractal Fract. 2022, 6, 675 8 of 12

Figure5. FrequencyFigure Frequency

5. (a)
plots of plots of (a)(b)
pore circularity, pore circularity,
pore (b) pore
convexity, convexity,
(c) pore (c) pore and
elongation, elongation,
(d) and
(d) pore FeretAngle.
pore FeretAngle.
The seepage capacity of shale oil is controlled by the diameter of the pore along the
short axis (MinFeret diameter), thus the normalization distributions of MinFeret diameter
was calculated (Figure 6). The results show that micropores (100–1000 nm) are the main
component of the Qingshankou shale, occupying 69% of the total pore. Of the four pore
types, the IPs had the largest proportion of micropores (71.19%), followed by ICs (63.14%),
OPs (55.14%) and OCs (51.52%).

Figure 6. Pore size distributions of four pore types.

Figure6. Pore size distributions of four pore types.
3.3. Multifractal Characteristics
The multifractal characteristics of the four pore types are calculated and discussed in
this section. The q is set from −20 to 20 with an interval of 0.5. The distribution curves of
q~Dq and α~f(α) for different pore types are shown in Figure 7.
Fractal Fract. 2022, 6, 675 9 of 12

Figure 7. (a) Generalized dimensional and (b) multifractal spectra of four different pore types.
Figure7. (a) Generalized dimensional and (b) multifractal spectra of four different pore types.
Dq decreases rapidly with increasing q when q < 0, and the Dq varies greatly among
pore type. Dq decreases slowly with the increase of q when q > 0, and the difference of
Dq between pore type was small (Figure 7a). A multitude of pore structure parameters
(D0 , D1 , D2 ) can be obtained by Dq-q. D0 refers to the capacity dimension characterizing the
average pore structure distribution and reflecting the pore structure complexity [48]. The in-
formation dimension and the correlation dimension are denoted as D1 and D2 , respectively.
D1 /D0 indicates the dispersion of the porosity to the pore diameter [33] (Table 2).

Table 2. Multifractal parameters of four pore types.

Pore Type D0 D1 D2 D1 /D2 αmin αmax α0 ∆α A

Organic pore 1.92 1.55 1.32 0.81 0.95 2.97 2.25 2.02 1.82
Inorganic pore 1.21 1.01 0.91 0.84 0.71 2.14 1.43 1.44 1.02
Organic crack 1.92 1.74 1.62 0.91 1.27 3.42 2.13 2.16 0.67
Inorganic crack 1.86 1.73 1.63 0.93 1.16 3.14 2.01 1.97 0.75

Each pore type has typical multiple fractal characteristics, exhibiting D0 > D1 > D2 . Ops
and OCs have the largest D0 values (1.918/1.919), while IPs have the smallest D0 values
(1.208) (Table 2). It can be concluded that Ops and OCs have the most complex structure,
while OPs are simpler. The distribution of D1 and D2 for each pore type was the same
as each other. OCs and ICs have higher D1 /D0 value (0.91/0.93), thus the pore volume
varies greatly at different scales; whereas the pore volume of the Ops and IPs are more
concentrated at different scales with lower D1 /D0 value (0.81/0.84) (Table 2).
The multifractal spectral singularity index can effectively scale the pore structure
heterogeneity, αmax and αmin indicate the maximum and minimum values of multifractal
spectral singularity index, respectively (Figure 7b). ∆α is the difference between αmax and
αmin , namely ∆α = αmax −αmin . α0 indicates the singularity index when q = 0. Multifractal
spectral skewness A = (α0 −αmin )/(αmax −α0 ). ∆α and α0 indicate the non-homogeneity
of the pore structure. The larger the values are, the greater the heterogeneity of the pore
structure is [48] (Table 2).
The distribution of ∆α and α0 is consistent. The ∆α varies from 1.44 to 2.02 (α0
between 1.43 and 2.25), which is significantly lower than the value of Shahejie shales in the
Bohaiwan basin, which is the most important shale oil producer in China [48]. It can be
concluded that the pore heterogeneity of Qingshankou Shale is weaker. OPs (∆α = 2.02,
α0 = 2.25) are more heterogeneous than IPs (∆α = 1.44, α0 = 1.43), and cracks (average
α0 = 2.07, average ∆α = 2.06) are more heterogeneous than pores (average α0 = 1.84, average
∆α = 1.73). This conclusion is consistent with the result from Dq-q. In addition, the As
of the pores (average A = 0.71) are all greater than those of the cracks (average A = 0.92)
which indicates that the pores have a lower index and fluctuations (Table 2).
Fractal Fract. 2022, 6, 675 10 of 12

4. Conclusions
The stringent REA selection requires that the porosity of each pore type does not
change significantly with box size. The stringent REA of high-resolution SEM image was
identified to be 35 × 35 for lacustrine oil-prone shale from the Qingshankou Formation.
A stringent REA was chosen to analyse the pore structure. The multifractal theory was
applied to explain the heterogeneity of each pores-type network.
Four pore types were found in the stringent REA: organic pores, organic cracks,
inorganic pores, inorganic cracks. Inorganic pores and inorganic cracks were the main pore
types and accounted for 87.8% of the total pore volume, and organic cracks were the least
important to the Qingshankou shale.
Organic pores are the most complex for pore structure, the least average MinFeret
diameter (38 nm), but the largest D0 , α0 , A (1.92, 2.25, 1.82). Moreover, the extension
distance of the organic cracks is short and cannot effectively connect the organic pore.
Inorganic pores are the simplest for pore morphologies, the maximum average Min-
Feret diameter (101 nm), and the minimum D0 , D1 , D2 , D1 /D2 , α0 , ∆α (1.21, 1.01, 0.91, 0.84,
1.43, 1.44). In addition, the inorganic cracks have long extension distances and stronger
homogeneity, which can effectively connect the inorganic pores.
Therefore, we conclude that inorganic pores and inorganic cracks are a key factor in
the storage and seepage capacity of Qingshankou shale. Organic pore and organic cracks
provide limited storage space.

Author Contributions: Conceptualization, S.T. and Z.D.; methodology, S.T.; software, Y.G.; vali-
dation, Y.G.; formal analysis, Z.D.; investigation, S.T.; resources, S.T.; data curation, S.T.; writing—
original draft preparation, S.T.; writing—review and editing, Z.D.; visualization, Z.L.; supervision,
Z.D.; project administration, S.T.; funding acquisition, S.T. All authors have read and agreed to the
published version of the manuscript.
Funding: This study was partly funded by China Postdoctoral Science Foundation (2020M670878),
Superior Youth Foundation of Heilongjiang Province (YQ2020D002), Project of Daqing Guiding
Science and Technology Plan (zd-2019-18), Science and Technology Project of Heilongjiang Province
(No.2020ZX05A01).
Institutional Review Board Statement: Not applicable.
Data Availability Statement: Not applicable.
Conflicts of Interest: We declare that we have no financial or personal relationships with other
people or organizations that can inappropriately influence our work, there is no professional or
other personal interest of any nature or kind in any product, service and/or company that could be
construed as influencing the position presented in, or the review of, the manuscript entitled, “Pore
microstructure and multifractal characterization of lacustrine oil-prone shale using high-resolution
SEM: A case sample from natural Qingshankou shale”.

References
1. Mănescu, C.B.; Nuño, G. Quantitative effects of the shale oil revolution. Energy Policy 2015, 86, 855–866. [CrossRef]
2. Caineng, Z.; Zhi, Y.; Jingwei, C.; Rukai, Z.; Lianhua, H.; Shizhen, T.; Xuanjun, Y.; Songtao, W.; Senhu, L.; Lan, W. Formation
mechanism, geological characteristics and development strategy of nonmarine shale oil in China. Pet. Explor. Dev. 2013, 40, 15–27.
3. Yang, Y.; Wang, K.; Zhang, L.; Sun, H.; Zhang, K.; Ma, J. Pore-scale simulation of shale oil flow based on pore network model.
Fuel 2019, 251, 683–692. [CrossRef]
4. Chen, Z.; Jiang, W.; Zhang, L.; Zha, M. Organic matter, mineral composition, pore size, and gas sorption capacity of lacustrine
mudstones: Implications for the shale oil and gas exploration in the Dongying depression, eastern China. AAPG Bull. 2018, 102,
1565–1600. [CrossRef]
5. Su, S.; Jiang, Z.; Xuanlong, S.; Li, Z.; Zhu, R. The effects of shale pore structure and mineral components on shale oil accumulation
in the Zhanhua Sag, Jiyang Depression, Bohai Bay Basin, China. J. Petrol. Sci. Eng. 2018, 165, 365–374. [CrossRef]
6. Shuangfang, L.; Haitao, X.; Min, W.; Dianshi, X.; Wenbiao, H.; Junqian, L.; Liujuan, X.; Shansi, T.; Sen, W.; Jijun, L. Several key
issues and research trends in evaluation of shale oil. Acta Pet. Sin. 2016, 37, 1309.
7. Feng, Q.; Xu, S.; Xing, X.; Zhang, W.; Wang, S. Advances and challenges in shale oil development: A critical review. Adv.
Geo-Energy Res. 2020, 4, 406. [CrossRef]
Fractal Fract. 2022, 6, 675 11 of 12

8. Suyun, H.; Wenzhi, Z.; Lianhua, H.; Zhi, Y.; Rukai, Z.; Songtao, W.; Bin, B.; Xu, J. Development potential and technical strategy of
continental shale oil in China. Pet. Explor. Dev. 2020, 47, 877–887.
9. Saraji, S.; Piri, M. The representative sample size in shale oil rocks and nano-scale characterization of transport properties. Int. J.
Coal Geol. 2015, 146, 42–54. [CrossRef]
10. Zhao, J.; Jin, Z.; Hu, Q.; Jin, Z.; Barber, T.; Zhang, Y.; Bleuel, M. Integrating SANS and fluid-invasion methods to characterize pore
structure of typical American shale oil reservoirs. Sci. Rep. 2017, 7, 15413. [CrossRef]
11. Hu, T.; Pang, X.; Jiang, S.; Wang, Q.; Zheng, X.; Ding, X.; Zhao, Y.; Zhu, C.; Li, H. Oil content evaluation of lacustrine organic-rich
shale with strong heterogeneity: A case study of the Middle Permian Lucaogou Formation in Jimusaer Sag, Junggar Basin, NW
China. Fuel 2018, 221, 196–205. [CrossRef]
12. Wang, Y.; Cheng, H.; Hu, Q.; Liu, L.; Jia, L.; Gao, S.; Wang, Y. Pore structure heterogeneity of Wufeng-Longmaxi shale, Sichuan
Basin, China: Evidence from gas physisorption and multifractal geometries. J. Pet. Sci. Eng. 2022, 208, 109313. [CrossRef]
13. Guan, M.; Liu, X.; Jin, Z.; Lai, J. The heterogeneity of pore structure in lacustrine shales: Insights from multifractal analysis using
N2 adsorption and mercury intrusion. Mar. Pet. Geol. 2020, 114, 104150. [CrossRef]
14. Suyun, H.; Bin, B.; Shizhen, T.; Congsheng, B.; ZHANG, T.; Yanyan, C.; LIANG, X.; Lan, W.; Rukai, Z.; Jinhua, J. Heterogeneous
geological conditions and differential enrichment of medium and high maturity continental shale oil in China. Pet. Explor. Dev.
2022, 49, 257–271.
15. Wang, F.; Yang, K.; You, J.; Lei, X. Analysis of pore size distribution and fractal dimension in tight sandstone with mercury
intrusion porosimetry. Res. Phys. 2019, 13, 102283. [CrossRef]
16. Vavra, C.L.; Kaldi, J.G.; Sneider, R.M. Geological applications of capillary pressure: A review. AAPG Bull. 1992, 76, 840–850.
17. Labani, M.M.; Rezaee, R.; Saeedi, A.; Al Hinai, A. Evaluation of pore size spectrum of gas shale reservoirs using low pressure
nitrogen adsorption, gas expansion and mercury porosimetry: A case study from the Perth and Canning Basins, Western Australia.
J. Pet. Sci. Eng. 2013, 112, 7–16. [CrossRef]
18. Rani, S.; Padmanabhan, E.; Prusty, B.K. Review of gas adsorption in shales for enhanced methane recovery and CO2 storage. J.
Petrol. Sci. Eng. 2019, 175, 634–643. [CrossRef]
19. Wang, X.; Wang, M.; Li, Y.; Zhang, J.; Li, M.; Li, Z.; Guo, Z.; Li, J. Shale pore connectivity and influencing factors based on
spontaneous imbibition combined with a nuclear magnetic resonance experiment. Mar. Petroleum Geol. 2021, 132, 105239.
[CrossRef]
20. Liu, Z.; Liu, D.; Cai, Y.; Yao, Y.; Pan, Z.; Zhou, Y. Application of nuclear magnetic resonance (NMR) in coalbed methane and shale
reservoirs: A review. Int. J. Coal Geol. 2020, 218, 103261. [CrossRef]
21. Liu, Y.; Yao, Y.; Liu, D.; Zheng, S.; Sun, G.; Chang, Y. Shale pore size classification: An NMR fluid typing method. Mar. Petrol. Geol.
2018, 96, 591–601. [CrossRef]
22. Klaver, J.; Desbois, G.; Urai, J.L.; Littke, R. BIB-SEM study of the pore space morphology in early mature Posidonia Shale from the
Hils area, Germany. Int. J. Coal Geol. 2012, 103, 12–25. [CrossRef]
23. Klaver, J.; Desbois, G.; Littke, R.; Urai, J.L. BIB-SEM characterization of pore space morphology and distribution in postmature to
overmature samples from the Haynesville and Bossier Shales. Mar. Petroleum Geol. 2015, 59, 451–466. [CrossRef]
24. Ji, L.; Lin, M.; Cao, G.; Jiang, W. A multiscale reconstructing method for shale based on SEM image and experiment data. J. Pet.
Sci. Eng. 2019, 179, 586–599. [CrossRef]
25. Li, J.; Lu, S.; Xie, L.; Zhang, J.; Xue, H.; Zhang, P.; Tian, S. Modeling of hydrocarbon adsorption on continental oil shale: A case
study on n-alkane. Fuel 2017, 206, 603–613. [CrossRef]
26. Zheng, J.; Wang, Z.; Gong, W.; Ju, Y.; Wang, M. Characterization of nanopore morphology of shale and its effects on gas
permeability. J. Nat. Gas Sci. Eng. 2017, 47, 83–90. [CrossRef]
27. Kelly, S.; El-Sobky, H.; Torres-Verdín, C.; Balhoff, M.T. Assessing the utility of FIB-SEM images for shale digital rock physics. Adv.
Water Resour. 2016, 95, 302–316. [CrossRef]
28. Tahmasebi, P.; Javadpour, F.; Sahimi, M. Three-dimensional stochastic characterization of shale SEM images. Transp. Porous Media
2015, 110, 521–531. [CrossRef]
29. Tian, S.; Bowen, L.; Liu, B.; Zeng, F.; Xue, H.; Erastova, V.; Greenwell, H.C.; Dong, Z.; Zhao, R.; Liu, J. A method for automatic
shale porosity quantification using an Edge-Threshold Automatic Processing (ETAP) technique. Fuel 2021, 304, 121319. [CrossRef]
30. Liu, J.-F.; Cao, X.-L.; Xu, J.; Yao, Q.-L.; Ni, H.-Y. A new method for threshold determination of gray image. Geomech. Geophys.
Geo-Energy Geo-Resour. 2020, 6, 72. [CrossRef]
31. Tian, S.; Brown, L.; Zeng, F.; Xue, H.; Lu, S.; Erastova, V.; Greenwell, C. In A Method for Automatic Shale Porosity Quantification
Using New Advanced Edge Threshold Technique. Geophys. Res. Abstr. 2019, 21, EGU2019-1446.
32. Zhao, S.; Li, Y.; Wang, Y.; Ma, Z.; Huang, X. Quantitative study on coal and shale pore structure and surface roughness based on
atomic force microscopy and image processing. Fuel 2019, 244, 78–90. [CrossRef]
33. Liu, K.; Ostadhassan, M. Multi-scale fractal analysis of pores in shale rocks. J. Appl. Geoph. 2017, 140, 1–10. [CrossRef]
34. Wang, Q.; Hu, Y.; Zhao, J.; Ren, L.; Zhao, C.; Zhao, J. Multiscale apparent permeability model of shale nanopores based on fractal
theory. Energies 2019, 12, 3381. [CrossRef]
35. Song, W.; Wang, D.; Yao, J.; Li, Y.; Sun, H.; Yang, Y.; Zhang, L. Multiscale image-based fractal characteristic of shale pore structure
with implication to accurate prediction of gas permeability. Fuel 2019, 241, 522–532. [CrossRef]
Fractal Fract. 2022, 6, 675 12 of 12

36. Wang, M.; Xue, H.; Tian, S.; Wilkins, R.W.; Wang, Z. Fractal characteristics of Upper Cretaceous lacustrine shale from the Songliao
Basin, NE China. Mar. Petroleum Geol. 2015, 67, 144–153. [CrossRef]
37. Zhou, N.; Lu, S.; Wang, M.; Liu, W.; Guan, Y.; Tan, H.; Wang, Z. Applicability of fractal capillary pressure models to sandstones. J.
Pet. Sci. Eng. 2020, 185, 106626. [CrossRef]
38. Zhang, S.; Yan, J.; Hu, Q.; Wang, J.; Tian, T.; Chao, J.; Wang, M. Integrated NMR and FE-SEM methods for pore structure
characterization of Shahejie shale from the Dongying Depression, Bohai Bay Basin. Mar. Petroleum Geol. 2019, 100, 85–94.
[CrossRef]
39. Singh, D.P.; Wood, D.A.; Hazra, B.; Singh, P.K. Pore properties in organic-rich shales derived using multiple fractal determination
models applied to two Indian Permian basins. Energy Fuels 2021, 35, 14618–14633.
40. Xu, J.; Wu, K.; Li, R.; Li, Z.; Li, J.; Xu, Q.; Chen, Z. Real gas transport in shale matrix with fractal structures. Fuel 2018, 219, 353–363.
[CrossRef]
41. Liu, B.; Sun, J.; Zhang, Y.; He, J.; Fu, X.; Yang, L.; Xing, J.; Zhao, X. Reservoir space and enrichment model of shale oil in the first
member of Cretaceous Qingshankou Formation in the Changling sag, southern Songliao Basin, NE China. Pet. Explor. Dev. 2021,
48, 608–624. [CrossRef]
42. Liu, B.; Wang, H.; Fu, X.; Bai, Y.; Bai, L.; Jia, M.; He, B. Lithofacies and depositional setting of a highly prospective lacustrine shale
oil succession from the Upper Cretaceous Qingshankou Formation in the Gulong sag, northern Songliao Basin, northeast China.
AAPG Bull. 2019, 103, 405–432. [CrossRef]
43. Liu, Y.; Liu, B.; Cheng, L.; Xing, J.; Tian, S.; Huang, S.; Dong, S. Modeling of tectonic-thermal evolution of cretaceous qingshankou
shale in the changling sag, southern Songliao Basin, NE China. Front. Earth Sci. 2021, 9, 694906. [CrossRef]
44. Yang, W.; Yongkang, L.; Ruiqi, G. Formation and evolution of nonmarine petroleum in Songliao Basin, China. AAPG Bull. 1985,
69, 1112–1122.
45. Xu, Z.; Jiang, S.; Liu, L.; Wu, K.; Li, R.; Liu, Z.; Shao, M.; Jia, K.; Feng, Y. Natural gas accumulation processes of tight sandstone
reservoirs in deep formations of Songliao Basin, NE China. J. Natural Gas Sci. Eng. 2020, 83, 103610. [CrossRef]
46. Dong, T.; He, S.; Wang, D.; Hou, Y. Hydrocarbon migration and accumulation in the Upper Cretaceous Qingshankou Formation,
Changling Sag, southern Songliao Basin: Insights from integrated analyses of fluid inclusion, oil source correlation and basin
modelling. J. Asian Earth Sci. 2014, 90, 77–87. [CrossRef]
47. Liu, B.; Wang, Y.; Tian, S.; Guo, Y.; Wang, L.; Yasin, Q.; Yang, J. Impact of thermal maturity on the diagenesis and porosity of
lacustrine oil-prone shales: Insights from natural shale samples with thermal maturation in the oil generation window. Int. J. Coal
Geol. 2022, 261, 104079. [CrossRef]
48. Zhang, P.; Lu, S.; Li, J.; Chang, X.; Li, J.; Li, W.; Chen, G.; Wang, S.; Feng, W. Broad ion beam-scanning electron microscopy pore
microstructure and multifractal characterization of shale oil reservoir: A case sample from Dongying Sag, Bohai Bay Basin, China.
Energy Explor. Exploit. 2020, 38, 613–628.
49. Cosenza, P.; Prêt, D.; Fauchille, A.-L.; Hedan, S. Representative elementary area of shale at the mesoscopic scale. Int. J. Coal Geol.
2019, 216, 103316. [CrossRef]
50. Medina, F.J.; Jausoro, I.; Addato, M.A.F.; Rodriguez, M.J.; Tomassini, F.G.; Caneiro, A. On the evaluation of Representative
Elementary Area for porosity in shale rocks by Field Emission Scanning Electron Microscopy. Energy 2022, 253, 124141. [CrossRef]