Coal-forming sedimentary model and its control on vertical reservoir heterogeneity of the upper carboniferous Benxi formation in Ordos basin, China

Abstract

The Upper Carboniferous Benxi Formation in the Ordos Basin contains thick coal layers, showing good potential for deep coalbed methane exploration. In this study, we analyze the macroscopic coal type, macerals, industrial composition, and pore characteristics to determine the coal-forming environment and reservoir characteristics of deep coal in the Benxi Formation. Macroscopic Benxi coal types show strong vertical heterogeneity, with a lower layer comprising mainly semi-bright coal, overlain by bright coal, and capped with semi-bright coal and gangue. The ash content of the coal also exhibits strong vertical heterogeneity; from bottom to top, it increases, then decreases, then increases again in the uppermost layer. Vitrinite is the main maceral of Benxi coal, exhibiting the lowest concentration in the lower layer, and increasing in the vertical direction. Obvious vertical heterogeneity also exists in the pore characteristics of the coal, with the total pore volume, total specific surface area, and pore size all varying substantially; the upper layer has the highest porosity with most abundant macropores and micropores. Upper Carboniferous Benxi coal was mainly deposited in a saline, peat-swamp environment, under weak hydrodynamic conditions. The lower layer, controlled by fluctuation of sea level, was deposited in aerobic environment with a certain amount of hydrodynamic force, resulting in the formation of low-vitrinite, high-ash, low-porosity coal with strongly heterogenous characteristics. In contrast, the upper layer was deposited in a reducing, high-level swamp environment during the later stage of a retreating sea. During this stage, a bright coal layer with high vitrinite content, low ash content, and high porosity was deposited under stable conditions and weak hydrodynamic force. Therefore, the upper coal layer exhibits good reservoir properties, conducive to the storage and migration of coalbed methane, and is identified as the key layer for exploration and development of deep coalbed methane in the Ordos Basin. This research provides a theoretical basis for upcoming exploration and development of deep coalbed methane in the Benxi Formation, and provides a reference for future theoretical and technical research that will benefit deep coalbed methane development in other areas of the Ordos Basin.

Introduction

The Ordos Basin is a large-scale coal and natural gas production base in China. The basin is rich in natural gas resources and has great exploration potential. Deep coalbed methane in the Ordos Basin is characterized by the coexistence of conventional and unconventional reservoirs, symbiosis of free and adsorbed gas. As a type of unconventional natural gas, it has broad development prospects¹. According to a preliminary estimation, coalbed methane resources located at depths of ≥ 2000 m in China may total 40.71 trillion m³with recoverable resources of 10.01 trillion m³. Among these resources, those in the Ordos Basin may be as much as 12.99 trillion m³with recoverable resources of 3.08 trillion m³². Therefore, the Ordos Basin has become the top priority for achieving breakthroughs in deep coalbed methane exploration.

At present, the study of deep coalbed methane in the Ordos Basin is concentrated in four Upper Paleozoic coal-formed gas fields: Sulige, Wushenqi, Yulin, and Changdong (Mizhi), where major breakthroughs have been made in shallow coalbed methane exploration^3,4. The Benxi Formation has not seen commensurate breakthroughs; research progress has been slow, mainly focusing on the coal-forming environment and model^5,6,7,8reservoir-forming conditions^9,10,11and reservoir characteristics^12,13,14 at equivalent stratigraphic levels.

In terms of the study of sedimentary facies and coal-forming environment/model of the Benxi Formation in the Ordos Basin, different sedimentary model types are subdivided using differences in geochemical characteristics, microscopic characteristics, and coal petrological characteristics in horizontal direction. Prior studies have been carried out on the coal-facies characteristics, coal-forming environment, sedimentary model, pore characteristics, and storage properties of the Benxi coal in the Ordos Basin^5,6,8,11. Previous studies mainly focus on coal-forming and coalbed methane storage relationships in the horizontal direction across the Ordos Basin. The relationships between coal-facies and storage characteristics are then combined to derive the physical properties of the coal reservoirs. Ultimately, a planar sedimentary model of the Benxi coal is obtained.

The development of deep coalbed methane in the Benxi Formation is currently focused on some favorable intervals. Considering the strong heterogeneity of coal, there are evident differences in the vertical development of coal properties and reservoir characteristics, causing difficulties for the exploration and development of favorable target layers. However, there have been few studies on the vertical heterogeneity of coal in the Benxi Formation. Moreover, there remains a lack of systematic understanding and related research on the potential vertical sedimentary model and dominant storage section facilitated by the use of vertical changes in macerals, coal-facies characteristics, and pore characteristics.

Owing to the considerable resource potential of the Benxi coal seam of the Benxi Formation, lying at a buried depth of > 2000 m in the eastern Ordos Basin, this seam is expected to be an important base for the large-scale production of deep coalbed methane in China. Previously, Niu et al.^15,16 have carried out coal property analysis, reservoir characteristics analysis¹⁶, gas content analysis¹⁵ and other research on the 8 # coal seam of Benxi Formation in several wells in the central and eastern Ordos Basin. It is found that the upper layer of the 8 # coal seam reservoir has low ash yield, high vitrinite content and better pore development¹⁶. In view of the change of vertical gas content, it is found that the upper layer of the 8 # coal seam has the highest gas content, with an average value of 19.48 cm³/g^15,16, and the proportion of free gas has also increased¹⁵. This study believes that the upper layer is the dominant reservoir section for natural gas exploration and development of the 8 # coal seam reservoir of Benxi Formation in Ordos Basin. Therefore, in order to further study, we here studied the coal petrology and sedimentology of this coal seam using drill-core material obtained from the eastern Ordos Basin. We considered vertical heterogeneity, and studied the sedimentary environment of the coal seam using maceral and coal-facies characteristics, industrial composition, and total sulfur analyses. Through analysis of the controlling factors of pore development and pore distribution characteristics, the relationship between the vertical sedimentary environment and reservoir conditions and pore development of Carboniferous Benxi deep coal was clarified. The study also clarified the vertical heterogeneity characteristics and causes of carboniferous coal in benxi formation, as well as its dominant reservoir sections. This provides a theoretical basis for upcoming exploration and development of deep coalbed methane in the Benxi Formation, and provides a reference for future theoretical and technical research that will benefit deep coalbed methane development in other areas of the Ordos Basin.

Geological setting

The Ordos Basin spans the five provinces of Shaanxi, Gansu, Ningxia, Inner Mongolia, and Shanxi. It originates from the long-term stable development of a multi-cycle superimposed cratonic basin that has undergone stable subsidence, depression migration, and later torsion^17,18. It covers an area of approximately 250,000 km² and is the second largest basin on land in China^19,20,21. The Ordos Basin has experienced four evolutionary stages: Lower Paleozoic shallow marine platform, Upper Paleozoic offshore plain, Mesozoic intracontinental basin, and Cenozoic faulting and subsidence^22,23. It can be divided into six secondary tectonic units, including the Yimeng uplift, Western fold‒thrust belt, Tianhuan depression, Shanbei slope, Jinxi flexural fold belt, and Weibei uplift^24,25 (Fig. 1a). From the Archean onwards, the Ordos Basin has experienced multi-stage tectonic movements, including exposure to the Qianxi, Fuping, Wutai, Lyuliang, Jinning, Caledonian, Hercynian, Indosinian, Yanshan, and Himalayan orogenies^24,26,27.

Fig. 1

(a) Geological structure map of Ordos Basin. (b) Stratigraphic histogram of Q36 well in Benxi Formation 8 # coal seam of Ordos Basin. (Fig. 1a was created by Yu Li and was drawn using CorelDRAW 2020 (64-Bit) software. The version number of this software is v22.0.412 and the URL link is https://www.coreldrawchina.com.).

From Paleozoic to Mesozoic, the sedimentary environment of Ordos Basin is from marine facies to transitional facies, and then to continental facies²⁸. In the Late Carboniferous, the Ordos Basin sank as a whole, and sedimentation occurred on the unconformity surface at the top of the Ordovician. The predecessors further divided the Carboniferous Benxi Formation into Hutian, Pangou and Jinci member according to the lithology marker layer (Fig. 1b). The Jinci member mainly develops dark mudstone, limestone, sandstone and coal. In the early stage of Jinci member, the sea level rose slowly, forming the largest Jinci sandstone of Benxi Formation, and this stage was reformed by tide and storm. In the middle of Jinci member, the sea level entered a relatively stable high state, and the seawater inundated a wide range. In the late Jinci period, the sea level dropped rapidly, forming a paleogeomorphological pattern of high in the north and low in the south, and there was a short-term sedimentary hiatus. In the later period, with the slow rise of sea level, the tidal flat in the study area developed, the plants in the coastal marsh area were luxuriant, and the stable coal measure strata were formed²⁹. The coal-bearing strata at the top of Jinci section of Benxi Formation is a set of good coal-bearing source rocks in the whole area. The Benxi coal seam is the main coal seam in the Ordos Basin, with a thickness of 3–12 m and a burial depth of 1600–2800 m. The Benxi coal is mainly composed of bright coal and semi-bright coal with moderate metamorphism. It is one of the favorable strata for deep coalbed methane enrichment in China and has good development prospects^30,31.

The coal measure strata of Benxi Formation in Ordos Basin are widely developed in the central and eastern parts, with abundant gas reservoirs and great resource potential. From 2016 to 2020, the resource evaluation of natural gas resources in the Ordos Basin exceeded 11.04 × 10¹²m³. No.8 coal seam of Benxi Formation is the most favorable exploration target of coalbed methane in Benxi Formation. It has the characteristics of large thickness, wide distribution (area of about 16 × 10⁴m²) and continuous and stable development. The plane distribution area with buried depth of 2000–3500 m is about 6.9 × 10⁴km², which is a set of good coal-measure gas source rock²⁹.

Samples and experiments

Samples

The Carboniferous Benxi coal seam in the Ordos Basin was taken as the research object, and sampling was carried out in full core well of the Benxi Formation. According to differences in the macroscopic properties of the coal seam, 12 representative coal samples were collected from a depth range of 2801.80–2813.13 m in well Q36. The macroscopic characteristics of the samples were analyzed by hand; they were then roughly classified into two categories, bright and semi-bright, based on the overall relative luster intensity of each sample (Fig. 2). Most of the Benxi coal samples are bright coal, having a dark and shiny color. Banding is not evident, the structure is rather homogeneous, cracks are visible, and fractures are evident.

Fig. 2

Typical core samples of 8 # coal Q36 well in Benxi Formation, Ordos Basin. (a) The depth is 2803.50 m, it is semi-bright coal, the cross-section is irregular large block, the texture is light mainly composed of mirror coal and bright coal, its luster is very strong, shell-like ports, brittle and easy to crumble, the cut texture is developed; (b) The depth is 2804.60 m, it is in between semi-bright coal and bright coal. It is dominated by bright coal, the stripe structure is obvious, the face cut texture is developed, facing cuts with angular or stepped breaks; (c) At a depth of 2805.34 m, it is a bright coal, its cross-section is cut into three-dimensional blocks, and the face cuts are easily recognized; (d) At a depth of 2806.78 m, it is a bright coal, its cross-section is relatively smooth, the fissures can be seen to be filled by calcite, and many face cuts can be seen in the vertical direction; (e) At a depth of 2808.57 m, it is a semi-bright coal, its cross-section is relatively developed with obvious cut fissures, and 15 face cuts are developed/5 cm; (f) The depth is 2811.59 m, it is semi-bright coal, the face cuts are developed with 3 ~ 5 strips/5 cm; (g) The depth is 2811.87 m, it is bright coal, its cross section is in the form of three-dimensional granularity, the cuttings are developed with 12 ~ 13 strips/5 cm, and the physical properties of its reservoir are better; (h) The depth is 2812.60 m, it is semi-bright coal, its cross section is damaged and not easy to be recognized, the vertical visible cuttings are extremely developed with 5 ~ 10 strips/5 cm, and its size is bigger; (i) The depth is 2812.90 m, it is semi-bright coal, its face cuts are developed with 8 ~ 10 strips/5 cm, and vertical cuttings are obviously visible, its texture is lighter, and the physical properties of its reservoir are better.

Experimental methods

Industrial ingredients

According to “GB/T 212–2008: industrial analysis method of coal”, analysis of all samples (refer to the national standard GB/T 212–2001) was carried out to determine the ash, moisture, fixed carbon and volatile contents of the coal. Ash refers to the coal sample burned at 815℃±10℃, and the mass fraction of the residue in the coal sample is taken as the ash of the coal sample. Fixed carbon refers to the residue after removing moisture, ash and volatile matter from coal. Usually, the higher the fixed carbon content of coal, the higher the calorific value and the better the coal quality. The forms of sulfur were determined according to “GB/T 215–1996: determination of different forms of sulfur in coal”. Sulfate sulfur was obtained by determining the mass of barium sulfate in the coal; iron-sulfide sulfur was determined using the oxidative method; and organic sulfur was obtained by calculating total sulfur and deducting sulfate sulfur and iron-sulfide sulfur.

Micro-component analysis

Micro-components of coal are the constituent units of coal that can be recognized under the microscope. In this paper, the macerals of coal samples are classified according to the international standard “ICCP System 1994”³². According to differences in origin, morphology, and physical properties, the micro-components were classified into three main groups: vitrinite, inertinite, and liptinite³². DM4500P polarizing microscope electron and number point method was used. The 50-fold oil-immersed objective lens was used to determine the maceral content by the number point method. A total of more than 300 effective points were recorded. Through the observation of electron microscope, the organic macerals and coal quality characteristics of Benxi coal sample were analyzed and identified, and the contents of vitrinite, inertinite and other components of the studied coal sample were analyzed.

Coal-facies parameters

$$\:\text{G}\text{I}=\frac{\text{V}\text{i}\text{t}\text{r}\text{i}\text{n}\text{i}\text{t}\text{e}+\text{M}\text{a}\text{c}\text{r}\text{i}\text{n}\text{i}\text{t}\text{e}}{\text{F}\text{u}\text{s}\text{i}\text{n}\text{i}\text{t}\text{e}+\text{S}\text{e}\text{m}\text{i}\text{f}\text{u}\text{s}\text{i}\text{n}\text{i}\text{t}\text{e}+\text{I}\text{n}\text{e}\text{r}\text{t}\text{o}\text{d}\text{e}\text{t}\text{r}\text{i}\text{n}\text{i}\text{t}\text{e}}$$

(1)

The gelification index (GI) refers to the ratio of gelatinized fraction to non-gelatinized fraction, and reflects the degree and duration of wetting of peat bogs during peat accumulation³³.

$$\:\text{T}\text{P}\text{I}=\frac{Telinite+Collotelinite+Fusinite+Semifusinite}{Collodetrinite+Macrinite+Inertodetrinite}$$

(2)

The tissue preservation index (TPI) is used to indicate the intensity of degradation of plant tissues and degree of preservation of cells of peat-forming plants; it can also be used to trace changes in pH.

$$\:\text{V}\text{I}=\frac{Telinite+Collotelinite+Corpogelinite+Fusinite+Semifusinite+Resinite+Suberinite}{Collodetrinite+Vitrodetrinite+Inertodetrinite+Liptodetrinite+Sporinite+Cutinite}$$

(3)

The vegetation index (VI) is the ratio of components of preserved cellular structure to matrix, fragmentary material, and granular components; it mainly reflects the type of coal-forming vegetation and its preservation degree.

$$\:\text{G}\text{W}\text{I}=\frac{Gelinite+Corpogelinite+Vitrodetrinite+Minerals}{Telinite+Collotelinite+Collodetrinite}$$

(4)

The groundwater flow index (GWI) is an indicator based on gelatinization and mineral inputs, and represents the water level at the time of peat-bog formation³⁴.

$$\:\frac{\text{F}}{\text{M}}=\frac{Telinite+Collotelinite+Fusinite+Semifusinite}{Vitrodetrinite+Inertodetrinite+Macrinite+Collodetrinite}$$

(5)

The framework matrix ratio (F/M) is the ratio of skeletal components to matrix components, mainly reflecting the degree of destruction of plant cells and strength of water flow activity in a peat bog³⁵.

$$\:\frac{\text{V}}{\text{I}}=\frac{Vitrinite}{Inertinite}$$

(6)

The ratio of vitrinite to inertinite (V/I) reflects the degree of water cover of the marsh and how wet or dry the coal-forming climatic conditions were³⁶.

$$\:\text{O}\text{I}=\frac{Macrinite+Inertodetrinite}{Other\:microscopic\:components\:oxidation\:index}$$

(7)

The oxidation index (OI) reflects the degree of oxidation or reduction of a coal-forming peat bog³⁷.

Scanning electron microscope (SEM) analysis

Scanning electron microscopy involves the excitement of electronic signals of different functions point-by-point on the surface of a tested sample using a finely focused electron beam; the received electrical signals are amplified by a detector and an image is then formed enabling information to be obtained regarding the surface morphological characteristics of a sample. Here, we used a field-emission-SEM to analyze the coal samples. It is a magnification technique capable of magnification ranging between 10× and 300,000× revealing the information about elements present and their topographical data on the surface through unlimited field depth virtually. FE-SEM is typically performed in a high vacuum because gas molecules tend to disturb the electron beam and the emitted secondary and backscattered electrons used for imaging. It was operated in secondary electron mode, with an accelerating voltage of 3 kV and analysis time of 3 h. The SEM images enabled the characteristics of pores and cracks, including their sizes, shapes, connectivity, and surface morphology, to be analyzed using various image analysis techniques.

Gas adsorption experiment

Gas adsorption is the main effective method for analyzing micropores and macropores in coal. The gas adsorption test was performed using an AutosorbiQ-MP automatic gas adsorption analyzer according to the national standards GB/T 21650.2–2008 and GB/T 21650.3–2011. Gas adsorption experiments include low-temperature liquid N₂ and low-pressure CO₂ adsorption experiments. Liquid N₂ adsorption experiments are carried out at liquid nitrogen temperature of 77.36 K. When the adsorption‒desorption amount reaches equilibrium, the function of adsorption amount and relative pressure p/p₀ can be obtained. The adsorption curve can be obtained by the correlation between the adsorption amount and the relative pressure, and the effective data such as the specific surface area, pore volume, and distribution of pore sizes can be further obtained. Specific surface areas (SSAs) were calculated based on Brunauer–Emmet–Teller (BET) theory. Pore volumes (PVs) and pore size distributions (PSDs) for pore diameters from 1.7 to 250 nm were calculated based on Barrett–Joyner–Halenda (BJH) model. Carbon dioxide adsorption is defined as CO₂ adsorption at 273.15 K. The gas adsorption at different relative pressures is determined using CO₂ as the adsorbent; after completion of the test, density functional theory (DFT) and Dubinin-Radushkevich (DR) equation can be used to process the data to obtain the specific surface area, pore volume, pore size distribution, and other related information concerning the micropores of interest^38,39.

The high-pressure mercuric pressure test was carried out using a PoreMaster Hg porosimeter from Kantar, with a pressure range of 20‒413 MPa. Pore sizes can be calculated according to the pressure using the Washburn equation. Because Hg is non-wetting, pressurization causes it to enter larger pores. The Hg curve was then obtained from the Hg porosimeter by recording the interaction between the pressure and volume of the Hg during pressurization; then the parameters such as pore volume, specific surface area and distribution of pore sizes are calculated by using Washburn equation and geometric model hypothesis.

Results

Industrial ingredients

Industrial analysis, also known as coal-quality analysis, is the basic means of testing coal quality. The industrial analysis of coal mainly includes the determination of moisture, ash, volatile matter, sulfur, and calculation of fixed carbon within coal. According to these data, the type, basic properties, and industrial use of coal can be preliminarily judged^40,41,42; these data are shown in Table 1; Fig. 3.

Table 1 Coal petrographic analysis, macerals, industrial composition and total sulfur determination results of carboniferous Benxi formation in well Q36.

Fig. 3

Comprehensive histogram of 8 # coal Q36 well in Benxi Formation of Ordos Basin. It can be found that with the change of depth, the lower layer shows low porosity, high inorganic sulfur, low vitrinite and high ash content, and has strong heterogeneity characteristics; the upper layer shows high porosity, high organic sulfur, high vitrinite and low ash content.

Macroscopic coal properties

According to the relative gloss intensity of each layer and relative content of bright lithotypes⁴³two types of samples, namely semi-bright coal and bright coal, were identified in coal seams of well Q36 in the Benxi Formation, with a higher proportion of bright coal compared with semi-bright coal. The macroscopic coal components of bright coal were mainly vitrain and clarain, banding was not generally evident, the structure was approximately uniform, endogenous cracks were developed, and conchoidal fractures were frequently observed. The macroscopic coal components of semi-bright coal were mainly clarain, including vitrain and durain bands, with banded and layered structures, and conchoidal fractures.

Industrial composition

Coal ash refers to the residue remaining after complete combustion under certain conditions (805‒825 ℃). The ash yield of Well Q36 in the Benxi Formation was between 2.68% and 45.51%, with an average value of 17.60%. In general, this is characteristic of low-ash coal. From the bottom to top, the ash content first decreased and then increased. The fixed carbon in coal refers to the solid material produced by the pyrolysis of organic matter under certain heating conditions, which belongs to a part of the coke slag⁴⁴. The fixed carbon content in the samples ranged from 33.58 to 88.62%, with an average value of 72.26%. In general, this is characteristic of high-fixed-carbon coal. From the well bottom to top, the fixed carbon content first increased and then decreased, a pattern opposite to that of the ash content (Table 1).

Total sulfur analysis

Total sulfur (St) can be divided into organic sulfur (So) and inorganic sulfur. The St content is an important indicator in evaluating the characteristics of coal quality. Inorganic sulfur is divided into two categories: sulfur in iron sulfide (Sp) and sulfur in sulfate (Ss). The St content of the samples ranged between 0.33% and 6.05%, with an average of 4.56%. The So content ranged between 0.3% and 5.46%, with an average of 3.83%. The Sp content ranged between 0.01% and 1.85%, with an average of 0.43%, and the Ss content ranged between 0.02% and 1.41%, with an average of 0.30%. Hence, coal seam of the Benxi Formation can be classed as high-sulfur coal.

Microscopic composition

Under the optical microscope, it can identify the organic macerals and mineral matter of coal. And organic macerals are composed of vitrinite, inertinite, and liptinite^40,41,42,45. The organic maceral content of the samples ranged from 68.67 to 99.33%, with an average of 88.03%. These macerals were mainly vitrinite (50.33%‒95.00%; average 75.58%), mostly comprising telinite and vitrodetrinite, followed by gelovitrinite. The telinite content was 19.30%‒91.70%, with an average of 52.4%; the vitrodetrinite content was 0.70%‒63.70%, with an average of 20.20%; and the content of gel specular debris was < 9.30%, with an average of 2.90%. The inertinite content ranged from 2.67 to 32.33%, with an average of 12.39%. This was dominated by detrital inertinite (2.70%‒10.00%; average 6.60%), followed by semifusinite (< 14.70%; average 3.00%). The content of fusinite was < 6.70%, with an average of 1.90%, that of macrinite was < 3.30%, with an average of 0.60%, and that of micrinite was < 1.70%, with an average of 0.20%. The lipid content was the lowest of all the macerals at < 0.33%, with an average of 0.06%, mainly comprising resinous bodies. The mineral content ranged between 0.67% and 31.33%, with an average of 11.97%, and mainly comprised pyrite and clay, followed by carbonates, and a small amount of other sulfides (Fig. 4; Tables 1 and 2). The remainder of each sample could not be observed under the microscope, and was ignored.

Fig. 4

Photos of maceral of Benxi Formation 8 # coal sample. (a) The maceral photos of coal samples with a depth of 2803.10 m. Under the oil-immersed reflected light, the content of vitrinite is more, the vitrodetrinite can be seen, and there is also a small amount of inertodetrinite; (b) The maceral photos of coal samples with a depth of 2803.50 m. Under oil-immersed reflected light, the vitrinite content is very high, and most of them are collotelinlte; (c) The maceral photos of coal samples with a depth of 2804.60 m. Under oil-immersed reflected light, the vitrinite content is also very high, and most of them are collophanite, and also contain a small amount of inertodetrinite; (d) The maceral photos of coal samples with a depth of 2805.34 m. Under oil-immersed reflected light, the vitrinite content is also very high, and most of them are collophanite; (e) The maceral photos of coal samples with a depth of 2806.78 m. Under oil-immersed reflected light, the vitrinite content is also very high, and contains a small amount of inertodetrinite; (f) The maceral photos of coal samples with a depth of 2807.62 m. Under oil-immersed reflected light, the vitrinite content is also relatively high, but it is reduced. The content of inertinite increased, among which the content of inertodetrinite, fusinite and semifusinite increased; (g) The maceral photos of coal samples with a depth of 2808.57 m. Under oil-immersed reflected light, the vitrinite content is further reduced compared with the sample at 2807.62 m, and the inertinite content is increased. The content of fusinite and semifusinite increased; (h) The maceral photos of coal samples with a depth of 2810.09 m. Under oil-immersed reflected light, the vitrinite content increases compared with the sample at 2808.57 m, and the pores are mostly filled with clay minerals; (i) The maceral photos of coal samples with a depth of 2811.59 m. Under oil-immersed reflected light, the vitrinite content is reduced compared with the sample at 2810.09 m, the pores are mostly filled with clay minerals, and a small amount of pyrite appears; (j) The maceral photos of coal samples with a depth of 2811.87 m. Under oil-immersed reflected light, a large number of fusinite and semifusinite can be seen. And there are more clay minerals; (k) The maceral photos of coal samples with a depth of 2812.60 m. Under oil-immersed reflected light, the vitrinite content is relatively low, most of the pores are filled with clay minerals, and a large amount of pyrite can be seen; (l) The maceral photos of coal samples with a depth of 2812.90 m. Under oil-immersed reflected light, the content of vitrinite is also relatively low, a large number of fusinite and semifusinite bodies can be seen, and a large amount of pyrite can be seen.

Table 2 Coal rock macerals and coal facies parameters of Benxi coal in well Q36.

Pore size distribution

Reservoir pore structure generally refers to the geometry, size, distribution, and interconnection of pores and throats, important factors affecting reservoir physical properties. In the study of coal reservoir pores and fractures, IUPAC(1972) classification scheme is usually adopted internationally. Pores are divided into micropores (< 2 nm), mesopores (2–50 nm), and macropores (> 50 nm) according to their radius. The low-pressure CO₂ adsorption method is used to describe the pore distribution characteristics of micropores; the low-pressure N₂ adsorption/desorption method is used to describe the pore distribution characteristics of mesopores; and the Hg intrusion method is used to describe the pore distribution characteristics of macropores. By analyzing the Hg injection‒demercuration curve, N₂ adsorption‒desorption curve, and CO₂ adsorption curve of a Hg intrusion experiment, the morphology of pores in coal can be studied and the factors affecting pore structure can be discussed (Fig. 5; Table 3).

Fig. 5

CO₂ adsorption (a), N₂ adsorption and desorption (b), mercury adsorption (c) isotherms of coal in Qi 36 well, 8 # coal, Benxi Formation, Ordos Basin. (a) The low-pressure CO₂ adsorption method is used to describe the pore distribution characteristics of micropores. From the figure, it can be found that in general, the samples with high ash content have less CO₂ adsorption; the samples with low ash content have more CO₂ adsorption. (b) The low-pressure N₂ adsorption/desorption method is used to describe the pore distribution characteristics of mesopores. From the figure, it can be found that in general, the samples with high ash content have more amount of N₂ adsorption and desorption; the samples with low ash content have less amount of N₂ adsorption and desorption. (c) The mercury intrusion method is used to describe the pore distribution characteristics of macropores. From the figure, it can be found that in general, the samples with high ash content have more amount of Hg intrusion and removal; the samples with low ash content have less amount of Hg intrusion and removal.

Table 3 The results of pore volume and pore specific surface area for Benxi coal.

From Fig. 5, a hysteresis loop can be observed, and the type of coal pores can be preliminarily determined. The CO₂ adsorption capacity ranged from 2.98 to 20.20 cm³/g, with an average of 16.65 cm³/g. The CO₂ adsorption capacity of all samples exceeded 13 cm³/g, and the average adsorption capacity was 17.89 cm³/g. The adsorption capacity of N₂ ranged from 0.06 to 1.86 cm³/g, with an average of 0.59 cm³/g and large overall range. The cumulative amount of Hg intrusion ranged from 0.01 to 0.05 cm³/g, with an average of 0.03 cm³/g and small overall range (Table 3). It can be seen from Fig. 5 that the ash content had a certain influence on the amount of Hg intrusion and removal, N₂ adsorption and desorption, and CO₂ adsorption. For samples with high ash contents, the amount of CO₂ adsorption was small, amount of N₂ adsorption and desorption was large, and amount of Hg intrusion and removal was large. For samples with low ash contents, the amount of CO₂ adsorption increased, amount of N₂ adsorption and desorption decreased, and amount of Hg intrusion and removal decreased (Fig. 5).

On the basis of the aforementioned data, the pore distribution characteristics of coal macropores, mesopores, and micropores (Table 3) were obtained. The porosity of the samples ranged from 3.44 to 9.12%, with an average of 6.25% and large overall range. The total pore volume of the samples ranged from 0.0069 to 0.0483 cm³/g, again with a large range of variation, and an average of 0.0354 cm³/g. The total specific surface area of the samples ranged from 23.55 to 152.36 cm³/g, again with a large range of variation, and an average of 123.20 cm³/g. It can be seen from Table 3 that the porosity, specific surface area, and pore volume of the gangue samples at a depth of 2803.1 m were the lowest, while those of bright coal samples at a depth of 2807.62 m were the highest. Micropores comprised the highest proportion of pore volume, followed by macropores, then mesopores. The proportion of micropores ranged from 57.77 to 87.85%, with an average of 72.81%; the proportion of macropores ranged from 9.52 to 40.99%, with an average of 23.17%; and the proportion of mesopores ranged from 0.68 to 30.35%, with an average of 4.01%. The proportion of pore specific surface area was dominated by micropores, again followed by macropores, then mesopores. The specific surface area of micropores ranged from 22.02 to 144.63 cm²/g, with an average of 119.02 cm²/g, and the proportion ranged from 93.49 to 97.86%, with an average of 96.42%. The specific surface area of macropores ranged from 0.99 to 8.89 cm²/g, with an average of 4.00 cm²/g, and the proportion ranged from 2.02 to 6.15%, with an average of 3.28%. Finally, the mesoporous specific surface area ranged from 0.06 to 0.54 cm²/g, with an average of 0.18 cm²/g, and the proportion ranged from 0.04 to 2.30%, with an average of 0.30%.

The average pore size of the samples ranged between 13.97 and 228.39 nm, with an overall average of 44.61 nm. The sample obtained from a depth of 2803.1 m had the smallest pore size, and compared with other samples, had a higher content of mesopores, and lower content of micropores and macropores. The sample obtained from a depth of 2807.62 m had the largest pore size, far exceeding the pore size of the other samples. The pore size range of all other samples ranged between 20 and 50 nm (Table 3). According to a histogram of pore diameter and pore volume (Fig. 6), the main pore volume of the coal samples originated from micropores, followed by macropores; there were very few mesopores observed in the samples. The sample from a depth of 2803.1 m was abnormal, having more mesopores and smaller micropores.

Fig. 6

The proportion of pores with different pore sizes in the pore volume of some coal samples from Well Qi 36, Benxi Formation, Ordos Basin.

Discussion

Influence of vertical sedimentary environment on the formation of the Benxi coal seam

Coal-rock sedimentary facies of the Benxi coal seam

Qualitative and quantitative maceral results are shown in Table 1, where it can be seen that coal from the Benxi coal seam is mainly composed of vitrinite (75.58%), with less amounts of inertinite (12.39%) and liptinite (0.06%)^32,40,41,42 (Fig. 7), and also contains small amounts of clay minerals and pyrite. The 12 coal samples in this seam are mainly bright and semi-bright coal, and the Ro is about 1.4%. Therefore, according to the international coal classification standard ISO11760 (2005), due to 0.5% < Ro < 2%, the coal rank of the coal samples in our study is medium rank bituminous coal. In addition, according to Table 1, the volatile content of the 12 coal samples is mostly about 14%, and the volatile content of the top coal sample is relatively high. Therefore, with reference to “ASTM D388-23: Standard Classification of Coals by Rank”, they also can be classified as medium-low volatile bituminous coal. Because vitrinite mainly results from the gelatinization of plant tissues in a water-covered reducing environment, it indicates a relatively stable water level and reducing environment for coal formation. In contrast, inertinite is formed by fungal decomposition and wildfire burning of similar plant tissue under dry conditions, reflecting a lower water level and oxidizing environment^46,47,48. Therefore, a higher vitrinite content and lower inertinite content indicates that the sedimentary environment of the Benxi coal seam was a partially reducing water-covered peat swamp.

Fig. 7

Triangle diagram of macerals of 8 # coal. It is found from the figure that there is almost no liptinite.

Moreover, the use of coal petrological parameters (e.g., GI, TPI, VI, and GWI) can provide an enhanced understanding of the overall sedimentary environment of seam Benxi coal^49,50. Diessel⁵¹ first proposed the GI‒TPI coal facies diagram, and the use of GI and TPI to determine coal-forming conditions. He interpreted the GI to be the ratio of the gelation component to the non-gelation component, reflecting the dryness or humidity of the coal-forming environment; that is, the change in water level in the coal-forming swamp and degree of gelation of plant remains. A higher GI value equates to a more humid peat-forming environment and deeper overlying water, whereas a drier peat-formation environment results in a lower GI value. The TPI is used to reflect the degree of transport, degradation, and oxidation of plant debris or cells and the degree of preservation of plant cell structure. A low TPI indicates intense degradation and a poorly preserved structure, whereas a high TPI indicates a low intensity of degradation and well preserved plant cells⁵². In this paper, a GI‒TPI facies diagram (Fig. 8a) was established according to these GI and TPI parameters. From this diagram, it can be observed that all coal sample parameter points essentially plotted in open-water swamp facies (e.g., tidal marsh), except for sample Q36-9 from a depth of 2811.59 m, which plotted in the moist forest swamp facies. Coal in the upper layer formed under deeper water, while coal in the middle and lower layers formed under shallower water. The overall GI is 1‒10, indicating a high gelation index for this coal seam, along with a swamp-like sedimentary environment characterized by deep overlying water, weak hydrodynamic conditions, and low plant accumulation rate. The TPI is mostly < 1, indicating that the plant cell structure has been destroyed to a certain extent, the degree of gelation is high, and the reducing power of the sedimentary environment was high.

Fig. 8

Coal facies diagram of 8 # coal in Benxi Formation, Ordos Basin. (a) GI-TPI coal facie diagram; (b) VI-GWI coal facie diagram.

The VI and GWI, defined by Calder et al.⁵³are derived by adding stable macerals and minerals to the TPI/GI ratio proposed by Diessel⁵¹and reflect the dynamic groundwater characteristics of the original coal-forming environment. The VI reflects the preservation degree of coal-forming vegetation and its affinity with herbaceous or woody plants; a VI of < 1 indicates that the coal-forming plants were herbaceous or aquatic plants, whereas a VI > 1 indicates an affinity with woody plants³⁵. The GWI is used to indicate changes in groundwater level, reflecting the degree of control of groundwater on the original coal-forming environment and its mineral content. A higher GWI equates to stronger degradation and higher mineral input of coal macerals^34,53and stronger dynamic groundwater conditions. From the GWI‒VI facies diagram shown in Fig. 8 (b), it can be seen that the parameter values are essentially located in two areas: water-covered swamp and marsh swamp. The VI values of most samples are > 1, indicating that the coal-forming plants were mainly woody plants, and the GWI values are mostly < 1, indicating generally weak hydrodynamic conditions³⁵.

In summary, the environment of the coal-forming period corresponding to the Benxi coal seam was a water-covered moist peat swamp with generally weak hydrodynamic conditions. However, through the study of vertical changes in macroscopic coal characteristics, maceral content, and coal-facies characteristics, it is evident that there are certain differences in the sedimentary environment recorded with changing depth in the coal seam.

Changes in sedimentary environment indicated by changing ash content

Coal ash is composed of oxides and salts of various elements such as Si, Al, and Fe. During the process of precipitation or migration in peat swamps, these elements are affected by factors such as water oxidation state, pH, hydrodynamic conditions, and salinity. Therefore, the composition of coal ash can reflect certain changes in the coal-forming environment⁵⁴. From Fig. 3, it can be seen that the ash content gradually decreases from bottom to top within the drill core, and there is also a trend of high values at both ends of the core and low values in the middle. The ash content of the lower layer is higher than that of the upper layer, and shows no obvious change. The ash index is expressed as (SiO₂ + Al₂O₃)/(Fe₂O₃ + CaO + MgO); a weakly reducing environment is dominated by SiO₂ and Al₂O₃, and a strongly reducing environment is dominated by Fe₂O₃, CaO, and MgO; hence, the ash index is inversely proportional to reducibility⁵⁵. From the data presented herein, it is evident that the lower layer and very top of the core mainly correspond to an oxidizing environment, with strong hydrodynamic force, bringing large amounts of terrestrial debris, resulting in a higher ash content. In contrast, the upper layer mainly corresponds to a reducing environment, with relatively stable water, resulting in a low ash content.

Changes in sedimentary environment indicated by changing sulfur content

Total sulfur in coal seams includes organic sulfur and inorganic sulfur. Organic sulfur mainly comes from coal-forming plants, and inputs of inorganic sulfur are mainly controlled by the aqueous environment during deposition. Usually, when the reducibility of a peat swamp is strong, sulfur-rich coal forms⁵⁴. The sulfur in the samples obtained herein was mainly organic sulfur (Fig. 3), indicating that the depositional area comprised a partially reducing environment. In addition, in the lower layer of the coal seam, the inorganic sulfur content was high and organic sulfur content was low, indicating that strong hydrodynamic force and an oxidizing aqueous environment. In contrast, in the upper layer of the coal seam, the inorganic sulfur content was low, and organic sulfur content was high, indicating weak hydrodynamic force, and a generally reducing aqueous environment.

Changes in sedimentary environment indicated by changing coal petrological characteristics

Coal type is the most direct material manifestation of coal sedimentary facies. Different coal types can reflect different geochemical, hydrodynamic, water-cover, and burial conditions in peat swamps during coal-forming periods. The Benxi coal seam formed in a wet peat swamp environment with overlying water. In the early stage of coal formation, organic matter was deposited and then overlain with water. Therefore, a reducing‒weakly oxidizing environment existed, leading to good gelation, high vitrinite content, and low inertinite content; thus, Benxi coal mainly comprises bright coal and semi-bright coal (Fig. 2). According to observed changes with depth, the lower layer of the Benxi coal seam is mainly semi-bright coal, containing a thin layer of bright coal and argillaceous bands, and the upper layer is mainly bright coal, with the topmost layer being semi-bright coal. These changes are mainly a result of the ancient geographical environment. In the early and latest stages of peat formation, the deposition of organic matter occurred under oxidizing conditions, the GI value was low (Table 2), and the degree of gelation was insufficient. In the middle stage of peat formation, the swamp conditions changed, and the deposition of organic matter occurred under highly reducing conditions and high levels of water cover, leading to a high degree of gelation (Table 2) and high vitrinite content. Therefore, the upper coal seam is dominated by bright coal.

Microscopic coal components are good indicators of the coal-forming environment. In general, plant debris in a humid, reducing environment mainly generates vitrinite via the gelatinization process, while in a dry, oxidizing environment, it mainly generates inertinite by the fusinization process⁵⁴. Through the analysis of macerals within Benxi coal (Figs. 4 and 9; Table 1), it can be seen that in the lower coal seam, the vitrinite content is relatively low, inertinite content is relatively high, and maceral content changes greatly within this lower layer. This indicates that the hydrodynamic force was strong, sedimentary environment was unstable, gelation was weak, and environment was oxidizing during the development of this part of the coal seam. In the upper coal seam, the vitrinite content is relatively high, inertinite content is relatively low, and the maceral content changes little within this upper layer. This shows that the period is the oxidation environment of water retention, the condition of overlying water is better, and the gelatinization is better. The content of vitrinite in the topmost layer of the coal seam decreased again, while the content of inertinite increased, indicating that the environment changed from reducing to oxidizing at this time, and the transgression occurred, accompanied by enhanced hydrodynamic force.

Fig. 9

Vertical distribution sequence of coal-forming environment indicator parameters of 8 # coal in Benxi Formation. The indication parameters mainly include vitrinite, inertinite, F/M, OI and V/I.

In addition, the V/I proposed by Harvey et al.³⁶ can also intuitively reflect the degree of water cover of coal peat swamps, and dry versus wet climate conditions. In an environment of high temperature and partial oxidation, fusinization is dominant, and the content of inertinite in coal macerals increases. In a warm and humid, partially reducing environment, gelatinization is dominant, and the vitrinite content increases. According to V/I values, coal-forming environments can be divided into four types: class I: V/I > 4, representing strong overlying deep water; class II: 1 < V/I ≤ 4, representing extremely humid-covered water; class III: 0.25 < V/I ≤ 1, representing wet-weak water cover; class IV: V/I ≤ 0.25, indicating a dry—high dry environment, where there have been fires. The V/I value gradually increases from the bottom to top of the Benxi coal seam in the study area (Fig. 9). In the lower layer, the V/I value is low, degree of water coverage is low, with some exposure to an oxidizing environment; the V/I value of the upper layer is much larger than 4, so the upper layer was completely covered with water, and the coal-forming environment was reducing. Therefore, the coal seam presents a pattern of oxidation‒reduction‒oxidation from bottom to top.

The OI value can also reflect the redox conditions of coal peat, wherein a higher OI value equates to a stronger degree of oxidation. In general, an OI > 1 represents an oxidizing environment, and an OI < 1 represents a reducing environment³⁷. According to the data presented herein (Fig. 9; Table 2), the OI value is < 1 throughout the Benxi coal seam, indicating a partially reducing environment, supporting the overall sedimentary environment described above. However, there is a certain fluctuation from bottom to top through the coal seam. The OI values of the lower layer are higher, representing a partially oxidized environment, and the fluctuation is relatively large, indicating strong hydrodynamic force and substantial changes in water coverage. The OI values of the upper layer are low, indicating a partially reducing environment. The OI value then increases slightly at the very top of the coal seam, indicating a return to a partially oxidizing environment.

The ratio of skeletal components to matrix components and detrital components in coal-forming plants (F/M) can reflect the fluidity of swamp water and degree of plant fragmentation^35,37. An F/M value > 1 indicates a stagnant environment with weak water flow activity, whereas an F/M value < 1 indicates an active aqueous environment with strong water flow. The F/M values of the lower layer of the Benxi coal seam are < 1, indicating that the sedimentary environment was an active aqueous environment with strong water flow, accompanied by the input of terrigenous clastic material; hence, the ash yield of this layer is relatively high (Fig. 3). In contrast, the F/M values in the upper layer are much larger than 1, indicating that the sedimentary environment had stagnant water with weak hydrodynamic conditions, and the water body was relatively stable. The F/M value at the very top of the coal seam is < 1, marking a return to an active aqueous environment, with accompanying higher ash yield (Fig. 9).

In summary, Benxi coal of the Benxi Formation in the Ordos Basin is a peat-swamp sedimentary facies, formed under partially reducing overlying water and weak hydrodynamic conditions. In the vertical direction, owing to the influence of sea-level change, the ash content gradually decreases from bottom to top, and the content of vitrinite gradually increases. This indicates an oxidizing aqueous environment with strong hydrodynamic conditions during deposition of the lower layer of the coal seam, and a reducing waterlogged environment during deposition of the upper layer of the coal seam. The very top layer of the coal seam marks a return to oxidizing conditions.

Reservoir development characteristics and controlling factors

Influencing factors of coal reservoir properties

The reservoir space and properties of coal have an important influence on the occurrence of coalbed methane. Therefore, it is of great significance to study the influencing factors of reservoir properties.

Ash content and fixed carbon content have an effect on the storage properties of coal. According to Fig. 10, ash content shows a significant negative correlation with porosity, while fixed carbon content shows a significant positive correlation with porosity. The main reason behind this is that an increase in ash content reflects large amounts of supplementary terrestrial debris, which continuously fills the primary pores of the coal and increases the mineral content of the coal. Hence, the continuity of pores can be destroyed, and the total pore volume of the coal can be reduced. Therefore, ash content and fixed carbon content can affect the reservoir properties of coal.

Fig. 10

Correlation between ash content, fixed carbon content and porosity. (a) Correlation between ash and porosity; (b) Correlation between fixed carbon and porosity.

It has been confirmed by the data presented herein (Fig. 3) that the coal ash yield in different parts of the same coal seam changes substantially in the vertical direction. The ash content in the lower layer is higher than that in the upper layer, which indirectly indicates that the pores in the upper layer are well developed. Therefore, the upper layer is conducive to the storage and migration of coalbed methane and performs better as a reservoir.

In addition to the influence of ash and fixed carbon contents on reservoir performance, the sedimentary environment also exerts a certain control on pore development and reservoir properties. For samples of similar metamorphic grade, the coal-forming environment, as the main controlling factor affecting the macerals and industrial components of coal reservoirs, can affect the pore structure characteristics by influencing coal composition. Different coal-forming environments contain different plant assemblages, hydrodynamic conditions, and mineral supplies, which will inevitably affect the pore structure characteristics of the resultant coal. Therefore, it is important to discuss the controlling effect of the coal-forming environment on the pore characteristics of coal reservoirs. According to the data presented herein, the coal-forming sedimentary environment changed with depth through the Benxi coal seam; hence, the corresponding coal facies will also change with depth. To discover which type of coal facies controlled the pore structure of the Benxi coal as comprehensively as possible, we analyzed correlations between the four parameters of TPI, GWI, V/I, and F/M and the porosity, specific surface area, total pore volume, and other parameters obtained via the Hg intrusion experiment; this permitted a determination of the coal facies index that had the greatest influence on the pore structure.

It was found that not all coal facies parameters had an impact on pore structure characteristics. Figure 11 (a) shows that a correlation between TPI and total pore volume was not evident, indicating that the preservation of plant cells had little effect on these coal pores. According to Fig. 11 (b-c), the correlations between F/M and total pore volume, and VI and total pore volume, were also weak, indicating that pore structure was not greatly affected by vegetation type or bone base ratio.

Fig. 11

The correlation between coal facie parameters and ash content, total pore volume and specific surface area. (a) Relationship between total pore volume and TPI; (b) Relationship between total pore volume and VI; (c) Relationship between total pore volume and F/M; (d) The relationship between ash and GWI; (e) The relationship between total pore volume and GWI; (f) The relationship between specific surface area and GWI.

The GWI can be used to reflect changes in peat accumulation rate, groundwater level, and terrigenous debris and mineral contents. The correlations between pore characteristic parameters and GWI were analyzed (Fig. 11d‒f) and it was found that GWI had a significant positive correlation with ash content, and significant negative correlations with pore specific surface area and total pore volume. This can be explained as follows: an increase in GWI reflects an increase in the water level of the peat swamp; at this time, the hydrodynamic force is enhanced, resulting in stronger degradation of macerals and increased mineral input, in turn leading to an increase in ash content; these exotic minerals will fill the larger primary pores in coal, resulting in negative correlations between pore specific surface area and GWI, and between total pore volume and GWI.

In general, therefore, the coal facies parameter that has the greatest influence on pore structure characteristics in this area is the GWI, which can reflect changes in water level and mineral input. According to data presented herein, the GWI in the upper coal seam layer is comparatively low (Table 2), F/M is high, and V/I is high; this corresponds to a low groundwater level, weak hydrodynamic force, low mineral input, and low ash content. The plant texture was robust, with a high proportion of woody plants, and the plant structure is well preserved. Hence, coal pores in the upper layer of the Benxi coal seam are well developed, being more conducive to the storage and migration of coalbed methane. The GWI in the lower coal seam layer is comparatively high, and F/M and V/I are low. This coal phase was dominated by herbaceous plants, whose texture was easy to break, and the plant structure is poorly preserved. Moreover, groundwater dynamics were strong, bringing abundant terrigenous detrital minerals. Therefore, the lower layer of the Benxi coal seam is not conducive to the storage and migration of coalbed methane.

Distribution characteristics of different pore types

Coalbed methane mainly occurs in various pore fissures of the coal matrix. The development of coal pores plays a key role in the adsorption, storage, and seepage of coalbed methane, and is also one of the main controlling factors of coal reservoir physical properties. Therefore, we studied the distribution characteristics of different pore types.

Studies have shown that pore structure has heterogenous characteristics, and the pore system of a coal reservoir is mainly divided into adsorption pores (pore size < 100 nm) and seepage pores (pore size > 100 nm)⁵⁶. According to previous studies, the low temperature liquid N₂ experiment can successfully describe the pore structure of 2–50 nm mesopores, the CO₂ adsorption experiment can be used to characterize the pore structure of 0–2 nm micropores, and high-pressure Hg intrusion porosimetry can be used to determine the pore structure characteristics of > 50 nm macropores^57,58. From Fig. 3, it can be seen that Benxi coal mainly contains micropores, followed by a few macropores, and mesopores are basically not developed. According to prior research^59,60when Ro is low, coal seam compaction is also low. This results in a loose coal structure, where large and medium-sized pores are well developed, leading to high coal porosity. With an increase in Ro, the compaction effect increases, and coal undergoes a series of changes under the action of increasing temperature and pressure. As coal is continuously compacted, water is excluded, the coal becomes denser, and some primary pores are destroyed or filled. Eventually, the pore structure becomes dominated by micropores. Owing to the high maturity of coal samples in our study area, the pore structure is mainly dominated by micropores, with relatively low contents of macropores and mesopores.

The data presented herein showed that the amount of CO₂ adsorption was negatively correlated with ash content (Fig. 12a) and positively correlated with fixed carbon content (Fig. 12d), indicating a negative correlation between ash content and micropore content, and a positive correlation between fixed carbon content and micropore content.

Fig. 12

The correlation between carbon dioxide adsorption curve, nitrogen adsorption curve, mercury intrusion curve and ash, fixed carbon. (a) The relationship between ash and CO₂ adsorption capacity; (b) The relationship between ash and N₂ adsorption capacity; (c) The relationship between ash and mercury intrusion; (d) The relationship between fixed carbon and CO₂ adsorption capacity; (e) The relationship between fixed carbon and N₂ adsorption capacity; (f) The relationship between fixed carbon and mercury intrusion.

According to previous studies, coal pores can be divided into four categories according to their origin: primary pores, metamorphic pores, exogenous pores, and mineral pores⁶¹. Primary pores are those that existed during coal deposition, and mainly include cellular pores and inter-chip pores. A cell cavity pore (or plant tissue pore) is a cell structure pore of the coal-forming plant itself, while an inter-chip pore refers to various detrital microscopic bodies in the coal. When there is an increase in ash content, a large number of argillaceous minerals fill primary pores and block micropores; hence, the micropore content is reduced and the adsorption performance deteriorates.

In terms of secondary micropores left by the generation, accumulation, and gaseous dissipation of coalbed methane during coal metamorphism, previous studies have shown that most of these are filled with adsorbed gas. Primary pores and stomata are common in the vitrinite group, and because the plant cell cavity in the vitrinite group is relatively intact, the pore size is generally small, often dominated by transition pores and micropores. If the vitrinite content of a coal sample is high, the coalbed methane adsorption and storage capacity of the sample is also high. Thus, porosity is positively correlated with vitrinite content, while inertinite has a weak hydrocarbon generation ability⁶². The data presented herein revealed that the ash content was low in areas of the coal seam with high vitrinite content (Fig. 3). Hence, higher secondary porosity equates to better hydrocarbon generation ability, accompanied by a higher vitrinite content, and lower ash content. Therefore, in general, micropore content is negatively correlated with ash content and positively correlated with fixed carbon content. Through our investigations reported herein, we found that ash content was higher in the lower and topmost layers, and lower in the upper layer of the Benxi coal seam. Therefore, micropores in the upper layer are well developed, contain a large amount of adsorbed gas, and have good reservoir properties.

The development of pores is related to the composition, rank, and sedimentary environment of coal. Vitrinite contains a large number of pores, so the development of micropores in the upper layer of the coal seam is better than that in the lower layer; this is consistent with the sedimentary environment characteristics of the upper layer, being a reducing stable environment, with a high vitrinite content and low ash content. Hence, the upper layer contains a large number of pores, indicating a high hydrocarbon generation material basis and high gas generation capacity.

Mesopores are weakly developed in Benxi coal (Fig. 6). The main reason for this is that the samples in the study area are highly mature and its Ro value is about 1.4%, indicating that mesopores will have been gradually destroyed and filled via compaction. Vitrinite mainly contains micropores, while mesopores and macropores are mainly developed in inertinite⁶³. Samples in this study area are mainly composed of vitrinite; hence, the development of mesopores is generally poor.

There is no evident correlation of mesopores with ash or fixed carbon contents (Fig. 12b and e). For mesopores of 2–50 nm, atomic force microscopy observations have shown that these comprise interchain and molecular pores in coal⁶⁴. The interchain pores decrease with increasing coal maturity. Molecular pores tend to be larger than interchain pores, mainly concentrated in the size range of 2–50 nm. Qin et al.⁶⁵ also pointed out that with an increase in coal rank, owing to the continuous reduction of coal side chains and gradual and orderly arrangement of coal macromolecules, the gap between molecules is squeezed, resulting in the gradual reduction of intermolecular pores in coal. This shows that coal mesopores should be closely related to the side chains and basic structural unit gaps in coal molecules, and the formation of mesopores should be controlled by changes in these parameters. Moreover, during weak deformation, structural deformation leads to the degradation and shedding of alkyl side chains and functional groups, resulting in the formation of small molecular free radicals. Some of these exist in mesopores in an adsorbed state, while others are transferred to mesopores and combine with the carbon atomic skeleton of the pore wall, resulting in the destruction of connectivity between mesopores and segmentation of various forms of micropores. Mesopores in raw coal are squeezed by tectonic stress, directional shrinkage occurs to different degrees along the direction perpendicular to the stress action, and differences in pore structure are reduced. During strong deformation, macromolecules are condensed under the action of strong tectonic stress, and the interlayer spacing is reduced, forming a large number of micropores dominated by molecular interlayer pores. Therefore, the formation of mesopores is mainly related to tectonic stress. In turn, the correlations between mesopores and ash and fixed carbon contents are poor^66,67.

Figure 12 (c) and (f) shows that the amount of Hg injection is negatively correlated with ash content, and positively correlated with fixed carbon content. This shows that macropores are better developed in coal with less ash content. This is mainly because a high ash content leads to the filling of macropores, thereby reducing the overall macropore content. As shown in Fig. 3, the upper layer was characterized with lower ash content, showing greater abundance of larger pores with less pore filling. As a result, macropores are predominantly concentrated in the upper section of the Benxi coal seam. This contributes to enhanced permeability and the presence of free gas in the upper layer, facilitating the desorption of coalbed methane.

In summary, there are a large number of macropores and micropores in the upper layer of the Benxi coal seam. This shows that, in addition to containing a large amount of adsorbed gas, this layer also contains a large amount of free gas. In addition, the vitrinite content of the upper layer is high and the ash content is low. Therefore, the production potential of the upper coal seam is large, meaning that it should be a key layer for exploration and development.

Microscopic characteristics of pore development

Scanning electron microscopy observations confirmed the heterogeneity of pore distribution. The coal at the very top (Fig. 13a‒c) and lower layer (Fig. 13g‒i) of the Benxi coal seam contains a large number of micropores, including primary porosity and secondary porosity. Some of the primary porosity is filled with large quantities of clay minerals. However, in the upper layer (Fig. 13d‒f), there are a large number of pores that have been subjected to structural changes and destruction. Most of these are plant cell pores and are not filled with clay minerals. Therefore, the porosity of the upper layer is good and has good reservoir performance. In summary, macropores and micropores are mainly developed in the upper layer of Benxi coal. Therefore, the upper layer is an important channel for gas storage and migration, and has favorable reservoir properties.

Fig. 13

Field emission scanning electron microscope observation photos of Benxi Formation 8 # coal sample. (a) The SEM image of a coal sample with a depth of 2803.5 m (Bar = 3 μm). The content of clay minerals is high, and some primary pores are filled with clay minerals; (b) The SEM image of a coal sample with a depth of 2803.5 m (Bar = 10 μm). A large number of micropores can be seen, some of which are filled with clay minerals; (c) A maceral photo of a coal sample at a depth of 2803.5 m; (d) The SEM image of a coal sample with a depth of 2805.34 m (Bar = 10 μm). A large number of mesopores can be seen, most of which are plant cytoplasmic pores, and are hardly filled with clay minerals; (e) The SEM image of a coal sample with a depth of 2805.34 m (Bar = 10 μm). A large number of pores can be seen, some of which are subjected to structural changes and destruction, and almost not filled with clay minerals; (f) A maceral photo of a coal sample at a depth of 2805.34 m; (g) The SEM image of a coal sample with a depth of 2810.09 m (Bar = 10 μm). It can be found that the pore content is relatively small; (h) The SEM image of a coal sample with a depth of 2810.09 m (Bar = 3 μm). The content of clay minerals is relatively high, and some primary pores are filled with clay minerals; (i) A maceral photo of a coal sample at a depth of 2810.09 m.

Sedimentary model of Benxi coal in the Benxi Formation

In different sedimentary systems, coal seams occur at different locations in genetic stratigraphic units (cycles). It forms a variety of combinations with the roof and the surrounding rock within a certain distance above the roof, forming a variety of coal accumulation environments with a particular distribution law in a region. According to the results reported herein, we can conclude that the Carboniferous Benxi coal seam in the Ordos Basin belongs to the transitional facies coal-bearing series and was a peat flat deposit. Its formation was closely related to changes in sea level. Studies have shown that the climate was mild and humid before this coal-forming period, which was conducive to the development of a large number of plants⁵. In the Late Carboniferous, owing to tectonic action, the ancient terrain of the study area sank as a whole, and silt depressions were formed in some areas. At this time, due to poor burial conditions, plants were degraded and stable coal seams were not formed. With a rise in sea level, tidal-flat mudstone deposition occurred. Because of tidal action, a large-scale regression occurred in the study area (Fig. 14 ①). The shelf sedimentary facies were exposed to the water surface, and plants continued to grow and gradually occupy the basin. In addition, large-scale regression was accompanied by small-scale transgressive events, resulting in peat swamps in local offshore areas covered by seawater. Here, the soil was prone to swamping, resulting in the formation of nutrient-rich low peat bogs (Fig. 14 ②). Because of the low-lying terrain, there was an abundant water supply, so the peat water became rich in minerals. This was coupled with the mixing of terrigenous debris, so peat accumulated alongside the deposition of a large amount of minerals. At this time, the water body was a active water environment with strong fluidity, and the content of reactive oxygen species in the water body increased. Subsequently, with a further fall in sea level, coastal zones where peat had previously accumulated would have once again accumulated peat, and plants would have further extended their growth to the peat flats (Fig. 14 ③). In this stage, the water body comprised a relatively reducing environment, and the peat flat sedimentary environment was stable. During the later stage of coal formation, sea level rose. The original peat swamp was covered by seawater, and the spreading growth of plants was terminated by seawater inundation; plant growth receded to the tidal flats. At this time, the water body was a deep-water retention reduction environment, and peatification stopped (Fig. 14 ④). Finally, with the occurrence of large-scale transgressive events, previously deposited peat was covered by seawater. In this reducing environment, mudstone, sandstone, and limestone were deposited, and coalification began to occur, ultimately forming coal. The swamp environment of coal-forming becomes a reduced tidal flat environment, indicating the end of coal-forming period (Fig. 14 ⑤).

Fig. 14

Sedimentary model of Benxi Formation 8 # in Ordos Basin.

After the formation of coal, owing to the influence of the sedimentary environment, the development characteristics of coal reservoir pores in different coal-forming environments were evidently different. In general, primary pores, such as plant structure, and secondary pores were affected during the diagenesis stage. Because of the early oxidizing environment with strong water fluidity in the lower coal seam, the gelatinization effect was weak. This coal phase is dominated by herbaceous plants, and the structure of plant residues is poorly preserved. The content of inertinite in this coal is high and the content of vitrinite low, and the strong hydrodynamic force brought a large quantity of terrigenous debris and minerals, resulting in an increase in ash content, which is not conducive to the development of pores. Therefore, the pore development degree of the lower coal seam layer is low and the connectivity poor, which is inconducive to the storage and migration of coalbed methane. Moreover, owing to the intermittent occurrences of regression and transgression, the water body was unstable. Therefore, the change of each component content is not stable enough, and there are up and down fluctuations. The macroscopic coal types formed were mostly semi-bright coal and bright coal. For the middle portion of the coal seam, the water body became a retention reduction environment due to a medium-term regressive event. The vitrinite content increased and the inertinite content decreased, the hydrodynamic conditions became weaker, and the ash content decreased, which is conducive to the development of pores. Therefore, the coal reservoir properties of this layer are good, and it generally comprises semi-bright coal. Due to a large-scale transgression during the later stage of deposition of the upper coal seam, a large amount of seawater coverage resulted in the formation of a deep-water retentive environment. At this time, the content of vitrinite continued to increase, while the content of inertinite decreased. The water body became a hydrostatic environment, with weak hydrodynamics and low ash content. Therefore, pore development improved, and more macropores and micropores developed, which is conducive to the migration and storage of adsorbed and free gas. Therefore, the upper layer has good reservoir properties and generally comprises bright coal. Because of the late deposition of the roof, the sedimentary environment changed, resulting in the increase of terrigenous materials during the deposition period. The sedimentary argillaceous composition is increased, and the lithology is transformed into mudstone, mostly gangue.

In summary, the lower layer of the Benxi coal seam comprised an active water oxidizing environment with strong activity due to regression and transgression. The inertinite content is high and the vitrinite content is relatively low. The hydrodynamic conditions were strong, ash content is high, and degree of pore development is low. This layer mainly comprises semi-bright coal and is not conducive to gas storage and migration. In contrast, large-scale transgression occurred during deposition of the upper layer, resulting in a stagnant reducing environment. The vitrinite content increased and inertinite content decreased. The hydrodynamic force weakened, ash content decreased, and there was improved pore development. This layer mainly comprises bright coal and has good reservoir properties. The roof is in the late stage of deposition, so a large amount of debris is added during the deposition period, and finally reach the end of the coal-forming stage.

Conclusions

(1) The Benxi coal seam, as a whole, exhibits a higher average vitrinite content than inertinite content, reflecting a sedimentary environment of a partially reducing peat swamp overlain by water. The TPI‒GI facies diagram indicates overall conditions of deep overlying water, weak hydrodynamics, and high degree of gelation, equating to an open water area. The GWI‒V/I facies diagram indicates that the coal-forming plants were mostly woody plants, and the hydrodynamic force was weak; hence, the general environment comprised a water-covered swamp.

(2) Through analysis of the industrial composition, sulfur content, and petrological characteristics of the Benxi coal seam, vertical heterogeneity was revealed. Variations in ash and sulfur content from bottom to top of the coal seam reflect changes in hydrodynamic conditions from strong to weak and back to strong, and changes in redox environment from oxidizing to reducing and back to oxidizing. The study of macerals showed that the content of vitrinite in the coal seam gradually increased from bottom to top, and the content of inertinite gradually decreased. Meanwhile, V/I and OI values demonstrated that the lower and topmost layers in the vertical direction were dominated by an oxidizing environment, while the upper layer was dominated by a reducing environment. Finally, F/M data showed that the hydrodynamic force of the lower and topmost layers was strong, and that of the upper layer was weak.

(3) Through the study of reservoir development characteristics and their controlling factors, it was found that porosity negatively correlated with ash content, and positively correlated with fixed carbon content. The GWI also had a strong control effect on porosity, with a larger GWI equating to lower porosity. The GWI of the upper layer of the coal seam was low owing to a low ash content. Therefore, we conclude that pores in the upper layer of the coal seam are well developed; both macropores and micropores are better developed in the upper layer compared with the lower layer, with the former having good reservoir potential.

(4) On the basis of the sedimentary environment and pore development characteristics, a sedimentary model diagram of the Benxi coal seam can be built. We contend that a peat-flat sedimentary system resulting from sea regression and transgression developed during the coal-forming period of the Benxi coal seam in the Ordos Basin. The upper layer of this coal seam has the highest vitrinite content, lowest ash content, and highest porosity; hence, it is the dominant reservoir section.