Investigation of Coal Structure and Its Differential Pore–Fracture Response Mechanisms in the Changning Block

Abstract

The deep coal seams in the southern Sichuan region contain abundant coalbed methane resources. Determining the characteristics and distribution patterns of coal structures in this study area, and analyzing their impact on pore and fracture structures within coal reservoirs, holds substantial theoretical and practical significance for advancing coal structure characterization methods and the efficient development of deep coalbed methane resources. This paper quantitatively characterizes coal structures through coal core observations utilizing the Geological Strength Index (GSI) and integrates logging responses from different coal structures to develop a quantitative coal structure characterization model based on logging curves. This model predicts the spatial distribution of coal structures, while nitrogen adsorption data are used to analyze the development of pores and fractures in different coal structures, providing a quantitative theoretical basis for accurately characterizing deep coal seam features. Results indicate that density, gamma, acoustic, and caliper logging are particularly sensitive to coal structure variations and that performing multiple linear regression on logging data significantly enhances the accuracy of coal structure identification. According to the model proposed in this paper, primary-fragmented structures dominate the main coal seams in the study area, followed by fragmented structures. Micropores and small pores predominantly contribute to the volume and specific surface area of the coal samples, with both pore volume and specific surface area increasing alongside the degree of coal fragmentation. Additionally, the fragmentation of coal structures generates more micropores, enhancing pore volume and suggesting that tectonic coal has a greater adsorption capacity. This study combines theoretical analysis with experimental findings to construct a coal structure characterization model for deep coal seams, refining the limitations of logging techniques in accurately representing deep coal structures. This research provides theoretical and practical value for coal seam drilling, fracturing, and reservoir evaluation in the southern Sichuan region.

1. Introduction

As a heterogeneous porous medium, tectonic coal exhibits a complex internal pore–fracture structure primarily resulting from uneven stress distribution caused by horizontal stress compression on the coal seam. This intricate pore structure not only influences the presence of coalbed methane within the original coal matrix but also plays a critical role in the desorption, diffusion, and seepage of coalbed methane during coal seam extraction [1]. Consequently, investigating the development characteristics of regional coal structures and variations in their pore–fracture structures is of substantial theoretical and practical importance for predicting coal and gas outbursts and evaluating the recoverability of coalbed methane resources [2,3,4,5,6].

In recent years, numerous scholars have conducted extensive research on the characteristics of coal structure and pore structures in tectonic coal. However, current methods for characterizing coal structure still require refinement. Previous studies have predominantly examined the primary controlling factors of coal structure in isolation, such as macroscopic classification descriptions, microscopic observations at various scales, molecular structure simulations, and interpretations of logging and seismic data. These studies have mainly focused on characterizing the pores and fractures in tectonic coal from different microscopic perspectives and on qualitative and quantitative analyses and predictions based on logging data [7]. In quantitative studies of coal structure, mathematical formulas derived from logging parameter response analysis are commonly used for quantitative analysis and identification. Currently, machine learning methods such as Support Vector Machines (SVMs), Backpropagation Neural Networks (BPNNs), and XGBoost are also employed to classify coal structures [8,9,10]. Among these, the Geological Strength Index (GSI) is established through relationships with logging curves, enabling the quantitative identification of coal structures. This approach significantly enhances the applicability of coal structure quantification and has therefore been widely adopted [11,12,13,14].

Second, varying levels of tectonic deformation result in different degrees of change in coal structures, which subsequently influence the development and distribution characteristics of pores and fractures [15,16,17]. In coalbed methane exploration and development, understanding the pore–fracture structure characteristics of coal reservoirs is theoretically significant for selecting favorable areas and predicting coalbed methane distribution [18,19]. Researchers analyzed the pore structure of tectonic coal using mercury intrusion porosimetry, concluding that tectonic processes increase both the total pore volume and specific surface area in coal [20]. Some researchers also researched faulted tectonic coal using low-temperature nitrogen adsorption experiments, finding that the specific surface area, pore size, and pore volume of faulted tectonic coal exceed those of primary-structure coal [21]. Some researchers contend that the influence of coal structure on permeability and pore characteristics, as analyzed through liquid nitrogen experiments, is a critical area in coalbed methane extraction. Liquid nitrogen adsorption experiments have elucidated the pore characteristics of coals with varying structures. For instance, gas adsorption in tectonically deformed coals predominantly occurs in pores with diameters of approximately 3.3 nm. As coal damage intensifies, both the specific surface area and the fractal dimensions of pore volume increase [22]. These findings establish a theoretical basis for understanding the relationship between coal structure and permeability, providing scientific insights for the efficient extraction of coalbed methane. While most scholars analyze the characteristics of tectonically deformed coals by comparing them with primary-structure coals, research focusing on specific structural categories of coal remains relatively scarce. Additionally, prior studies have largely concentrated on shallow, easily accessible coal seams, with limited investigation into deeper, less accessible seams. This paper advocates for a comprehensive approach to describing coal structure, combining macro- and micro-scale features with quantitative logging interpretation, representing a relatively novel perspective in the field.

As stress distribution during the formation of syncline tectonic coal varies across regions, pore structure characteristics are impacted accordingly, leading to varying degrees of change in the pore structure of coal at different locations. Therefore, systematic research on these patterns is necessary.

The study area, Changning in southern Sichuan, is situated within the low-amplitude fold belt of southern Sichuan and lies in the coal-bearing region encompassing southern Sichuan and northern Guizhou. The Longtan Formation in this region is rich in coal resources and is located in a transitional marine–terrestrial environment, which is favorable for coal seam development. The sedimentary facies of the Longtan Formation evolve from east to west, transitioning through platform, tidal flat, delta, braided river, and alluvial fan facies, spanning marine, transitional marine–terrestrial, and terrestrial environments. The area is abundant in coal resources, with coal seams primarily composed of semi-bright coal in a high to over-mature stage. The coal is of high quality, with favorable gas-bearing properties. Additionally, the coal rock and the mechanical properties of the roof and floor rocks are advantageous for mining, offering favorable conditions for industrial coalbed methane development.

This study, therefore, focuses on the coal rocks with varying coal structures of the Longtan Formation in the Changning Block of southern Sichuan. By employing both macroscopic descriptions of coal structures and microscopic descriptions through microscopy, a GSI quantitative model of coal structure based on logging curves is developed to predict the spatial distribution of coal structures and to identify the factors influencing spatial distribution differences. Subsequently, mercury intrusion data are used to characterize pore and fracture development across different coal structures. The objective is to uncover the distribution characteristics of various coal structures in the study area and to assess the impact of these structures on pore–fracture characteristics, thereby providing technical support for selecting optimal areas in future deep coalbed methane exploration and development.

2. Geological Settings

The Sichuan Basin developed atop the Yangtze Craton platform’s basement and exhibits the typical characteristics of a superimposed basin [23,24]. Based on its basement structure, deformation characteristics, and evolutionary processes, the Sichuan Basin can be divided into six structural belts: the Eastern Sichuan high-amplitude fold belt, the Southern Sichuan low-amplitude fold belt, the Northern Sichuan depression belt, the Western Sichuan depression belt, the Central Sichuan gentle fold belt, and the Southwestern Sichuan low-amplitude fold belt [25,26]. The Changning Block, within the study area, is located at the southern edge of the Sichuan foreland basin in the Xuyong-Junlian superimposed fold belt. To the east and west, it borders the north–south structural belts of the Chuan-Qian and Chuan-Dian regions, while to the north, it connects with the southern low-gentle fold belt of the Huaying Mountains and the Weiyuan uplift. To the west, it is adjacent to the Leibo uplift, part of the east–west structural belt of the Yanjin-Weixin area. This structural system significantly influences the region, resulting in a complex morphology where components from the east–west, northeast, northwest, and north–south directions coexist. The Changning Block is predominantly characterized by northeastward transverse fold structures, often constrained by north–south structures. East–west structures further impact the formations, creating a composite connection of northeast structures. The folds in the area exhibit S-shaped bends in plan view, with generally broad anticlines featuring gentle dips, well-developed secondary structures, and complex formations. In contrast, synclines are narrower, with steeper dips, fewer secondary structures, and relatively intact syncline forms (Figure 1).

The primary coal-bearing strata in the study area are the Upper Permian Changxing Formation (P_3c) and the Longtan Formation (P_3l). The total thickness of these strata ranges from 201.64 to 284.62 m, with an average thickness of 252.25 m, classified as nearshore transitional deposits of an upper delta plain. The lithology of the Longtan Formation mainly comprises gray and dark gray mudstone, sandy mudstone, claystone, carbonaceous mudstone interbedded with multiple layers of fine sandstone, siltstone, silty mudstone, and coal seams. The base includes light gray pyrite-bearing aluminous mudstone, tuffaceous sandstone, dark gray shale, and thin coal seams. The Changning area contains 5 to 10 coal seams, with the thickest cumulative coal seam reaching 5.84 m and an average thickness of 3.52 m. According to core sampling data from the Changning area, the maximum single seam thickness is 3.14 m, primarily consisting of anthracite. The Longtan Formation in the Changning region primarily develops four coal seam sets: C₃, C₇, C₈, and C₂₅. These seams are mostly concentrated at the top and middle–lower sections of the coal-bearing strata. The upper two coal seams and the lower seam sets, labeled sequentially from top to bottom as C₃, C₇, and C₂₅, are characterized by gray–black anthracite with dull to bright luster.

3. Characterization of Coal Structure

3.1. Structural Characteristics of Different Coal Bodies at a Macroscopic Scale

According to the national standard (GB/T30050-2013), coal structure classification is divided into primary structure, fragmented structure, granular structure, and mylonitic structure [27]. Due to the damage to coal structure during coring, it is challenging to identify granular and mylonitic coal from coal cores. Therefore, the coal structure in the study area is categorized into three classes: Class I coal, Class II coal (fragmented coal and some granular coal), and Class III coal (granular coal and some mylonitic coal).

The primary structural feature of the region is the Changning anticline. Accordingly, four coring wells from the northern and southern sections of the anticline were selected as sample sources. Among these, coal and rock samples with larger fracture spans were specifically chosen for the experiment. During the core logging process, key characteristics were identified through hand specimen observations. Class I coal exhibits well-defined boundaries between macroscopic coal rock types, distinguishable components, an intact structure, undisturbed layering, and a prominent primary banded structure. It is hard and difficult to break by hand, with discernible internal and external fractures. The coal core appears mainly as columns or large, angular blocks with no relative displacement between blocks, indicating a relatively hard coal texture. Class II coal cores are fragmented, with distinct boundaries between macroscopic coal rock types and identifiable components. This type results from slight damage to the primary structure, with discontinuous primary banded structures still visible. The coal body may be relatively intact or fragmented, with the coal core appearing primarily as angular fragments that exhibit relative displacement between blocks. It can be easily broken by hand, has a relatively hard texture, and displays well-developed cleats and fractures with good connectivity. Class III coal lacks discernible macroscopic coal rock types due to severe damage to the primary structure, with layering lost and a powdery, dull structure. Fractures are unobservable, and the coal body is severely damaged, presenting as granular or powdery with very low hardness and strength, easily crumbling into coal powder or dust, with a major particle size of less than 1 mm (Figure 2).

3.2. Fissure Characteristics of Coal with Different Coal Structures at a Microscopic Scale

Microscopic fractures are linear spaces that are invisible to the naked eye and can only be observed using an optical microscope or scanning electron microscope (SEM). These fractures are primarily developed in vitrain and clarain and are significantly smaller in scale than macroscopic fractures. Based on the observation scale, they are further classified into microfractures, observable under an optical microscope, and submicroscopic structures, visible only under SEM [28,29].

3.2.1. Characteristics Under an Optical Microscope

After analyzing and observing the hand specimens, the collected samples were promptly sealed in airtight bags to minimize contamination and oxidation. In the sample preparation room, the samples were crushed and sieved to obtain 18–40 mesh fractions for preparing coal powder polished sections. Microscopic and mineralogical characteristics were then analyzed using an optical microscope and a scanning electron microscope (SEM). For microscopic observation, a Leica DM4P (Leica Camera AG, Wetzlar, Hesse, Germany) optical microscope was employed to examine and quantify the morphology of coal components. The experiment was conducted in the Coalbed Methane Laboratory at the China University of Mining and Technology. Under an optical microscope, individual fractures in Class I coal appear as linear, serrated, or zigzag patterns, with multiple fractures forming a network. In contrast, Class II and Class III coals, subjected to stronger tectonic stress, exhibit an increase in both the size and number of fractures. The two sides of these fractures display relative displacement, with altered particle arrangement on both sides, resulting in fracture formation. The banded structure and bedding in Class II and Class III coals are no longer visible, and macroscopic coal components (vitrain, clarain, durain, fusain) cannot be distinguished. The luster is dull, and fractures develop without directional preference. Displacement (micro-faults) may occur on both sides of the fractures, accompanied by signs of wrinkling (micro-folds), and may even exhibit characteristics of fibrous brecciation. Organic macerals frequently show signs of compressional or torsional deformation; under severe stress, they may fracture or displace and are often re-cemented. Slickensides and plastic flow phenomena are also commonly observed (Figure 3).

3.2.2. Characteristics Under a Scanning Electron Microscope

The impact of structural forces on the macroscopic features of coal can be discerned visually; however, microscopic characteristics such as pore and fracture morphology require observation through an optical microscope. Scanning electron microscopy (SEM) has rapidly advanced and is now widely utilized as a key tool for analyzing the microscopic surface structure of coal [30].

In this study, microscopic observations using a scanning electron microscope (SEM) were performed with a Hitachi S4800 (Hitachi Ltd., Tokyo, Japan) instrument, adhering to the relevant standard (SY/T5162-2021) [31]. By examining multiple points under the microscope, the following characteristics were identified and summarized.

A comparison of SEM images of coal samples with varying degrees of damage (Figure 4) reveals that primary structural coal typically exhibits a relatively uniform pore distribution, with diverse pore shapes including round, oval, and other irregular forms. The size and connectivity of these pores vary depending on the maturity of the coal and its geological history. In contrast, structurally deformed coal shows a more complex pore structure due to tectonic forces, characterized by an increased number of fractures and irregular pores, which may interconnect to form a complex network structure [32]. Primary pores, metamorphic pores, and mineral pores predominantly develop in primary structural and fragmented coal, where the structure is relatively well preserved; exogenous pores are more prevalent in severely damaged granular and mylonitic coal, displaying poorer stability and irregular shapes [33]. As the degree of coal body damage increases, the integrity and connectivity of the coal progressively deteriorate. Intense shearing effects can lead to the formation of highly distorted and deformed pores with irregular edges and shapes. These changes may disrupt the adsorption–desorption equilibrium of methane in tectonically deformed coals, resulting in storage states that differ from those of conventional coal seams and ultimately impacting the efficiency of coalbed methane extraction [34,35].

3.3. Logging Identification of Coal Structure

The Geological Strength Index (GSI) is a method for quantifying rock strength, introduced by E. Hoek and colleagues, based on factors such as rock integrity, surface roughness, and fracture development. Due to the high correlation between the GSI concept and the intrinsic characteristics of coal structure, this method has been widely used by scholars for the quantitative characterization of coal structures. This system has been employed for issues such as instability prediction, residual strength calculation of jointed rock masses, slope stability, and the prediction of rock strength and deformation modulus. The quantitative characterization of rock mass structure has provided a reference for understanding coal body structure. After extensive experimentation, scholars adapted the GSI, originally used to characterize rock mass strength, for coal body structure prediction and refined it to better suit the classification of different coal body structures. The improved GSI quantification system replaced the original rock surface weathering condition index with a comprehensive index that considers fracture development and structural surface conditions in coal samples [36].

Specifically, the GSI value is assigned based on factors such as the size of coal fragments, the degree of fracture and joint development, and the condition of structural surfaces. The basic quantification criteria are shown in Figure 5a. According to the classification criteria in the figure, coal samples with a complete block structure on the surface, clear banded layers with a layered or laminated structure, minimal fractures over a large area of the coal wall, distinguishable internal and external fractures, no noticeable mirror surfaces, and no relative displacement of the blocks are classified as Type I coal, which corresponds to a score from the first and second rows. Next, by observing the surface conditions under five levels, the corresponding column is determined. The final score is obtained by intersecting the row and column, as shown on the far right of the figure, with a score ranging from 80 to 90. The macroscopic characteristics of various coal structures in the study area and their corresponding GSI values are displayed in Figure 5b.

Based on the GSI quantification of core samples, a correlation analysis between GSI and logging curves was conducted. The results indicate that both DEN (Density) and AC (Acoustic) logging show a strong correlation with GSI, with correlation coefficients (R²) of 0.635 and 0.528, respectively. This strong correlation is primarily due to the significant relationship between sound wave propagation velocity in coal seams and coal structure. As the degree of coal fragmentation increases, the overall looseness of the coal rock also increases, leading to a decrease in sound wave propagation speed and a notable increase in time delay [37]. Similarly, different coal structures exhibit varying degrees of damage; as fragmentation increases, the number of pores and fractures in the coal rock rises, resulting in a relative decrease in coal rock density, which enhances response characteristics.

In contrast, the relationships between CNL (Compensated Neutron Log), CAL (Compensated Acoustic Log), GR (Gamma Ray), and R (Resistivity) with GSI are not significant. Origin (24) software was used for correlation analysis, yielding the following results. Specifically, the correlation coefficient (R²) for CNL and GSI is 0.371, that for CAL and GSI is 0.142, that for GR and GSI is 0.2, and that for R and GSI is 0.259 (Figure 6). The figure shows that some isolated points are the reason why the trend line does not perfectly fit the results. However, these points occur during the actual drilling process and can also be explained by the linear trend observed in increasingly fractured coal. Therefore, retaining the trend line can enhance the practical applicability of the prediction formula.

Due to the low correlation between individual logging curve values and the GSI, a more accurate quantitative characterization of coal structure was achieved through multiple linear regression between GSI and logging values. This led to the establishment of a quantitative characterization model for coal structure based on logging curves, as shown in Equation (1). The fitting coefficient (R²) for the multiple regression of GSI and logging curves is 0.846, representing a significant improvement in correlation compared to that for single logging fit values. This indicates that using multiple logging curves together provides a more effective method for evaluating coal structure. Additionally, analysis using SPSS (27) revealed that the model demonstrates a good fit and is applicable. Second, a correlation analysis of the actual values of GSI with the predicted values of GSI shows that their correlation is considerably high (Figure 6). This indicates that combining multiple logging curves is a more effective method for evaluating coal structure.

For the study area, the identification intervals for coal structure are as follows: Class I coal has a GSI between 70 and 100, Class II coal has a GSI between 45 and 70, and Class III coal has a GSI between 0 and 45.

GSI = 3.678DEN − 0.871AC + 4.27CAL + 0.291CNL − 0.025GR − 0.026R + 81.908
y = 81.908 − 0.026*R − 0.025*GR + 4.27*CAL + 0.291*CNL − 0.871*AC + 3.678*DEN

(1)

In the equation, GSI represents the quantitative value of coal structure and is dimensionless; DEN is the density logging value in g·cm⁻³; GR is the natural gamma logging value in API; AC is the acoustic time delay logging value in μs·ft⁻¹; CAL is the borehole diameter logging value in inches; R is the resistivity logging value in Ω·m; and CNL is the compensated neutron logging value in %.

4. Pore Structure Characteristics of Coal Reservoirs with Different Coal Structures

Coal exhibits dual structural features of pores and fractures, containing pore fissures that range from the nanometer to millimeter scale. These structures form a link between pores and fractures and act as migration pathways for coalbed methane [38]. The characteristics and proportions of pores at different scales significantly influence the processes of adsorption, desorption, diffusion, and migration of coalbed methane [39]. The distribution and connectivity of these pores and fractures impact the development characteristics of coalbed methane. Studying the pore characteristics of coal with different structural types enhances the understanding of variations in coal reservoir properties due to differences in coal structure, providing valuable foundational data and technical support for gas disaster prevention in coal mines, as well as for the evaluation and development of coalbed methane resources in regions with diverse coal structural developments.

4.1. Adsorption Characteristics of Liquid Nitrogen

Pores in coal refer to the spaces within the coal matrix that are not filled with organic matter or minerals. These pores are key structural elements of coal and serve as critical spaces for coalbed methane storage. Evaluation factors primarily include pore size distribution, specific surface area, pore volume characteristics, and pore morphology [40]. Coal samples with varying degrees of fragmentation were crushed into appropriate particle sizes (60–80 mesh) and analyzed using an ASAP2460 automatic surface area and porosity analyzer (Micromeritics Corporate Headquarters, Norcross, GA, USA) following the national standard GB/T 21650.2-2008 [41]. Low-temperature liquid nitrogen adsorption experiments were conducted to analyze the pore structure characteristics of coal reservoirs. A classification method specifically adapted for this study was applied, categorizing pore characteristics of different coal structures based on pore size. The categories are as follows: micropores (<2 nm), mesopores (2–10 nm), macropores (10–100 nm), and ultralarge pores (>100 nm) [42].

According to adsorption and condensation theories, solid materials with capillary pores exhibit both overlapping and separation phenomena during adsorption–desorption tests. The separation of the adsorption and desorption branches creates an adsorption hysteresis loop, the shape of which reflects specific pore structures. As coal is a porous material, the low-temperature liquid nitrogen adsorption–desorption curves and results provide valuable insights into coal pore morphologies and the pore size distribution that primarily influences adsorption.

In the study area, the low-temperature nitrogen adsorption–desorption curves generally present an inverted “S” shape (Figure 7). It is evident that some previous research on isothermal adsorption, such as the Langmuir and Freundlich isotherms, cannot adequately explain the adsorption–desorption curves shown in the figure. According to the six types of adsorption isotherms proposed by IUPAC, the nitrogen adsorption curve for the deep coal reservoir of the Longtan Formation in the Changning Block is classified as Type IV, indicating a significant presence of mesopores and macropores in the tested coal rock samples. The data show that all coal samples exhibit varying degrees of adsorption hysteresis loops on their adsorption isotherm curves. Morphologically, the hysteresis loops of the samples display characteristics of both H3 and H4 types. These loop characteristics suggest a complex, multi-scale pore system in the coal, comprising micropores, mesopores, and macropores, with pore shapes such as plate-like, wedge-shaped, slit, and ink-bottle pores.

Comparing the adsorption isotherm curves of coal samples with different structural types, it is observed that as coal damage increases, the distance between the two branches of the hysteresis loop widens, and the steep drop point on the desorption curve becomes more pronounced. This suggests that greater coal damage results in an increase in open pores, ink-bottle pores, and slit pores within the coal.

4.2. Pore Size/Specific Surface Area Distribution Characteristics

The pore volumes and specific surface areas across various pore size segments in each coal sample are primarily dominated by micropores and small pores (Figure 8). As the degree of damage to the coal body increases within the same coal seam, the overall pore volume and specific surface area generally increase. The specific surface area for Type I coal ranges from 0.0971 to 0.3137 m²/g, that for Type II coal from 0.2693 to 0.5304 m²/g, and that for Type III coal from 0.6443 to 2.7138 m²/g. It is evident that the specific surface areas of Type II and Type III coals are significantly higher, indicating that damage to the coal body leads to an increase in the specific surface area.

Coal pore volume significantly impacts the adsorption storage space and capacity of coalbed methane, as well as the permeability and production capacity of coalbed methane wells. It is a critical parameter for studying coal pore structure characteristics. The pore morphology in coal is diverse and varies in size; based on connectivity, pores are classified into open pores and “dead pores”. “Dead pores” are non-communicating, preventing the adsorbate from entering, whereas open pores have good permeability, allowing the adsorbate to fill the pores. Thus, the pore volume obtained through adsorption experiments reflects the equivalent gas (adsorbate) volume adsorbed within the open pores of a unit mass of adsorbent. The relationship between pore volume and pore size distribution in coal shows that the peak pore volume for Type I coal is primarily concentrated around 5 to 10 nm, that for Type II coal around 3 to 5 nm, and that for Type III coal around 1 nm (Figure 9). This suggests that as the coal structure becomes more fragmented, the pore diameters corresponding to peak pore volumes decrease, indicating that structural fragmentation generates more micropores, contributing to greater pore volume.

Coalbed methane primarily exists in an adsorbed state on the surface of coal pore matrix particles. The development of coal porosity and the specific surface area significantly influence the existence state, adsorption characteristics, storage capacity, and gas content (gas content, gas saturation, geological reserve abundance) of coalbed methane. The relationship between the specific surface area of coal and pore size distribution reveals that the specific surface areas of all three types of coal are mainly contributed by micropores and small pores, with peaks primarily distributed among micropores. As the coal structure becomes more fragmented, the contribution of small pores to the specific surface area decreases (Figure 10). It is evident that some previous research on isothermal adsorption, such as the Langmuir and Freundlich isotherms, cannot adequately explain the adsorption–desorption curves shown in the figure. In Type I coal, micropores contribute the most to gas adsorption. Some researchers suggest that pores, as critical pathways for gas diffusion, influence the mobility and permeability of gas in coal seams through their shape and size. Based on the curve patterns, Type I coal predominantly consists of pores with openings at both ends, resulting in better permeability. In Type II coal, the contribution of small pores to adsorbed gas decreases while that of micropores increases; the adsorbed gas occupies one-end-open pores, ink-bottle-shaped pores, and slit-shaped pores, which have poorer permeability. In Type III coal, micropores contribute more to adsorbed gas than small pores, mainly comprising ink-bottle pores and slit-type flat pores.

Changes in coal structure lead to variations in the proportion of micropores, which in turn affect gas adsorption in the coal seam. Micropores with pore diameters around 1.3 nm have the largest specific surface area and, consequently, the highest gas adsorption capacity. This means that when sufficient pores of about 1.3 nm are present in the coal, gas will preferentially adsorb in these pores. Type III coal contains the most such pores, resulting in increased gas adsorption. Among the adsorption–desorption curves of different coal structures, Type III coal exhibits the highest gas adsorption capacity.

5. Conclusions

To determine the integrity of underground coal rock samples without coring and to understand the differences in pore structure among coal rock samples with varying degrees of fragmentation, this study focuses on coal rock samples obtained from drilling in the Changning Block of the Sichuan Basin. The coal structure in the Changning Block was characterized through macroscopic description, microscopic observation, and analysis of logging parameters. Experimental tests on coal rocks with different levels of integrity were compared and discussed, leading to the following conclusions:

(1)

Observations of core samples at various scales indicate the development of Type I coal (primary coal), Type II coal (fractured coal and partially primary coal), and Type III coal (fragmented coal and partially mylonitized coal) in the study area. Under an optical microscope, fractures in Type I coal appear as linear, jagged, or zigzag patterned, forming network fractures when multiple cracks intersect. In contrast, Type II and Type III coal, subjected to stronger tectonic stress, exhibit increased and enlarged fractures, with rearranged particle alignment on both sides, leading to fractures. Scanning electron microscope observations reveal that Type I and II coal mainly develops primary pores, metamorphic pores, and mineral pores, while Type III coal primarily develops exogenous pores, which are less stable and irregular in shape.

(2)

Using correlation analysis, a multiple linear regression model was established between the Geological Strength Index (GSI) and logging curves, delineating identification intervals for different coal structures. The GSI for Type I coal ranges from 70 to 100, that for Type II coal from 45 to 70, and that for Type III coal from 0 to 45.

(3)

Experimental results indicate that the volume and specific surface area of coal rock samples are primarily dominated by micropores and small pores. As coal seam fragmentation increases, the total pore volume and specific surface area also increase. The nitrogen adsorption curves of deep coal reservoirs in the Longtan Formation of the Changning Block exhibit Type IV characteristics, with adsorption hysteresis loops combining features of Types H3 and H4. The pore systems mainly include plate-shaped, wedge-shaped, slit-shaped, and ink-bottle-shaped pores.

(4)

Based on the experimental results, it can be concluded that coal structure fragmentation generates more micropores, contributing to greater pore volume. Changes in coal structure alter the proportion of micropores, affecting the coal’s gas adsorption capacity. Pores with diameters around 1.3 nm have the largest specific surface area and highest gas adsorption capacity. Type III coal contains the most such pores, resulting in the highest gas adsorption. The adsorption–desorption curves also show that Type III coal has the greatest adsorption capacity.

These findings are significant for predicting the integrity of deep coal seams and understanding the pore structure differences among various coal structures, especially for the development of coalbed methane resources in deep coal reservoirs. Although this study establishes a predictive formula for coal structure, the geological factors influencing underground coal fragmentation and their effects on gas adsorption performance require further investigation. Moreover, the parameters used in the formula are derived from samples within the block, which limits its accuracy when applied to other regions. Nonetheless, the methodology can be widely applied. In the future, more coal samples obtained by drilling will enrich and improve this study.