Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
Energy Reports 6 (2020) 1876–1887
Contents lists available at ScienceDirect
Energy Reports
journal homepage: www.elsevier.com/locate/egyr
Research paper
Synthetical study on the difference and reason for the pore structure of
the No. 3 coal reservoir from the southern Qinshui Basin, China, using
mercury intrusion porosimetry, low-temperature N2adsorption, low
field nuclear magnetic resonance, and nuclear magnetic resonance
cryoporometry
Huihu Liu a,∗, Ibrahim Issa Farida, Shuxun Sang b,∗∗, Jianhua Shang a, Haiyan Wu c,
Hongjie Xu a, Pingshong Zhang a, Qimeng Liu a
aSchool of Earth and Environment, Anhui University of Science & Technology, Huainan 232001, China
bSchool of Resource and Earth Science, China University of Mining and Technology, Xuzhou 221116, China
cSchool of foreign language, Anhui University of Science & Technology, Huainan 232001, China
article info
Article history:
Received 9 April 2020
Received in revised form 28 June 2020
Accepted 13 July 2020
Available online xxxx
Keywords:
Pore structure
Long field nuclear magnetic resonance
Nuclear magnetic resonance
cryoporometry
Low-temperature N2adsorption
Mercury intrusion porosimetry
Southern Qinshui Basin
abstract
This study aimed to synthetically investigate the pore structure characteristics of coal samples from
the southern Qinshui Basin in China, by mercury intrusion porosimetry (MIP), low-temperature N2
adsorption (LTNA), low field nuclear magnetic resonance (LFNMR), and nuclear magnetic resonance
cryoporometry (NMRC) methods and to reveal the reasons for the differences in the pore structures
of the coal samples. The results show the multimodality of the pore size distribution (PSD) for the
different pore diameters from all the samples using MIP and LFNMR, and the bimodality of the PSD of
all the samples using LTNA and NMRC. The peak representation in the micropores and transition pores
through MIP is generally consistent with that of LTNA, LFNMR, and NMRC, and the peak representation
of the PSD in the micropores through MIP is consistent with that of LFNMR. Owing to the differences
in the analysis principles and the calculation models from the different analysis methods applied, clear
differences are observed in the total volume of the transition pores and micropores based on MIP, LTNA,
LFNMR, and NMRC. The anthracitic samples have better connectivity in terms of the PSD than that of
the semianthracitic and low-volatile bituminous samples. As revealed through a synthetic analysis,
abundantly disconnected transition pores and micropores, as well as poorly connected micropores,
occur in coal, particularly in semianthracite and low-volatile bituminous coal. The difference in the PSD
between the different coal samples based on the different analysis methods applied is synthetically
controlled based on Ro,max, the mineral content in the coal, and the burial depth of the coal samples.
©2020 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc- nd/4.0/).
1. Introduction
The pores in a coal reservoir are considered important chan-
nels of coalbed methane (CBM) output (Fu et al.,2017;Liu et al.,
2017), and the pore size distribution (PSD) partly determines the
flow and permeability properties of both CBM and water (Yao and
Liu,2012;Wang et al.,2017).
Numerous methods are used for measuring the PSD of coal,
including mercury intrusion porosimetry (MIP), low-temperature
∗Correspondence to: School of Earth and Environment, Anhui University of
Science & Technology, Taifeng Street No.168, Huainan, 232001, China.
∗∗ Corresponding author.
E-mail addresses: xixiinformation@163.com (H.H. Liu),
Shuxunsang@163.com (S.X. Sang).
N2adsorption (LTNA), and high-resolution electron microscopy,
the most commonly used of method require mercury porosimetry
and gas adsorption using N2as the adsorbate. Mercury intrusion
is effective in characterizing the porosity with pore diameters
ranging from 2∼50 nm, particularly >7.2 nm; with pore diam-
eters of >50 nm (Clarkson et al.,2013), a structural distortion
of the pores may be generated at a lower limit of the pore size
(approximately 3 nm) owing to the influences of the compress-
ibility (Gane et al.,2004;Shao et al.,2018). In addition, mercury
intrusion may be difficult to achieve because of the destruction of
the coal reservoir under high pressure. LTNA has an upper pore
diameter limit of approximately 300 nm (Clarkson et al.,2011)
and is applicable to micropores (pore diameters of <10 nm) and
mesopores (pore diameters ranging from 10∼1000 nm). Further-
more, an overlapping pore size region exists, allowing MIP and
https://doi.org/10.1016/j.egyr.2020.07.011
2352-4847/©2020 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
H.H. Liu, I.I. Farid, S.X. Sang et al. / Energy Reports 6 (2020) 1876–1887 1877
LTNA to be compared. Although comparisons of MIP and LTNA
have largely been conducted regarding the PSD for coal (Clarkson
and Bustin,1999;Mastalerz et al.,2008), few discussions on their
application for different types of coal have been reported.
In comparisons of traditional MIP and LTNA, some nonde-
structive analytical methods such as low field nuclear magnetic
resonance (LFNMR) have been proposed to characterize the coal
porosity with a pore radius greater than 2 nm. Although LFNMR
can be used to measure the PSD of interconnected pores (Yao
et al.,2010;Yao and Liu,2012), it cannot be used to quantitatively
analyze the closed pores in coal because the PSD is determined
through a T2spectrogram, which can be obtained using an in-
verse calculation of the decrement signature of an echo from a
columnar sample under saturated water conditions. Furthermore,
nuclear magnetic resonance cryoporometry (NMRC) is a method
of measuring the phase transition of the testing liquid in the pores
of coal, that is, of the freezing–melting action experienced by
the liquid. The principle of NMRC has been described in previous
reports (Jackson and McKenna,1990;Mitchell et al.,2008), and
it has been shown that the PSD can be determined based on
the relationship between the melting point and the pore volume,
which can describe not only the PSD of the interconnected pores
but also that of the closed pores for coal (Mitchell et al.,2008;
Petrov and Furo,2009). This method has been applied in uncon-
ventional oil and gas reservoirs (Mitchell et al.,2008;Petrov and
Furo,2010;Kondrashova and Valiullin,2013;Zhao et al.,2017), in
which the closed pores can be opened after the coal samples are
crushed. Given the discussion, NMRC, unlike LTNA, can be applied
at elevated temperatures and pressures and is a nondestructive
approach, unlike MIP.
Based on the analysis above, MIP has difficulty revealing the
structure of the micropores, LTNA cannot be used to describe the
macropores, whereas LFNMR can describe the PSD of almost all
types of pores, and NMRC can characterize the PSD for a diameter
size ranging from 2∼500 nm. Moreover, MIP, LTNA, and LFNMR
can only describe the PSD of the interconnected pores in coal,
but cannot quantify the disconnected pores in coal; NMRC can
describe not only the interconnected pores in coal but also the
disconnected pores in coal. Above all, the different types of PSD
methods each have different applicabilities, and thus, how the
different results can be synthesized to reveal the PSD of the pores
in coal should be determined. Therefore, this paper will analyze
the PSD of coal not only in terms of the interconnected pores but
also in terms of the disconnected pores and synthetically reveal
the reason for the differences in the PSD of coal by the different
experimental methods.
The southern Qinshui basin (SQB) is a mature CBM basin. CBM
exploration and development has received widespread public
concern (Wang et al.,2015;Liu et al.,2018;Song et al.,2018;Luo
et al.,2019), and many studies have been conducted on the PSD of
the SQB (Li et al.,2017;Zhang and Fu,2018;Zhang et al.,2018;Xu
et al.,2019), with different analysis methods applied to evaluate
the PSD. However, as a more substantial problem, comprehensive
and systematic research on the differences in the multilevel pore
structures, from the nano and microscale to the macroscale, be-
tween the different types of coal materials in the SQB is relatively
inadequate. Therefore, it is necessary to strengthen the explo-
ration of coal reservoirs and the connected and disconnected pore
structures in multiple pore size ranges both comprehensively and
systematically. In this study, an integrated approach combining
traditional characterization methods such as MIP and LTNA for
the PSD using new nondestructive techniques applying LFNMR
and NMRC was conducted to accurately describe the differences
in the PSD in different types of coal materials from the SQB.
In this study, Hodot’s pore structure classification system was
employed to analyze the PSD. That is, the pore diameters of
the micropores, transition pores, mesopores, and macropores are
<10 nm, 10∼100 nm, 100∼1000 nm, and >1000 nm respectively
(Hodot,1961).
2. Experiments
2.1. Sampling and test for basic physical characteristics of coal sam-
ples
To reveal the differences in the PSD for different coal ranks
based on different testing methods, four coal samples with dif-
ferent types of coal ranks from the Sihe (SH), Chengzhuang (CZ),
Yuwu (YW), and Xinyuan (XY) coal mines of the SQB in China,
were used for analysis in this study. The samples were tested
for their basic characteristics, such as their Ro,max value, maceral
composition, proximate analyses, and permeability, according to
GB/T 6948-2008 (AQSIO,2008a), GB/T8899-2013 (AQSIO,2013),
GB/T212-2008 (AQSIO,2008b), and SY/T 6385-2016, respectively.
Moreover, an AXIO Imager M1m Microspectrophotometer was
used for measuring the coal petrography (Ro,max and maceral
composition), and the test for the coal permeability was con-
ducted with the steady state method under constant confining
pressure according to the original stress of the coal seam (NEA,
2016). The collection of the coal samples adopted columnar sam-
pling according to GB/T 475-2008, and the coal samples were
pulverized to particles with a size range of 0.18∼0.25 mm for
the determination of the Ro,max value, maceral composition, prox-
imation, and MIP analysis, according to GB/T 16773-2008 (AQSIO,
2008c).
2.2. Test on the PSD of coal samples
A set of experiments, including MIP, LTNA, LFNMR, and NMRC
methods, were conducted in series for each sample. MIP was
applied using an Autopore IV 9500 (Micromeritics Instrument,
USA) at a pressure of up to 60,000 psia (413.7 MPa), following ISO
15901-1-2016 (ISO,2016), and its experimental course adopted
constant pressure control. LTNA was conducted on an automated
surface area and pore size analyzer, Tristar II 3020 (Micromerit-
ics Instrument, USA), according to ISO 15901-2007 (ISO,2007).
Pulverized coal with a 45∼60 mesh particle size was used as the
experimental sample in LTNA, and nitrogen was applied as the
adsorbate under an analysis bath temperature of ∼195.85 ◦C. The
N2data collected for the crushed sample were interpreted using
multipoint BET analysis for the surface area and BJH analysis
for the PSD (Gregg and Sing,1982;Clarkson and Bustin,1999;
Clarkson et al.,2013).
LFNMR measurements are obtained using a MicroMR23-025V
LFNMR analyzer (Niumag Corporation Ltd, China), which applied
the relationship between the pore size and the condensation
temperature of the fluid in the pores to estimate the PSD. The
main preparation method of the coal samples for LFNMR includes
mechanical crushing of the coal samples, generating block coal
samples with a size range from 3∼5 cm, and saturating the coal
samples in water under a vacuum pressure of 20 MPa for a 48 h
period; the experimental principle and process were specified in
a previous study (Yao and Liu,2012). A schematic diagram of the
LFNMR is shown in Fig. 1.
The main principle of LFNMR is described as follows. The
spinning hydrogen nuclei in the fluid manifest NMR relaxation
behavior under the combined action of a radio frequency field
and a static magnetic field, which can be expressed as follows
through T2:
1
T2
=1
T2B
+1
T2S
+1
T2D
(1)
where T2is the relaxation time of the fluid, with unit ms; T2B
is the body relaxation time, with unit ms; T2Sis the surface
relaxation time, which is caused by the interaction between the
fluid and the pore surface, with unit ms; and T2Dis the relaxation
1878 H.H. Liu, I.I. Farid, S.X. Sang et al. / Energy Reports 6 (2020) 1876–1887
Fig. 1. Schematic diagram of LFNMR analyzer MicroMR 23-025V.
time caused by diffusion, with unit ms. In addition, when the
magnetic field is the same (the corresponding magnetic intensi-
ties are extremely low) and the echo time is sufficiently short, T2D
can also be neglected. Here, T2can be expressed as follows:
1
T2
≈1
T2B
=ρ(S
V) (2)
where Sis the surface area of the pores, cm2;Vis the pore
volume, cm3; and ρis the transverse surface relaxation strength
of the coal. Finally, the T2spectrogram can be converted into a
diagram of the PSD by the linear conversion method (Wang et al.,
2018).
NMRC measurements are obtained using Micro 12-010V-T
NMRC analyzers (Niumag Corporation Ltd, China). The main
preparation of the coal samples in the NMRC includes mechanical
pulverization, generating coal particles with a 35∼50 mesh; the
processing of the coal samples under a temperature of 373.15
K for a 24 h period; and the saturation of the coal samples in
water. The experimental procedure and calibration of the NMRC
are reported from a relevant study (Zhao et al.,2017). A schematic
diagram of the NMRC is shown in Fig. 2. The NMRC can be
used to directly obtain the pore volume from the linear relation
between the pore volume and the signal intensity. The NMRC is
applied to acquire the pore information from the melting process
and is not influenced by the background signals. The detailed
technical parameters and the detailed experimental parameters
of the LFNMR and NMRC analyzers are listed in Table 1.
In addition, XRD was employed to analyze the content and
types of minerals in the coal samples. Pulverized coal samples
with a mesh size of <325 (0.045 mm) were applied for the XRD
experiments. The observations and quantitative analyses of the
pores and the minerals for all the coal samples were conducted
on a ZEISS Sigma FE-SEM operating at 20 kV and equipped with
an energy dispersive spectrometer for the analyses of the min-
eral composition, based on SY/T 5162-2014 and SY/T 6189-1996
(CNPC,1996,2014). The small coal pillars used for FE-SEM were
polished using a cross-section polisher to grind off the oxide layer
(approximately 1∼2µm thick) of the coal pillars and were not
spluttered with a gold coating or other materials.
3. Results and discussion
3.1. Basic characteristics of the coal sample
Table 2 shows the burial depth of the coal samples, as well as
the permeability, maceral composition, coal quality, Ro,max value,
and bulk density. As shown in Table 2, the SH and CZ samples are
anthracite, the YW sample is semianthracite, and the XY sample is
low volatile bituminous coal. The coal permeability has a weak re-
lation with the burial depth, and the maceral composition and is
strongly associated with the coalification degree, mineral content,
and Aad. The coal permeability of the SH and CZ samples is clearly
higher than that of the YW and XY samples, which is related to
the coalification degree (Ro,max). However, the permeability of the
SH sample is lower than that of the CZ sample, which is relevant
to the mineral content and Aad. The SH sample has the highest
mineral content and Aad; a higher mineral content may result in
filling of the pores or fractures and may lead to a decrease in the
permeability. Among all samples, the YW sample has the deepest
burial depth, and its permeability is clearly lower than that of
the SH and CZ samples. Table 3 shows the mineral composition
of all the coal samples; as Table 3 indicates, the contents of
the clay minerals from the SH, CZ, and XY samples are clearly
higher than that of the YW sample. The filling amount of the
clay minerals strongly affects the coal permeability, and among
the selected coal samples, the XY sample has the highest content
of clay minerals; therefore, its permeability is the lowest. In the
anthracitic samples, the contents of the clay minerals of the SH
sample are higher than those of the CZ sample, which results in
a lower coal permeability of the SH sample than that of the CZ
sample. For the semianthracitic and low volatile bituminous coal,
the amount of clay mineral in the XY sample is higher than that
of the YW sample, which leads to a higher permeability of the
coal of the YW sample than that of the XY sample.
3.2. Genetic types of the pores in the coal samples
The genetic types of the pores in coal from the SQB were
previously discussed (Bustin et al.,2008;Liu et al.,2017). Fig. 3
shows the pore types in the coal samples. Clay minerals such as
kaolinite (Fig. 3a, b, g, j, k), dolomite (Fig. 3a, b, e, g), and illite
(Fig. 3g and k) are widely developed, whereas calcite (Fig. 3e)
and pyrite (Fig. 3j) are sporadically found in coal samples, filling
in the protogenetic pores of the coal samples. The pore types in
the coal samples mainly contain gas, intercrystal, and dissolved
pores, the size of which varies from several to several hundred
nanometers (Fig. 3c, f, i, l). The genetic types of pores in the coal
samples suggest that the PSD of the coal samples is affected by
the filling of the minerals.
3.3. Pore structure and its influence factor
3.3.1. PSD determined from MIP
The relationships of the cumulative pore volume with the
intrusion pressure and the pore diameter in the coal sample plot
using mercury intrusion and mercury extrusion are shown in
Fig. 4, the pore volume and percentage of the different types
of pores are indicated in Table 4, and the surface areas of the
different types of pores are shown in Table 4.
As indicated in Fig. 4 and Table 4, the PSD suggests that there
is significant substantial volume change in the transition pores to
the micropores; the ratio of the macropores is approximately 10%
with a low pore volume, and the ratio of the mesopores is lowest
in terms of the pore volume among all the samples, whereas that
of the micropores is the highest. The MIP suggests that there
is a multimodality of the PSD for all the coal samples, and the
peak values of the incremental pore volume correspond to pore
diameter of approximately 10, 30, 80, and >1000 nm, meaning
that the pore volumes experience a remarkable increase when
the pore diameter is below 100 nm, which further suggests that
the micropores develop well at a diameter of below 100 nm, and
develop partial macrospores in the coal. A comparison between
Fig. 4 and Table 4 indicates that the peak of the incremental
pore volume at the pore diameter is consistent with the volume
H.H. Liu, I.I. Farid, S.X. Sang et al. / Energy Reports 6 (2020) 1876–1887 1879
Fig. 2. Schematic diagram of NMRC pore analyzer Micro 12-010V-T.
Table 1
Specifications of LFNMR analyzer MicroMR23-025V and NMRC analyzer 12-010V-T.
Attribute Parameter
LFNMR NMRC
Magnet type Permanent magnet Permanent magnet
Magnetic field intensity (T) 0.055 ±0.01 0.3 ±0.05
Probe coil diameter (mm) 25 11
Temperature control range (◦C) −30 to 40 −30 to 40
Temperature control accuracy (◦C) 35 ±0.02 35 ±0.02
Cooling rate (◦C/min) Average 1 Average 1
Sample volume (cm3) 0.5–1 0.5–1
Aperture measurement effective range (nm) 10∼100,000 2–500
Echo time (ms) 0.16 0.1
90◦pulse width (µs) 8 5
180◦pulse width (µs) 15.6 10
Cumulative sampling number 32 256
Wait time (ms) 2000 500
Table 2
Characteristics for permeability, coal petrology, coal quality, and Ro,max, bulk density of coal samples.
Sample ID Burial depth (m) Permeability (mD) Maceral composition (wt %) Proximate (wt %) Ro,max (%) Bulk density (g/cm3)
Vitrinite Inertinite Mineral Mad Aad Vdaf
SH 326 0.064 68.61 15.78 15.61 1.48 13.12 6.32 3.33 1.30
CZ 457 0.077 72.22 20.39 7.39 2.71 12.18 6.94 2.96 1.26
YW 539 0.022 69.74 22.56 7.70 1.10 11.98 13.44 2.19 1.28
XY 400 0.013 74.93 17.84 7.23 0.81 5.35 15.26 1.81 1.19
Notes: wt %, weight percentage; Mad, moisture content; Aad , ash content; Vdaf, volatile content.
Table 3
Miner types in coal samples of SH sample, CZ sample, YW sample, XY sample.
Sample ID Miner type and its content (%)
Kaolinite Illite Chlorite Feldspar Calcite Dolomite Quartz Rutile Bauxite Apatite
SH 15.73 52.35 0 7.42 11.27 1.55 11.68 0.00 0 0
CZ 30.06 30.17 0 26.07 3.21 2.71 2.78 1.02 3.98 0
YW 8.21 38.00 0 2.01 0.00 1.41 33.01 3.20 0 14.16
XY 73.27 5.42 0 8.00 2.61 4.23 6.47 0 0 0
fraction of the PSD, and when the pore diameter is lower than
100 nm, the volume fraction between the micropores and the
transition pores has a remarkable ratio. However, when the curve
of the mercury extrusion lags behind the curve of the mercury
intrusion when the pore diameter is between 10 nm and 80 nm,
the cumulative pore volume is almost invariable when the pore
diameter is beyond 80 nm, which indicates that there may be
abundant disconnected pores in the range of the transitional
pores in the coal samples, The different coal samples have similar
characteristics in terms of the multimodality of the PSD, but it
can be seen that there is a difference in the peak amplitude of
the incremental pore volume, which can be generated by the
difference in the coal rank, maceral composition, and coal quality.
Meanwhile, all the samples show the greatest pore volume and
surface area within a pore size (diameter) of below 100 nm.
The surface area from the transition pores and the micropores is
dominant for all types of pores, which is consistent with the PSD.
Based on the results shown in Table 4, the XY sample has the
highest total pore volume, the CZ sample has the highest porosity,
the SH sample has the lowest total pore volume and porosity, and
the YW sample has the lowest surface area. For the anthracitic
samples, the total pore volume, porosity, and surface area of the
CZ sample are all higher than those of the SH sample. This result
is consistent with the coal permeability of the two samples, and
further reveals the effects of the clay mineral content, which
means that filling the clay minerals in the pores vastly affects the
1880 H.H. Liu, I.I. Farid, S.X. Sang et al. / Energy Reports 6 (2020) 1876–1887
Fig. 3. Pore types and mineral composition in SH sample (a, b, c), CZ sample (d, e, f), YW sample (g, h, i), and XY sample (j, k, l) by FE-SEM.
Fig. 4. Relationship of cumulative Hg pore volume with pressure and pore diameter in coal plots for SH sample (a), CZ sample (b), YW sample (c), and XY sample
(d).
H.H. Liu, I.I. Farid, S.X. Sang et al. / Energy Reports 6 (2020) 1876–1887 1881
Table 4
Pore volume and its percent in different coal samples by MIP.
Pore volume (10−4cm3/g) Volume fraction of PSD (%) ϕMIP (%)
Sample V1V2V3V4VtV1/VtV2/VtV3/VtV4/Vt
SH 29.18 9.28 86.12 200.02 324.60 8.99 2.86 26.53 61.62 4.22
CZ 37.49 16.45 101.50 228.36 383.80 9.77 4.28 26.45 59.50 4.84
YW 24.70 21.16 110.09 190.76 346.71 7.12 6.10 31.75 55.03 4.44
XY 60.48 20.05 100.70 207.08 389.03 15.55 5.15 25.87 53.23 4.65
Notes: V1∼V4, pore volume of macropore, mesopore, transition pore, and micropore respectively; Vt, total pore volume; V1/Vt∼V4/Vt, means percentage from
macropore, mesopore, transition pore, and micropore in the total pore volume respectively; ϕMIP means total porosity by MIP, %.
pore structure of the SH sample (Table 2). For the semianthracitic
sample and the low volatile bituminous sample, the total pore
volume, porosity, and surface area of the XY sample are all higher
than those of the YW sample, which is consistent with the coal
permeability of the two samples, and further reveals the effect
from the content of the clay minerals; that is, the filling of the
clay minerals in the pores substantially affects the pore structure
of the XY sample (Table 2). As indicated in Table 2, the total
pore volume of the coal sample is affected by the macropores
and the mesopores, thereby directly affecting the porosity. Based
on a comparison of the total pore volume with the porosity of
all the samples, the total pore volume of the XY sample is the
highest, followed by the CZ, YW, and SH samples; in addition,
the sequence of the total pore volume is consistent with the total
volume of the macropores and the mesopores, which results in a
higher total pore volume of the XY sample than that of the CZ,
YW, and SH samples. As shown in Table 4, the porosity of the CZ
sample is the highest, followed by the XY, YW, and SH samples.
This sequence of porosity is inconsistent with the sequence of the
total pore volume for all the coal samples, which occurs because
the porosity of a coal sample is related not only to its total pore
volume, but also to its bulk density. As shown in Table 2, the
sequence of the bulk density of all the samples is consistent with
the porosity of all the samples, and the XY sample has the lowest
bulk density and the largest volume of the coal matrix, which
results in a decrease in the porosity. This result demonstrates
why the XY sample has the largest total pore volume among all
the samples, and why the porosity of the XY sample is lower
than that of the CZ sample. Based on a comparison of the results
from the total pore volume (Table 4) with the coal permeability
(Table 2) of all the samples, it can be seen that the sequences of
the total pore volume and the coal permeability of all the samples
are inconsistent, which occurs because the coal permeability is
related not only to the PSD but also to the connectivity of the
pores of the coal sample.
A comparison shows that the SH sample has the lowest total
pore volume and total surface area among all of the coal samples
(Tables 4 and 5) and that the distribution of the total surface area
is inconsistent with the total pore volume of all the samples;
more specifically, the CZ sample has the highest pore volume
of the micropores, followed by the XY, YW, and SH samples,
whereas the CZ sample has the largest total surface area, followed
by the XY, SH, and YW samples. However, the surfaces of all
the coal samples show good consistency with the volume of the
micropores for all the samples, which occurs because the total
surface area of the coal sample is determined by the number of
pores, and is particularly related to the number of micropores;
the YW sample has a higher total pore volume and a lower total
surface area than the SH sample, in which is because the SH
sample having a higher pore volume and surface area of the
micropores.
3.3.2. PSD determined from LTNA
As discussed in corresponding studies (Bustin et al.,2008;
Clarkson et al.,2013), N2adsorption at ∼196 ◦C can be used to
investigate the volume of the vast micropores to the transition
pores. The adsorption curves from the SH, CZ, and XY samples are
shown in Fig. 5. As Fig. 5 indicates, the CZ sample has the highest
absorption, whereas the YW and SH samples have moderate
adsorption, and the XY sample has the lowest absorption, which
demonstrates that the pores in the CZ sample develop the best,
the pores in the YW and SH samples develop relatively well,
and the pores in the XY sample develop the worst. In addition,
there is a wide hysteresis loop and an inflection point shown
in the adsorption curve of the SH sample, and a relatively wide
hysteresis loop and an inflection point in the adsorption curves
of the CZ and YW samples, which indicate that the open pores
develop the best in the SH sample; however, they develop well
in the CZ and YW samples, but there is almost no hysteresis loop
or inflection point in the XY sample, suggesting that there is an
abundance of closed pores.
Plots of the cumulative and incremental pore volumes versus
the pore size for nitrogen show the pore volume distributions
(corresponding to the pore size) of the samples (Fig. 6).
As shown in Fig. 6a, when the pore diameter is less than
approximately 20 nm, the cumulative pore volume of the SH
sample is more than that of the CZ, YU, and XY samples, which
indicates that the development of the micropores from the SH
sample is superior to that of the CZ, YU, and XY samples. When
the pore diameter is more than 20 nm, the cumulative pore
volume of the CZ sample increases dramatically and is far greater
that of the other coal samples. When the pore diameter is higher
than approximately 30 nm, the cumulative pore volume of the
YW sample exceeds that of the SH sample. Fig. 6a shows that
the cumulative pore volume of the XY sample is always lower
than that of the SH, CZ, and YW samples. The results indicate that
the cumulative pore volumes of the SH and XY samples increase
slowly from Fig. 6a, which may be related to the high content of
minerals, such as clay and carbonate minerals (Tables 2 and 3).
Most of the coal samples exhibit a remarkable cumulative pore
volume when the pore diameter reaches 40 nm. As shown in
Fig. 6b, a bimodality of the incremental pore volume (peaks at
approximately 20 and 40 nm) and a trough in the incremental
pore volume (trough at ∼30 nm) occur in the PSD for all the
coal samples; the pore volume shows a remarkable increment
within the pore diameter ranges of 0∼20 and 30∼40 nm, and
when the pore diameter is beyond 40 nm, the incremental pore
volume of all the samples drastically decreases. A quick decrease
in the incremental pore volume from the transition pores with a
pore diameter of 20∼30 nm and beyond 40 nm may be related to
the growth of the transition pores and the filling of the minerals
within the pores.
The pore volumes and BET surface areas obtained from the
N2adsorption analysis are shown in Table 6. As indicated in
Table 6, the CZ sample exhibits the highest total volume of the
transition pores and micropores, whereas the XY sample exhibits
the least total volume of the transition pores and micropores,
the SH sample exhibits the highest BET surface area, and the XY
sample demonstrates the smallest BET surface area. The total pore
volume of a coal sample is mainly controlled by the ratio of the
1882 H.H. Liu, I.I. Farid, S.X. Sang et al. / Energy Reports 6 (2020) 1876–1887
Table 5
Surface area of pore with different pore diameter from coal samples by MIP.
Sample Surface area (m2/g) Percentage of PSD on surface area (%)
S1S2S3S4StS1/StS2/StS3/StS4/St
SH 0.001 0.02 1.80 16.66 18.49 0.01 0.11 9.75 90.13
CZ 0.002 0.03 2.10 19.03 21.17 0.01 0.14 9.93 89.92
YW 0.002 0.05 2.11 15.72 17.87 0.011 0.26 11.79 87.94
XY 0.005 0.03 2.07 17.05 19.16 0.03 0.17 10.81 88.99
Notes: S1∼S4means surface area of macropore, mesopore, transition pore, and micropore respectively; St, total pore
volume; S1/ St∼S4/ Stmeans percentage of surface area from macropore, mesopore, transition pore, and micropore
in total surface area respectively.
Table 6
Pore volume and BET surface area of coal sample determined by LTNA.
Sample ID Pore volume (10−4cm3/g) Volume fraction of PSD (%) BET surface (m2/g)
V3V4VtV3/VtV4/Vt
SH 9.85 6.84 16.69 59.02 40.92 1.36
CZ 22.43 3.89 26.32 85.22 14.78 0.45
YW 15.60 3.78 19.38 80.49 19.51 0.48
XY 4.14 0.07 4.21 98.41 1.59 0.06
Notes: V3∼V4, transition pore, and micropore respectively; Vt, total pore volume; V3/Vt∼V4/Vtmeans percentage
from transition pore, and micropore in the total pore volume respectively.
Fig. 5. N2adsorption isotherms for SH sample (a), CZ sample (b), YW sample (c), and XY sample (d).
Fig. 6. Cumulative (a) and incremental (b) N2pore volume plots for SH sample, CZ sample, YW sample, and XY sample.
transition pores, that is, the volume of the transition pores is
higher, and the total pore volume in the coal sample is higher.
As a comparison between the total pore volume determined from
MIP and the total pore volume determined from LTNA indicates,
the distribution of the pore volume from the different pore di-
ameters when using MIP is different from the distribution of the
pore volume from the different pore diameters when using LTNA;
that is, the total pore volume in a coal sample is controlled by the
H.H. Liu, I.I. Farid, S.X. Sang et al. / Energy Reports 6 (2020) 1876–1887 1883
volume of the transition pores and micropores when using MIP,
and the total pore volume in the coal sample is controlled by the
volume of the transition pores. It is worth mentioning that the XY
sample has the highest total pore volume when using MIP, but
has the lowest total pore volume when using LTNA, which may
be related to the XY sample having the highest amount of clay
minerals. The results when using LTNA shown in Table 6 indicate
that the total pore volume from the anthracitic samples is higher
than the total pore volumes from the semianthracitic sample and
the low volatile bituminous sample, which suggests that the coal-
ification degree (Ro,max) affects the PSD of the coal sample. The
BET surface area of the coal sample is mainly determined by the
micropores, that is, the ratio of the volume from the micropores
to the total pore volume is high. Because the BET surface area
of the coal sample is high, the growth of the micropores and the
filling degree of the clay minerals affect the BET surface area of all
the samples in the studied area. Through the integration with the
above analysis and the results from Tables 2,3, and 6and Figs. 5
and 6, the effects of Ro,max, and the mineral content on the PSD
of all the coal samples are clear.
3.3.3. PSD determined from LFNMR and NMRC
According to Eq. (2),T2spectral distributions can be converted
into the PSD of the coal samples. Figs. 7 and 8show the T2
spectra distributions and PSD of all the coal samples when using
the LFNMR, respectively. As shown in Figs. 7 and 8, with the T2
distribution with the smallest pores having the shortest relax-
ation times and the largest pores having the longest relaxation
times, the relaxation time is only 0.1 ms, which can be used to
detect micropores with a pore diameter of 10 nm; however, the
relaxation is only 0.9 ms, which can be used to detect transition
pores with a pore diameter of approximately 10∼100 nm. The T2
spectrum distribution of all samples (with two weak peaks shown
in the enlarged section in Figs. 7 and 8), particularly the T2spec-
trum distribution of the YW and XY samples, present three clear
peaks. The incremental pore volume of all the samples presents
multimodality, with the three peaks representing the transition
pores, macropores, and fracture. Among all of the samples, the
matching diameters of the transition pores with the first peak
of the incremental pore volume for the SH, CZ, YW, and XY
samples are approximately 60, 40, 10, and 18 nm, respectively. As
indicated in Fig. 8a and b, the porosity of all the samples is mainly
due to the contribution of the transition pores, whereas the
porosity of the mesopores and macropores from all the samples
is extremely low; in particular, the porosity of the mesopores is
the lowest for the all samples. When the pore diameter is greater
than 1000 nm, the porosity from all the samples is mainly due to
the contributions of the macropores and fractures. Through the
comparisons, the cumulative porosity of the micropores from the
SH and CZ samples is clearly lower than that of the YW and XY
samples, and the cumulative porosity of the transition pores from
the SH and CZ samples is obviously higher than that of the other
samples (Fig. 8a). The incremental porosity of the micropores
from the YW and XY samples is higher than that of the SH and CZ
samples (Fig. 8b). A comparison of the results for the pore volume
when using MIP, LTNA, and LFNMR shows that the ratio of the
micropores from the SH and CZ samples is lower than that of the
YW and XY samples when using LFNMR, which differs from the
ratio of the micropores when using MIP and LTNA. This indicates
that the SH and CZ samples mainly develop micropores with a
smaller pore diameter, in contrast to the YW and XY samples.
These micropores with a smaller pore diameter in the SH and CZ
samples may be residual gas pores, and perhaps even gas pores
filled with clay minerals, and are difficult to detect because of the
relaxation time (T2time) of LFNMR.
The pore volume fraction was calculated according to the
cumulative and incremental porosities, the results of which are
Fig. 7. LFNMR T2distribution at saturated water for SZ sample, CZ sample, YW
sample, and XY sample.
Table 7
Porosity of different pore and volume fraction of selected coal samples tested
by LFNMR.
Sample ID Porosity (%) Volume fraction of PSD (%)
ϕ1ϕ2ϕ3ϕ4ϕtϕ1/ϕtϕ2/ϕtϕ3/ϕtϕ4/ϕt
SH 0.70 1.49 7.05 0.01 9.25 7.57 16.11 76.21 0.11
CZ 0.51 0.61 8.00 0.27 9.39 5.43 6.50 86.49 2.88
YW 0.49 0.11 1.04 0.60 2.24 21.88 4.91 46.43 26.78
XY 0.66 0.03 1.25 0.30 2.24 29.46 1.34 55.81 13.39
Notes: ϕ1∼ϕ4, porosity of pore volume from macropore, mesopore, transition
pore, and micropore respectively; ϕt, total porosity; ϕ1/ϕt∼ϕ4/ϕtmeans per-
centage from macropore, mesopore, transition pore, and micropore in the total
pore volume respectively.
shown in Table 7. As indicated in Table 7, there is an obvious
difference in the porosity and volume fraction among each of the
samples. The SH and CZ samples show a higher total porosity
and in particular have a high porosity of the transition pores, in
contrast to the YW and XY samples. The YW and XY samples
have a higher porosity of the micropores, in contrast to the SH
and CZ samples. Matching the porosities, the volume fraction of
the transition pores from the SH and CZ samples is higher than
that of the YW and XY samples, and the volume fraction of the
micropores from the YW and XY samples is higher than that of
the SH and CZ samples. Comparing Fig. 8 and Table 7, the pore
diameter associated with the first peak from the transition pores
is 10∼60 nm; in addition, the transition pores have an advantage
in the volume fraction of the PSD for all the samples, although
the micropores have a higher proportion in the semianthracitic
sample and low volatile bituminous sample. The location where
the first peak of the incremental porosity occurs at the pore
diameter is consistent with the volume fraction of the transition
pores.
Integrating Tables 2,3, and 7, and Fig. 8, there is a clear control
from Ro,max in the coal reservoir, and the total porosity of the
anthracitic samples is higher than that of the semianthracite coal
samples and higher than that of the low volatile bituminous coal;
therefore, the total porosity of the SH and CZ samples is higher
than that of the semianthracite coal sample, such as the YW
sample and is higher than that of the low volatile bituminous coal
sample, such as the XY sample. Second, the mineral content and
burial depth of the coal both have a remarkable effect on the total
porosity of the coal samples; in addition, the SH sample has a high
mineral content, leading to a total porosity below that of the CZ
sample, whereas the YW sample has the deepest burial depth,
which leads to a total porosity equal to that of the XY sample.
Cumulative and incremental intrusion plots using the NMRC
are shown in Fig. 9. As indicated in Fig. 9, there are relatively
1884 H.H. Liu, I.I. Farid, S.X. Sang et al. / Energy Reports 6 (2020) 1876–1887
Fig. 8. LFNMR PSD curves at saturated water for SZ sample, CZ sample, YW sample, and XY sample (a: cumulative porosity; b: incremental porosity).
clear differences in the cumulative and incremental pore volumes
among all the samples. As shown in Fig. 9, the SH, CZ, and XY
samples have a high cumulative pore volume, whereas the YW
coal sample has a relatively low cumulative pore volume. It is
easy to see that the cumulative pore volumes of the XY and
SH samples are higher than those of the CZ and YW samples.
The results from NMRC are different from those of MIP, LTNA,
and LFNMR, which occurs because NMRC can detect not only the
interconnected pores but also the disconnected pores. According
to Tables 2 and 3, the total amounts of clay minerals, calcite,
and dolomite in the SH, CZ, YW, and XY samples are 80.90%,
66.15%, 47.62%, and 85.53%, respectively, which suggests that a
number of disconnected pores may occur in the SH, CZ, and XY
samples; in addition, the amount of disconnected pores from the
XY sample is the highest, followed by the SH and CZ samples,
which is in accordance with the analysis results from Fig. 5.
Meanwhile, the cumulative pore volume from NMRC has a high
negative correlation with the burial depth of the coal samples,
and the compaction of the burial depth on the coal seam results
in the pores closing. Among the coal samples, the burial depth of
the YW sample is the deepest, and the compaction of the burial
depth on the YW sample is the most significant; therefore, the
cumulative pore volume of the YW samples is the lowest.
According to Fig. 9b, the incremental pore volume shows
multimodality, and the incremental pore volume reaches a peak
at the different pore diameters. The incremental pore volume
of the SH sample shows a short peak at a pore diameter of
approximately 2 nm, a relatively obvious peak at a pore diameter
of 10 nm, and an obvious peak at a pore diameter of 30 nm,
whereas the incremental pore volume of the CZ shows three
main peaks at pore diameters of approximately 3, 10, and 60 nm.
The incremental pore volume of the YW sample shows peaks at
pore diameters of 2, 4, 6, 10, 20, and 60 nm. The incremental
pore volume of the XY sample shows peaks at pore diameters of
approximately 2, 30, and 100 nm. The incremental pore volume
reaches a maximum at a pore diameter of 30 nm for the SH
and XY samples, whereas the incremental pore volume reaches
a maximum at a pore diameter of 60 nm for the CZ and YW
samples. Comparing the cumulative and incremental pore vol-
umes, all samples have a relatively high pore volume when the
pore diameter is between 20 and 30 nm. As seen from Fig. 9b,
when the pore diameter is lower than 30 nm, the incremental
pore volume of the XY sample has the fastest growth, which is
far higher than the incremental pore volume of the SH, CZ, and
YW samples, indicating that the development of the disconnected
pores with pore diameters below 30 nm in the XY sample is
far better than that of the other coal samples. When the pore
diameter is beyond 30 nm, the incremental pore volumes of the
XY and SH samples show a downward trend overall, and the
incremental pore volumes of the CZ and SH samples are higher
than those of the XY and YW samples, which demonstrates that
the development of the disconnected pores with pore diameters
greater than 30 nm in the CZ and SH samples is better than that
of the XY and YW samples.
Table 8 shows the pore volume fraction of the PSD for all
samples according to the statistical results of NMRC. As shown in
Table 8, the volume fraction of micropores and transition pores
from all the samples is 77.80%∼90.77%, the volume fraction of
the micropores is slightly more than that of the transition pores,
and the volume fraction of the mesopores is the lowest. The
volume fraction of the mesopores from the SH, CZ, YW samples is
approximately 20%, which is approximately 10% higher than that
of the XY sample. Integrating the results shown in Tables 2,3, and
8, and Fig. 8, an obvious effect from the Ro,max, mineral content,
and burial depth on the generation of pores in the coal reservoir
is shown, all of which are factors that affect the pores; in addition,
the burial depth of the coal samples has remarkable control over
the pores of the coal samples.
3.4. Comprehensive analysis of the PSD based on MIP, LTNA, LFNMR,
and NMRC
As discussed above, the pore volume, porosity, and surface
area differ based on the different testing methods applied, and
the differences in the pore volume, porosity, and surface are
caused by the difference in the testing principle and the assumed
theory models of the different testing methods. MIP is effective
at analyzing the mesopores and macropores of coal samples but
has difficulty distinguishing the transition pores and micropores
of coal samples. LTNA is effective at analyzing the transition
pores and micropores of coal samples but is unable to analyze
the mesopores and macropores of coal samples. Moreover, the
destruction of the pore structure occurs under high pressure
when using MIP and LTNA, especially for MIP. LFNMR is effective
at analyzing pores with a wide pore diameter. NMRC is effec-
tive at analyzing pore sizes from micropores to mesopores. MIP,
LTNA, and LFNMR can detect the connected pores; in contrast,
MIP and LTNA have difficulty detecting poorly interconnected
pores, whereas LFNMR does not, and NMRC can detect not only
interconnected pores but also disconnected pores.
As indicated in Figs. 4–9, MIP suggests a multimodality for
all coal samples (peaks at pore diameters of approximately 10,
30, 80, and >1000 nm), LTNA suggests a bimodality for all coal
samples with peaks at pore diameters of approximately 20 and
40 nm, LFNMR suggests a multimodality for all coal samples with
peaks at pore diameters of approximately 10 and 20 nm, and
NMRC suggests a multimodality for all coal samples with peaks
at pore diameters of approximately 30 and 60 nm. The peak
representation in the micropores and transition pores when using
MIP is generally consistent with that of LTNA, LFNMR, and NMRC,
H.H. Liu, I.I. Farid, S.X. Sang et al. / Energy Reports 6 (2020) 1876–1887 1885
Fig. 9. Cumulative (a) and incremental (b) NMRC pore volume plots for SH sample, CZ sample, ZZ sample, YW sample, BF sample, and XY sample.
Table 8
Pore volume and volume fraction of different coal samples tested by NMRC.
Sample ID Pore volume (10−4cm3/g) Volume fraction of PSD (%)
V2V3V4VtV2/VtV3/VtV4/Vt
SH 168.96 334.91 257.41 761.28 22.20 43.99 33.81
CZ 152.57 339.76 225.94 718.27 21.24 47.30 31.46
YW 101.16 194.15 132.79 428.10 23.63 45.35 31.02
XY 66.88 363.98 294.62 725.48 9.22 50.17 40.61
Notes: V2∼V4, mesopore, transition pore, and micropore respectively; Vt, total pore volume; V2/Vt∼V4/Vtmeans percentage from
mesopore, transition pore, and micropore in the total pore volume respectively.
whereas the peak representation in the macropores when using
MIP is consistent with that of LFNMR. As shown in Tables 4,6,
and 7, the pore volume fraction from the micropores and transi-
tion pores are consistent when applying MIP, LNTA, and LFNMR,
which is not different from that of NMRC (Table 8), in which the
pore volume fraction from the micropores and transition pores is
approximately 80% except for that of the XY sample; in particular,
the pore volume fraction of the transition pores shows the highest
percentage in all types of pores from most of the samples. To
compare the differences from the results of the analysis methods,
the SH sample is chosen as an example, and the synthetic results
are as shown in Table 9, based on the results in Figs. 4–9, and
Tables 4,6,7, and 8.
Table 9 clearly shows the differences among the ranges of the
pore diameters, the dominant pore diameter of the pore volume,
and the volume fraction of the PSD for the SH sample. As shown
in Table 9, MIP and LFNMR can detect all interconnected pores
with different pore diameters, although the volume fraction of the
micropores from LFNMR is clearly lower than that of MIP, which
occurs because of the difficulty in testing the micropores owing
to the short echo time when using LFNMR. As a nondestructive
testing method, LFNMR can detect not only the connected pores
but also poorly connected pores, and as a result, the volume frac-
tion of the mesopores and the transition pores when using LFNMR
is higher than those when using MIP. The volume fraction of
the transition pores when using LTNA is clearly higher than that
when using MIP. Mercury has difficulty entering disconnected
and poorly connected mesopores and transition pores, and some
information from these pores is difficult to obtain through the
use of MIP. Based on a comparison with MIP, nitrogen easily
flows in poorly connected mesopores and transition pores, and
the volume fraction from the mesopores and transition pores is
high. Owing to the occurrence of the remarkably disconnected
pores, this leads to a high volume fraction of the transition pores
when using NMRC compared with MIP. Furthermore, the vol-
ume fractions of the mesopores and transition pores when using
NMRC are higher than those of MIP, which indicates that there
is an abundance of disconnected pores in coal as revealed using
NMRC. Based on the above analysis, there are remarkably poorly
connected and disconnected mesopores and transition pores in
the SH sample.
As indicated above, the pore volume in coal is mainly from the
contribution of the transition pores and micropores based on the
difference in the PSD from the different coal samples when tested
using MIP, LTNA, LFNMR, and NMRC; the comparison results of
the transition pores and micropores of the PSD based on the
different analysis methods are shown in Table 10. As shown in Ta-
ble 10, there are clear differences in the total pore volume of the
transition pores and micropores when using MIP, LTNA, LFNMR,
and NMRC, as discussed above, which is due to the differences
in the analysis principles and the calculation models from the
different analysis methods. There is no reason to compare the
total pore volume from the different analysis methods. There is an
obvious difference between the different coal samples in the total
pore volume when applying MIP, as analyzed in Sections 3.3.1
through 3.3.3, and the difference in the pore volume between
the different coal samples is controlled by Ro,max, the mineral
content in the coal, and the burial depth of the coal sample. The
SH sample has a relatively low pore volume when using MIP and
LTNA, which suggests that a remarkable transition pores change
into micropores with poor connectivity in coal owing to the filling
of the minerals. Integrating the results of the total pore volume
or total porosity when using MIP, LTNA, LFNMR, and NMRC, the
differences in the different coal samples in the transition pores
and the micropores are mainly controlled by Ro,max and the burial
depth of the coal samples. Meanwhile, the ratios of the transition
pores in the summation of the transition pores and micropores
when using LFNMR are higher than those when using LTNA
and MIP, which is related to the destruction of the micropores
when applying MIP and LTNA under high pressure, particularly
for MIP. The pore fraction of the micropores when using NMRC
is clearly higher than that when using LTNA and LFNMR, which
suggests that remarkably disconnected micropores develop in the
coal samples. A comparison of Tables 2 and 10 shows that the
permeability of the coal samples is basically consistent with the
total porosity or pore volume, particularly based on the results
from LFNMR.
Based on the above analysis, affected by the connected macro-
pores and mesopores, the XY and CZ samples show a high pore
volume and porosity when using MIP (Table 4), whereas affected
by the total pore volume and the bulk density (Tables 2 and 5),
1886 H.H. Liu, I.I. Farid, S.X. Sang et al. / Energy Reports 6 (2020) 1876–1887
Table 9
Comparison of PSD using by different analysis method from SH sample.
Analysis
method
Pore diameter
range (nm)
Dominant pore diameter of
pore volume (nm)
Volume fraction of PSD (%)
Macropore Mesopore Transition pore Micropore
MIP >7.2 <100; >1000 8.99 2.86 26.53 61.62
LTNA <100 2∼60 58.99 41.01
LFNMR 10∼100,000 10∼60; >1000 7.57 16.11 76.21 0.11
NMRC 2∼580 30∼60 22.20 43.99 33.81
Table 10
Comparison of PSD for different samples using by different analysis methods.
Sample MIP LTNA LFNMR NMRC
V3+4V3/V3+4V4/V3+4V3+4V3/V3+4V4/V3+4ϕ3+4ϕ3/ϕ3+4ϕ4/ϕ3+4V3+4V3/V3+4V4/V3+4
SH 286.14 30.10 69.90 16.69 59.02 40.92 7.06 99.86 0.14 592.32 56.54 43.46
CZ 329.86 30.77 69.23 26.32 85.22 14.78 8.27 96.74 3.26 565.70 60.06 39.94
YW 300.86 36.59 63.41 19.38 80.49 19.51 1.64 63.41 36.59 326.94 59.38 40.62
XY 307.78 32.72 67.28 4.21 98.41 1.59 1.55 78.62 21.38 658.60 55.27 44.73
Notes: V3+4, total pore volume of transition pore and micropore, 10−4Ml/g; V3/ V3+4∼V4/ V3+4, means percentage from transition pore, and micropore in the total
pore volume of transition pore and micropore respectively, %; ϕ3+4, total porosity of transition pore and micropore, %; ϕ3/ϕ3+4∼ϕ4/ϕ3+4, means percentage from
transition pore, and micropore in the total pore volume of transition pore and micropore respectively, %.
the CZ sample shows the highest surface area of the pores, the XY
shows a relatively high surface area of the pores, the YW sample
has the lowest surface area of the pores when using MIP. In ad-
dition, the PSD in the coal samples when using MIP is controlled
by Ro,max and the coal quality, which is particularly based on the
content of the clay minerals. Affected by the transition pores,
the CZ sample shows the highest pore volume, the SH and YW
samples show a relatively high pore volume, and the XY sample
shows a low pore volume when using LTNA, which is controlled
by the contents of the minerals such as clay minerals, calcite, and
Ro,max; in addition, the BET surface area when using LTNA shows
good consistency with the volume of the micropores from all the
samples. The SH and CZ samples have a higher volume fraction
of the transition pores than the YW and XY samples and have
a lower volume fraction of the micropores than the YW and XY
samples when using LFNMR, which is affected by the growth of
the micropores with a smaller pore diameter and is difficult to
detect using LFNMR. Affected by the disconnected pores, the SH,
CZ, and XY samples have a relatively high pore volume, and the
YW sample has the lowest pore volume, when using NMRC. Ac-
cording to the comprehensive analysis described above, the PSD
of anthracite is better than that of semianthracite, as well as that
of the low volatile bituminous coal, for all coal samples selected;
in addition, the SH and CZ samples develop better interconnected
pores and worse disconnected pores, respectively, whereas the
XY sample develops better disconnected pores. Moreover, the dif-
ferences in the PSD for all the coal samples are comprehensively
controlled by Ro,max, the mineral content in the coal, and the
burial depth of the coal sample.
4. Conclusions
Based on a synergetic analysis of the PSD of coal reservoirs
from the SQB, China, using by MIP, LTNA, LFNMR, and NMRC, and
comprehensively disclosing the reason for the difference in the
PSD, the main conclusions were acquired as follows:
(1) Multimodality from all the samples occurs when using MIP,
and anthracite coal has a higher pore volume, porosity,
and surface area than semianthracite and low volatile bi-
tuminous coal. Affected by the total pore volume and the
bulk density, the surface area of the pores in the coal is
irrelevant to the types of coal.
(2) Most of the coal samples show the greatest amount of pore
volume when using LTNA once the pore diameter reaches
40 nm, and a bimodality is shown in all the samples; the
total pore volume is mainly derived from the contribution
of the transitional pores. The growth of the micropores and
the filling degree of the clay minerals affect the BET surface
area of all the samples. Moreover, for the cumulative pore
volume, the BET surface area from the anthracitic samples
is basically higher than that of the semianthracitic and low
volatile bituminous samples, which is affected by the filling
of the minerals.
(3) The total porosity of all the samples when using LFNMR is
mainly from the contribution of the interconnected pores
including the transition pores and the micropores, and
the total porosities of the mesopores and the macropores
from all the samples are extremely low; in particular, the
porosity of the mesopores is the lowest for all the samples.
Anthracite has a higher total porosity than semianthracite
coal and a lower total porosity than volatile bituminous
coal. Anthracite coal develops micropores with a smaller
pore diameter than semianthracite, and a low volatile bi-
tuminous coal develops. Such pores are difficult to detect
owing to the value of T2of LFNMR.
(4) Differences in the cumulative and incremental pore vol-
umes when using NMRC occur, and the total pore volume
is mainly due to the contribution of the transition pores
and micropores. The incremental pore volume shows a
multimodality, and the incremental pore volume reaches
a high peak value at different pore diameters. Numerous
disconnected pores develop in all coal the samples, and
the number of disconnected pores from the low volatile
bituminous coal is the highest, followed by anthracite and
semianthracite coal.
(5) Owing to the differences in the analysis principles and
the calculation models from the different analysis methods
applied, there are clear differences in the total pore volume
of the transition pores and micropores when using MIP,
LTNA, LFNMR, and NMRC. Regarding the poorly connected
micropores in all the coal samples, the connectivity of the
micropores in the anthracite coal is better than that in the
semianthracite and low volatile bituminous coal samples;
moreover, the difference in the PSD from the different coal
samples based on the different analysis methods is com-
prehensively controlled through Ro,max, the mineral content
in the coal, and the burial depth of the coal samples.
H.H. Liu, I.I. Farid, S.X. Sang et al. / Energy Reports 6 (2020) 1876–1887 1887
Affected by the comprehensive factors, the permeability of
the coal samples shows good consistency with the PSD.
Declaration of competing interest
The authors declare that they have no known competing finan-
cial interests or personal relationships that could have appeared
to influence the work reported in this paper.
Acknowledgments
This study was supported by the University Natural Science
Research Project of Anhui Province (KJ2019A0100), the National
Natural Science Foundation of China (Grant Nos. 41727801,
41302129), the Anhui Provincial Natural Science Foundation
(2008085MD121), and the Anhui postdoctoral research project
(2017B171).
References
AQSIO, S.A.C., 2008a. GB/T6948-2008 Method of Determining Microscopically the
Reflectance of Vitrinite in Coal. Standards Press of China, Beijing.
AQSIO, S.A.C., 2008b. GB/T 212-2008 Proximate Analysis of Coal. Standards Press
of China, Beijing.
AQSIO, S.A.C., 2008c. GB/T 16773-2008 Method of Preparing Coal Samples for
the Coal Petrographic Analysis. Standards Press of China, Beijing.
AQSIO, S.A.C., 2013. GB/T 8899–2013 Determination of Maceral Group
Composition and Minerals in Coal. Standards Press of China, Beijing.
Bustin, R.M., Bustin, A.M.M., Cui, X., Ross, D.J.K., 2008. Impact of shale properties
on pore structure and storage characteristics. In: 2008 SPE Shale Gas
Production Conference. Fort Worth, Texas. SPE 119892. 16–18 November.
Clarkson, C.R., Bustin, R.M., 1999. The effect of pore structure and gas pressure
upon the transport properties of coal: a laboratory and modeling study. 1.
Isotherms and pore volume distributions. Fuel 78, 1333–1344.
Clarkson, C.R, Jensen, J.L., Blasingame, T.A., 2011. Reservoir engineering for
unconventional gas reservoirs: what do we have to consider? In: SPE North
American Unconventional Gas Conference and Exhibition. Woodlands, Texas.
SPE 145080. 12–16 June.
Clarkson, C.R., Solano, N., Bustin, R.M., Bustin, A.M.M., Chalmers, G.R.L., He, L.,
Melnichenko, Y.B., Radliński, A.P., Blach, T.P., 2013. Pore structure charac-
terization of North American shale gas reservoirs using USANS/SANS, gas
adsorption, and mercury intrusion. Fuel 103, 606–616.
CNPC, 1996. SY/T 6189–1996. In: Quantitative Analysis Method of Rock Mineral
By Energy Dispersive Spectrometry (EDS) in SEM. Standards Press of China,
Beijing.
CNPC, 2014. SY/T 5162–2014. In: Analytical Method of Rock Sample By Scanning
Electron Microscope. Standards Press of China, Beijing.
Fu, H.J., Tang, D.Z., Xu, T., Tan, S., Yin, Z.Y., Chen, B.L., Zhang, C., Wang, L.L., 2017.
Characteristics of pore structure and fractural dimension of low-rank coal:
A case study of Lower Jurassic Xishayao coal in the southern Junggar Basin,
NW China. Fuel 193, 254–264.
Gane, P.A.C., Ridgway, C.J., Lehtinen, E., Valiullin, R., Furo, I., Schoelkopt, J.,
Paulapuro, H., Daicic, J., 2004. Comparison of NMR cryoporometry, mercury
intrusion porosimetry, and DSC thermoporosimetry in characterizing pore
size distributions of compressed finely ground calcium carbonate structures.
Ind. Eng. Chem. Res. 43 (24), 7920–7927.
Gregg, S.J., Sing, K.S.W., 1982. Adsorption, Surface Area, and Porosity, second ed.
Academic Press, New York.
Hodot, B.B., 1961. Outburst of Coal and Coalbed Gas (Chinese Translation).
Industry Press, Beijing.
ISO, 2007. 15901-2007. In: Pore Size Distribution and Porosity of Solid Materials
By Mercury Porosimetry and Gas Adsorption. ISO Copyright Office, Geneva.
ISO, 2016. 15901-1-2016. In: Evaluation of Pore Size Distribution and Porosity of
Solid Materials By Mercury Porosimetry and Gas Adsorption—Part 1: Mercury
Porosimetry. ISO Copyright Office, Geneva.
Jackson, C.L., McKenna, G.B., 1990. The melting behavior of organic materials
confined in porous solids. J. Chem. Phys. 93 (12), 9002–9011.
Kondrashova, D., Valiullin, R., 2013. Improving structural analysis of disordered
mesoporous materials using NMR cryoporometry. Micropor. Mesopor. Mat.
178, 15–19.
Li, Y., Zhang, C.M., Tang, D.Z., Gan, Q., 2017. Coal pore size distributions
controlled by the coalification process: An experimental study of coals from
the Junggar, Ordos and Qinhsui basins in China. Fuel 206, 352–363.
Liu, S.Q., Sang, S.X., Wang, G., Ma, J.S., Wang, X., Wang, W.F., Du, Y., Wang, T.,
2017. FIB-SEM and X-ray CT characterization of interconnected pores in high-
rank coal formed from regional metamorphism. J. Nat. Gas Sci. Eng. 148,
21–31.
Liu, H.H., Sang, S.X., Xue, J.H., Lan, T.H., Xu, H.J., Ren, B., Cheng, Q., 2018.
Experimental study on the velocity sensitivity of coal reservoir during
coalbed methane drainage in southern Qinshui Basin. Energy Explor. Exploit.
36 (6), 1593–1608.
Luo, C.J., Zhang, D.F., Lun, Z.M., Zhao, C.P., Wang, H.T., Pan, Z.J., Li, Y.H., Zhang, J.,
Jia, S.Q., 2019. Displacement behaviors of adsorbed coalbed methane on coals
by injection of SO2/CO2binary mixture. Fuel 247, 356–367.
Mastalerz, M., Drobniak, A., Stra˛poc’, D., Acosta, W.S., Ruppa, J., 2008. Variations
in pore characteristics in high volatile bituminous coals: implications for
coalbed gas content. Int. J. Coal Geol. 76, 205–216.
Mitchell, J., Webber, J., Strange, J., 2008. Nuclear magnetic resonance
cryoporometry. Phys. Rep. 46, 1–36.
NEA, 2016. GB/T6948-2008 Porosity and Permeability Measurement under
Overburden Pressure. Natural Energy Administration of China, Beijing.
Petrov, O.V., Furo, I., 2009. NMR cryoporometry: Principles, applications and
potential. Prog. Nucl. Magn. Reson. Spectrosc. 54 (2), 97–122.
Petrov, O.V., Furo, I., 2010. A joint use of melting and freezing data in NMR
cryoporometry. Micropor. Mesopor. Mat. 136, 83–91.
Shao, P., Wang, X., Song, Y., Li, Y., 2018. Study on the characteristics of matrix
compressibility and its influence factors for different rank coals. J. Nat. Gas
Sci. Eng. 56, 93–106.
Song, Y., Ma, X.Z., Liu, S.B., Jiang, L., Hong, F., Qin, Y., 2018. Accumulation
conditions and key technologies for exploration and development of Qinshui
coalbed methane field. Petrol. Res. 3, 320–335.
Wang, B., Sun, F.J., Tang, D.Z., Zhao, Y., Song, Z.H., Tao, Y., 2015. Hydrological
control rule on coalbed methane enrichment and high yield in FZ Block of
Qinshui Basin. Fuel 140, 568–577.
Wang, A.M., Wei, Y.C., Yuan, Y., Li, C.F., Li, Y., Cao, D.Y., 2017. Coalbed methane
reservoir’s pore-structure characterization of different macrolithotypes in the
southern Junggar Basin of Northwest China. Mar. Pet. Geol. 86, 675–688.
Wang, F.Y., Yang, K., Cai, J.C., 2018. Fractal characterization of tight oil reservoir
pore structure using nuclear magnetic resonance and mercury intrusion
porosimetry. Fractals 26 (2), 1840006.
Xu, X.K., Meng, Z.P., Wang, Y., 2019. Experimental comparisons of multiscale
pore structures between primary and disturbed coals and their efforts on
adsorption and seepage of coalbed methane. J. Nat. Gas Sci. Eng. 174,
704–715.
Yao, Y.B., Liu, D.M., 2012. Comparison of low-field NMR and mercury intrusion
porosimetry in characterizing pore size distributions of coals. Fuel 95,
152–158.
Yao, Y.B., Liu, D.M., Chen, Y., et al., 2010. Petrophysical characterization of coals
by low-field nuclear magnetic resonance (NMR). Fuel 89, 1371–1380.
Zhang, M., Fu, X.H., 2018. Characterization of pore structure and its impact
on methane adsorption capacity for semi-anthracite in Shizhuangnan Block,
Qinshui Basin. J. Nat. Gas Sci. Eng. 60, 49–52.
Zhang, Z., Qin, Y., Zhuang, X.G., Li, G.Q., Wang, X.M., 2018. Poroperm character-
istics of high-rank coals from southern Qinshui Basin by mercury intrusion,
SEM-EDS, nuclear magnetic resonance and relative permeability analysis. J.
Nat. Gas Sci. Eng. 51, 116–128.
Zhao, Y.X., Sun, Y.F., Liu, S.M., Wang, K., Jiang, Y.D., 2017. Pore structure
characterization of coal by NMR cryoporometry. Fuel 190, 359–369.
