Effects of pre-fracturing fluids on pore-fracture structure and mechanical properties of deep coal

Abstract

Hydraulic fracturing is a key technology for enhancing coalbed methane recovery. Preconditioning the reservoir with acidizing or oxidative fluids can significantly improve stimulation outcomes; however, comparative studies evaluating these fluids in coal seams are still limited. In this work, deep coal samples (depth > 2500 m) were treated with five different pre-fracturing fluids—slick water (PAM), hydrochloric acid (HCl), hydrochloric and hydrofluoric (HCl + HF) acid, sodium hypochlorite (NaClO), and hydrogen peroxide (H₂O₂)—and their effects on porosity, permeability, and mechanical properties were systematically compared using low-field NMR, SEM, and mechanical testing. The potential relationships between changes in mechanical parameters and pore-fracture structure were also examined. Results show that all five fluids increased total porosity by 33.98 ~ 45.04%. The HCl + HF produced the most dramatic permeability enhancement (13,561.54%), whereas HCl alone reduced permeability by 78.01%. Moreover, HCl + HF led to the most pronounced reduction in both compressive and tensile strength. Kendall correlation analysis revealed several significant relationships: Young’s modulus (E) correlated negatively with total porosity (φ_t) and average roughness (R_a); Poisson’s ratio (ν) showed a positive correlation with φ_t; and tensile strength (R_t) was negatively correlated with R_a. These findings provide practical insights for selecting pre-fracturing fluids based on desired reservoir weakening or permeability enhancement objectives.

Introduction

Coalbed methane (CBM) is a clean and efficient fossil energy source which exists in coal seams in both adsorbed and free states¹. It represents a substantial resource, with more than 50% of in-place CBM located in coal seams deeper than 1,524 m (5,000 feet)². Deep coalbed methane (DCBM) reservoirs exhibit significant differences from the shallow counterparts in terms of geological settings, rock porosity characteristics, and mechanical properties^3,4. Conventional fracturing fluids can induce mineral migration and pore-throat blockage, which may ultimately impair CBM recovery. Consequently, chemical modification methods such as acidizing and oxidation treatments have emerged⁵. A detailed understanding of the impact of these fluids on the pore-fracture structure and mechanical properties of coal is therefore crucial for the efficient development of coalbed methane^6,7,8,9.

Influence of different pre-fracturing fluids on pore-fracture structure

Acidizing treatments enhance coal permeability by dissolving minerals and clearing flow pathways^10,11, while also altering the surface morphology of coal^12,13. Hydrochloric acid (HCl) reacts with mineral components in coal matrices, but has little effect on clay minerals^14,15,16,17. Hydrofluoric acid (HF) exhibits significant dissolving action on iron ores, kaolinite, and calcite^18,19, but excessively high concentrations of HF may lead to the formation of insoluble substances²⁰. A combined HCl + HF acid system can more effectively corrode mineral particles, improve pore structure, and enhance fracture connectivity²¹. However, some studies suggest that acids may cause the migration of insoluble minerals, leading to pore throat blockage or the expansion of clay minerals, which in turn can reduce the permeability of the coal sample^22,23.

Oxidation treatments primarily improve permeability by swelling, dissolving, or breaking down organic components in coal, converting them into soluble fractions²⁴, thereby increasing permeability^25,26. Agents such as sodium hypochlorite and potassium permanganate show particular promise for enhancing coal oxidation and permeability²⁷. Liu et al.’s study showed that the macromolecular polymers in active water fracturing fluid tend to adsorb onto the surface of coal fractures, reducing coal permeability. In contrast, the oxidant can dissolve and disperse blockages, mitigate damage caused by the fracturing fluid, and enhance the hydrophilicity and permeability of the coal²⁸.

Furthermore, studies have explored the use of composite acid-oxidant solutions^29,30. However, a systematic, comparative analysis of acidizing versus oxidative fluids using core samples from a single deep wellbore—essential for controlling geological variability—remains lacking. Consequently, a clear comparison of the applicability and modification effects of different acid and oxidant fluids in deep coal seams is still needed.

Influence of different pre-fracturing fluids on mechanical properties

Fracturing fluids are known to alter the mechanical properties of coal³¹, leading to complex changes in its mechanical behavior and failure characteristics after fluid immersion³². Acidizing treatments lead to a reduction in fracture initiation pressure and enhance permeability more effectively³³. Anthracite soaked in acid-based fracturing fluids shows a significant decrease in fracture toughness and requires less fracture energy compared to samples treated with water-based fluids³⁴. Research by Zhang et al.³⁵ also confirms that the corrosive action of acids can degrade coal’s mechanical properties. Hydrofluoric acid (HF) demonstrates a more pronounced effect in reducing the compressive strength of coal³⁶.

In contrast, limited research has directly examined the influence of oxidant solutions on the mechanical properties of coal. Some studies have indicated that treatment with oxidizing agents reduces the fracture toughness of coal³⁷. But this strength reduction may also facilitate easier fracture closure in later stages.

While previous work has begun to explore how certain fluids affect the pore-fracture structure and mechanical properties of coal, a systematic analysis of the intrinsic relationship between mechanical alterations and pore-fracture evolution remains lacking. Comparative studies evaluating the interaction effects of different fluid types are also insufficient. To address these gaps, this study investigates deep coal seams with depths exceeding 2,500 m, and compares the effects of several pre-fracturing fluids on the pore-fracture structure and mechanical properties of coal. The potential relationships between changes in mechanical parameters and pore-fracture structural parameters were examined. The findings aim to provide theoretical support for hydraulic fracturing design in deep coalbed methane reservoirs.

Experimental methods

Coal sample and fluid Preparation

The experimental coal samples were obtained from the No. 8 Coal Bed in the Ordos Basin, China, at a depth of approximately 2,700 m. Industrial and elemental analyses of the coal samples are presented in Table 1. The volatile matter content is less than 6.96%, and the hydrogen content is 3.36%. Based on ASTM D388-23, the coal is classified as anthracite. Whole-rock and clay mineral analyses are provided in Table 2. The coal contains quartz, calcite, and mica, while the clay minerals are dominated by kaolinite, illite, and chlorite.

A portion of the core samples was wire-cut into cylindrical specimens with a diameter of 25 mm for subsequent testing. For triaxial compression tests, specimens with a length of 50 mm are labeled CC-X (C = Columnar, C = Compression, X = sample number, 1 ~ 25). For Brazilian splitting (tensile) tests, specimens with a length of 25 mm are labeled CT-X (C = Columnar, T = Tension, X = sample number, 1 ~ 25). Additional slices were prepared and labeled CS-0 through CS-5 for further characterization.

Based on the whole-rock analysis data of the coal and the current field application of pre-fracturing fluids for coal reservoirs, five pre-fracturing fluid systems were selected. The formula is shown in Table 3.

Table 1 Industrial analysis and elemental analysis of coal sample.

Table 2 Whole-rock and clay mineral analysis results of coal sample.

Table 3 Experimental scheme.

Core immersion experiment

The core samples were first dried in a vacuum oven at 60 °C for 72 h. After drying, each sample was placed in a vacuum chamber for degassing, followed by saturation with the corresponding pre-fracturing fluid for 6 h at room temperature. Upon completion of saturation, residual fluid on the sample surface and within the pores was removed using an ultrasonic cleaner in low-frequency mode. The samples were then dried again under vacuum at 60 °C for 72 h before subsequent testing. The overall experimental procedure is illustrated in Fig. 1.

Fig. 1

Experimental procedure.

Nuclear magnetic resonance (NMR) testing

Nuclear magnetic resonance (NMR) measurements were conducted on the core samples both before and after immersion, under both brine-saturated and post-centrifugation conditions. The brine saturation process utilized a 2% KCl solution. Cores were saturated under a pressure of 1 MPa for 48 h, after which NMR measurements were performed. Subsequently, the coal samples were subjected to centrifugation. Given the absence of a unified centrifugation standard and considering that the coal contains a significant amount of microfractures after immersion, the centrifugal speed was set at 5000 rpm for a duration of 6 h. After centrifugation, NMR testing was conducted again. After testing, the cores were dried again at 60 °C for 72 h, followed by vacuum extraction and saturation with pre-fracturing fluid (at room temperature, for 6 h). Once saturation was completed, the core surfaces and any residual fluid in the pores were cleaned using an ultrasonic cleaner in low-frequency mode, followed by drying. Finally, the cores underwent a second round of vacuum extraction and saltwater saturation, followed by a second set of NMR tests, both before and after centrifugation. The core saturation device is manufactured by Huatong Petroleum Machinery, the NMR instrument used is the Niumai Technology MINI MR NMR system.

Mechanical property testing

(1) Triaxial compressive testing.

The triaxial mechanical property tests on the core samples treated with the five different pre-fracturing fluids were performed using a GCTS-RTR-1000 rock mechanics testing system. A confining pressure of 3 MPa was applied and kept constant throughout the experiment. The axial pressure was then increased at a displacement-controlled rate of 0.1 mm/min until the sample failed. The deviatoric stress versus axial strain curve was plotted during the loading process. Subsequently, key rock mechanical parameters, including the compressive strength, Young’s modulus, and Poisson’s ratio, were calculated. For each pre-fracturing fluid, five parallel experiments were conducted.

(2) Brazilian split testing.

The tensile strength of the coal samples treated with the five different pre-fracturing fluids was measured using a GCTS-RTX-1000 triaxial mechanical testing system via the Brazilian disc method. In accordance with the standard procedure, a concentrated load was applied along the diameter of the cylindrical specimen. The specimen was placed between the upper and lower loading platens, ensuring its axis was perpendicular to the loading direction and precisely aligned at the center. The test was initiated by applying a small contact pre-load of 0.1 ~ 0.5kN. Subsequently, the load was increased at a constant displacement rate of 0.02 mm/min until tensile failure occurred. For each pre-fracturing fluid, five parallel experiments were conducted.

Results and discussion

T ₂ spectral results

The T₂ spectra of coal samples before and after treatment with the five pre-fracturing fluids are shown in Fig. 2. The horizontal axis represents the relaxation time (T₂), which correlates with pore-size distribution (longer T₂ corresponds to larger pore radii). The vertical axis shows signal amplitude, reflecting the relative abundance of pores at each corresponding size.

As observed in Fig. 2, the T₂ spectra of the coal samples display two to three peaks, with the most prominent peak occurring around 1ms. Peaks in the range of 10 ~ 1000ms show lower intensity, indicating that the coal is predominantly micro-porous, with fewer meso- and macro-pores. After immersion in the five pre-fracturing fluids, the T₂ spectrum show a notable increase in signal intensity compared to the untreated sample, accompanied by a rightward shift of some peaks. These changes suggest that fluid treatment not only increased the total pore volume but also contributed to pore enlargement.

Fig. 2

NMR T₂ Spectra of Five Coal Samples.

Based on transverse relaxation time (T₂) values, coal pores were categorized into micropores (0.01 ~ 10ms), mesopores (10 ~ 100ms), and macropores (100 ~ 10,000ms). As illustrated in Fig. 3, micropores are the most developed, followed by mesopores and macropores.

The area under the T₂ spectrum curve (S) was calculated to quantify pore-volume changes. Specifically: S_i refers to the area corresponding to micropores (T₂ = 0.01 ~ 10ms); S_e refers to the area corresponding to mesopores (T₂ = 10~100ms); S_a refers to the area corresponding to macropores (T₂ =100~10000ms); The relative change in each pore-type area after treatment with a fracturing fluid was expressed as the rate of change 𝑅, calculated as follows:

$${R_j}=({S_{j - After}}--{S_{j - Before}})/{S_{j - Before}}$$

(1)

The areas corresponding to micropore, mesopore, macropore, and total porosity—both before and after treatment with each fracturing fluid—along with their relative changes (R_j), are summarized in Table 4; Fig. 3.

Table 4 T₂ spectrum areas and change rates for different pore types.

After immersion in the five pre-fracturing fluids, the areas corresponding to micropores, mesopores, macropores, and total porosity all increased. Total porosity increased by 28%~37%, with macropores showing the largest relative growth (250%~1109%), followed by mesopores (59%~217%) and micropores (22%~31%). Specifically, slick-water (PAM) treatment on sample CC-1 resulted in the greatest total porosity increase (36.70%). The increase in mesopores and macropores was the highest among the five pre-fracturing fluids, at 216.18% and 1108.36%, respectively, while the increase in micropores was relatively high (30.57%). In contrast, HCl treatment (CC-2) produced the smallest total-porosity gain (28.55%) and the lowest micropore increase (22.03%), although mesopores and macropores still grew appreciably (72.88% and 820.41%, respectively). The HCl + HF acid (CC-3) induced moderate, balanced growth across all pore classes: micropores + 29.21%, mesopores + 67.18%, macropores + 362.89%, yielding a total porosity increase of 36.04%. NaClO treatment (CC-4) produced the highest micropore increase (30.72%) with moderate mesopore (70.58%) and lower macropore (323.72%) growth, for a total porosity rise of 35.38%. Finally, H₂O₂ (CC-5) showed the lowest overall enhancement, with increases of 27.26% (micropores), 59.07% (mesopores), 250.26% (macropores), and 32.21% (total porosity).

Fig. 3

T₂ spectrum areas and change rates for different pore types.

T ₁- T ₂ fluid signal 2D spectrum analysis

Following methods described in references^38,39,40, the two-dimensional NMR hydrogen spectrum of coal was divided into seven distinct regions, as illustrated in Fig. 4. Region I represents hydrogen in hard organic compounds, region II represents hydrogen in the hydroxyl group, region III represents inorganic pores or bound water in kerogen, region IV represents the bound water in organic micropores/nanopores, region V represents water in inorganic mesopores, region VI represents water in inorganic macropores or cracks, region VII represents methane in porous media.

Fig. 4

The partition diagram of T₁-T₂ 2D spectra³⁸.

The two-dimensional NMR spectra of coal samples before and after treatment with various pre-fracturing fluids are presented in Fig. 5. Before treatment, the signal of sample CC-1 was concentrated mainly in Region III, with only weak signals in Regions IV, V, and VI. The discontinuous signal distribution between Regions III ~ IV and V ~ VI indicated that hydrogen occurred primarily as bound or adsorbed water in inorganic pores and micro-/nanopores, while its presence in meso- and macropores was limited and pore connectivity was poor. After PAM (slickwater) treatment, the signal in Region III shifted upward and rightward, and the intensity in Region V increased, reflecting pore enlargement within the micropore range and a greater abundance of mesopores, along with partial improvement in micropore–mesopore connectivity.

For sample CC-2 (HCl immersion), the untreated spectrum showed strong signals in Regions III and IV and a weak signal in Region V, with no clear connection between these regions, indicating that the sample was dominated by micro/nanopores with relatively few mesopores, and that connectivity between micropores and mesopores was poor. After HCl treatment, the signal in Region V intensified and became directly connected to Region III, indicating a notable increase in mesopores and enhanced pore connectivity.

A similar pre-treatment pattern was observed for sample CC-3 (HCl + HF), with signals concentrated in Regions III and IV and weak connectivity to Region V. Following HCl + HF treatment, the intensity in Region V rose substantially and the connecting signals between Regions III and V strengthened markedly, pointing to increased mesopore abundance and greatly improved overall pore connectivity.

In sample CC-4 (NaClO) before treatment, signals were concentrated in Regions III and IV, with relatively weaker signals in Regions V and VI. After NaClO immersion, signals in Regions III and V shifted upward and rightward, the intensity in Region V increased, and connectivity to Region III improved. These results suggest a substantial increase in the number of micropores, a moderate increase in mesopores, and significantly improved connectivity between micropores and mesopores.

Finally, for sample CC-5 (H₂O₂), the untreated signals were also mainly in Regions III and IV, with weak signals in Regions V and VI. After H₂O₂ treatment, upward-rightward shifts in Regions III and V suggested a slight increase in both pore size and pore number within the coal.

Fig. 5

Two-dimensional NMR spectra of coal samples before and after treatment with different pre-fracturing fluids.

NMR parameter analysis

Porosity

The total porosity (φ_t), producible porosity (φ_p), and irreducible porosity (φ_i) of the coal core can be calculated based on the NMR T₂ spectra obtained before and after centrifugation. The calculation formulas are as follows³⁸:

$$\varphi_{t}\hspace{0.17em}=\varphi_{p}\hspace{0.17em}+\varphi_{i}\hspace{0.17em}=\hspace{0.17em}\frac{\text{V'}}{\text{V}}$$

(2)

$$\varphi_{p}\hspace{0.17em}=\frac{\text{FFI}}{\text{FFI+BFI}}\times\varphi_{t}\hspace{0.17em}$$

(3)

$$\varphi_{i}\hspace{0.17em}=\frac{\text{BFI}}{\text{FFI+BFI}}\times\varphi_{t}\hspace{0.17em}$$

(4)

In the equation, V’ represents the pore volume, V represents the sample volume, FFI is the free fluid index, and BFI is the bound fluid index. As shown in Fig. 6, taking the NMR T₂ spectrum of the pre-immersion CC-5 coal sample as an example, the total porosity, producible porosity, and irreducible porosity of the coal sample were calculated based on the FFI and BFI values. The results are presented in Fig. 7.

Fig. 6

Schematic diagram of the porosity and T₂ cutoff value for the pre-immersion CC-5 coal sample.

Fig. 7

Porosity and growth rate of different coal samples before and after treatment.

After treatment with the five pre-fracturing fluids, all coal samples showed an increase in total porosity, ranked as follows: CC-1 (+ 45.04%, PAM immersion) > CC-4 (+ 43.96%, NaClO immersion) > CC-3 (+ 43.14%, HCl + HF immersion) > CC-5 (+ 39.82%, H₂O₂ immersion) > CC-2 (+ 33.98%, HCl immersion). Specifically, PAM treatment of sample CC-1 produced the largest rise in total porosity, accompanied by a decrease in bound porosity and an increase in free porosity. In contrast, HCl treatment (CC-2) resulted in the smallest increase in total porosity, along with a reduction in free porosity and the most marked rise in bound porosity. The HCl + HF acid (CC-3) yielded the greatest enhancement in free porosity. This is attributed to the ability of HF to dissolve quartz and certain silicate minerals, thereby improving pore-throat connectivity and significantly increasing free porosity^41,42. For NaClO treated sample CC-4, oxidation and leaching processes improved connectivity between organic-matter pores, leading to a notable increase in free porosity and a favorable total porosity response⁴³. H₂O₂ treatment (CC-5) induced a moderate improvement in porosity.

Permeability

The permeability of the coal samples was calculated using the Timur-Coates model (T-C model):

$$\:{{k}_{T-C}}_{}=A{\varphi_{t}\hspace{0.17em}}^{m}{\left(\frac{FFI}{BFI}\right)}^{n}$$

(5)

In the equation, A, m, and n are constants that are related to the properties of the coal rock. In this study, the values of A, m, and n are set as A = 10^− 4, m = 4, n = 2, respectively.

Fig. 8

Permeability and growth rate of different coal samples before and after treatment.

The permeability and relative changes of coal samples treated with different pre-fracturing fluids are shown in Fig. 8. Except for the HCl treated sample (CC-2), which experienced a reduction in permeability, all other fluids increased permeability to varying degrees. The order of improvement was as follows: CC-3(+ 13561.54%, HCl + HF immersion) > CC-4(+ 4104.30%, NaClO immersion) > CC-1(+ 1461.94%, PAM immersion) > CC-5(+ 1043.55%, H₂O₂ immersion) > CC-2(-78.01%, HCl immersion). The permeability reduction in CC-2 can be attributed to the limited ability of HCl to enhance flow pathways, along with the potential migration of insoluble mineral particles that may block pore throats, as supported by earlier observations of decreased producible porosity and increased irreducible porosity (Sect. “Porosity”) and consistent with earlier reports²². These effects reduce pore connectivity and may convert some free pores into bound pores. In contrast, the HCl + HF acid produced the greatest permeability enhancement. The addition of HF enables the dissolution of quartz and silicate minerals that are resistant to HCl alone, substantially increasing both porosity and pore connectivity. NaClO also significantly improved permeability—more effectively than H₂O₂—which aligns with earlier findings⁴⁴. The strong oxidizing capability of NaClO can break down certain organic macromolecular structures in coal, leading to pore expansion, creation of new pores, and dissolution-swelling of organic matter, thereby enhancing flow pathways. By comparison, H₂O₂ exerts a milder oxidative effect, primarily modifying the coal surface and exerting a more limited influence on the microstructure and permeability compared with NaClO.

T ₂ cutoff

Fig. 9

T₂ cutoff values and decreases rates of different coal samples before and after treatment.

As shown in Fig. 6, the T₂ cutoff value is determined by drawing a horizontal line from the maximum of the post-centrifugation cumulative porosity curve; its intersection with the pre-centrifugation cumulative curve defines the boundary between movable and bound fluids. A lower T₂ cutoff value indicates a higher proportion of movable fluids. The T₂ cutoff values and their percentage changes after treatment with the five pre-fracturing fluids are presented in Fig. 9. The trend in these values aligns with the observed permeability changes. Specifically, the T₂ cutoff values for samples CC-3 (HCl + HF) and CC-4 (NaClO) decreased by 95.92% and 94.09%, respectively, indicating a substantial increase in the movable-fluid fraction. In contrast, the T₂ cutoff increased for the remaining three samples, in the following order: CC-2(HCl immersion, + 6345.29%) > CC-5(H₂O₂ immersion, + 91.69%) > CC-1(PAM immersion, + 17.18%). The dramatic rise in the T₂ cutoff for HCl treated coal (CC-2) corresponds to a sharp reduction in the movable-fluid fraction, which is consistent with its significant permeability decline noted earlier.

Surface morphology and roughness

Figure 10 shows scanning electron microscopy (SEM) images at 4000× magnification of the untreated coal sample (CS-0) and samples CS-1 to CS-5 after immersion in the five pre-fracturing fluids. The surface of the untreated sample (CS-0) is relatively smooth, with only sporadic mineral deposits and no visible large pores or fractures. In contrast, after fluid treatment, all samples display a mottled and roughened surface texture with localized. The most severe surface alteration occurred in samples CS-3 (HCl + HF) and CS-4 (NaClO), where larger corrosion pits are evident. Specifically, the HCl + HF treated sample (CS-3) exhibits numerous small but deep dissolution cavities.

Fig. 10

SEM morphologies of different coal samples magnified 4000 times.

Fig. 11

Local scanning electron microscopy of coal samples treated with different pre-fracturing fluids.

Figure 11(a)~(f) show representative scanning electron microscopy (SEM) images of the untreated coal sample (CS-0). The sample contains calcite, mica, and quartz, along with naturally developed pores and fractures, some of which are infilled with clay minerals such as chlorite and kaolinite. Figure 11(f) illustrates the coexistence of organic matter and aluminosilicate minerals. SEM images of the slickwater treated sample (CS-1) are presented in Fig. 11(g)~(i). Dissolution pores and fractures with smooth inner walls are observed in the coal matrix. In Fig. 12(g), larger dissolution pores can be seen, whose edges are filled with rosette-like chlorite. This suggests that the ammonium persulfate gel breaker in slickwater can oxidize minerals and organic matter within pores and fractures, resulting in clean, smooth pore walls. Pronounced acid corrosion is evident in the HCl treated sample (CS-2). As shown in Fig. 11(j), mineral surfaces exhibit a strawberry-like etched texture. In Fig. 11(k), pores appear to be blocked by detached debris. As shown in Fig. 11(l)~(o), etching effects are even more pronounced in the HCl + HF treated sample (CS-3), where numerous dissolution cavities of varying sizes are formed on the coal surface, along with surface cracking. Quartz grains show a frosted appearance, and the increased surface cleanliness reveals abundant clay minerals. Residual mineral forms and multiple deep dissolution pores of different sizes are visible after the composite acid treatment. As shown in Fig. 11(p)~(r), after NaClO treatment (CS-4), the mica structure remains largely intact, although some dissolution pores are visible. Under the oxidative action, part of the coal matrix swelled, generating numerous swelling cracks. In the H₂O₂ treated sample (CS-5), as shown in Fig. 11(s)~(u), no obvious corrosion of mica or similar minerals was detected. The oxidative etching by hydrogen peroxide partially removed organic matter from the coal surface, increasing porosity, but did not produce deep dissolution pores.

Fig. 12

3D morphology images of coal samples treated with different fracturing fluids.

Table 5 The average roughness R_a and root mean square roughness R_q of different coal samples.

Figure 12 presents three-dimensional surface topography obtained by atomic force microscopy (AFM) for the untreated coal sample (CS-0) and the five treated samples (CS-1 to CS-5). The average roughness (R_a) and root-mean-square roughness (R_q), calculated using Nanoscope Analysis software, are summarized in Table 5. R_a represents the arithmetic mean absolute deviation of surface height from the mean plane, whereas R_q reflects the root-mean-square deviation; lower values indicate a smoother surface. After treatment with the pre-fracturing fluids, both R_a and R_q increased substantially, indicating significant alteration of surface morphology. The extent of increase followed the order:

R_a

CS-3(101.82%, HCl + HF immersion) > CS-2(67.27%, HCl immersion) > CS-4(62.73%, NaClO immersion) > CS-5(53.64%, H₂O₂ immersion) > CS-1(4.55%, PAM immersion);

R_q

CS-3(108.03%, HCl + HF immersion) > CS-2(65.69%, HCl immersion) > CS-4(61.31%, NaClO immersion) > CS-5(54.74%, H₂O₂ immersion) > CS-1(8.76%, PAM immersion).

Acidizing fluids produced the most pronounced roughening, consistent with substantial mineral dissolution. The HCl + HF caused the strongest dissolution effect, leading to the highest roughness values. HCl and NaClO yielded comparable roughness increases, whereas H₂O₂ had a weaker effect. In contrast, PAM induced only minimal changes in surface topography.

Mechanical properties

Triaxial compressive testing

Fig. 13

Stress-strain curves of coal samples immersed in different pre-fracturing fluids.

Figure 13 shows the stress–strain curves of coal samples treated with the five pre-fracturing fluids. From these curves, key mechanical parameters—Young’s modulus, Poisson’s ratio, and uniaxial compressive strength—were derived and are summarized in Fig. 14.

Fig. 14

Mechanical parameters of coal samples treated with different pre-fracturing fluids.

The average Young’s modulus of coal samples treated with different pre-fracturing fluids followed an ascending order of: HCl + HF (1473.077MPa) < HCl (2444.570MPa)< NaClO (2738.597MPa) < PAM (2904.863MPa)< H₂O₂ (4590.290 MPa). Samples treated with the HCl + HF acid showed the lowest Young’s modulus, indicating the greatest loss of deformation resistance. This is likely attributed to extensive dissolution of cementing minerals (e.g., calcite, dolomite) and quartz, which weakens inter-particle bonding and softens the rock framework³⁶. Consequently, the coal becomes more deformable under stress. The similar Young’s modulus values for HCl- and NaClO treated samples suggest that both fluids effectively soften coal—HCl through mineral dissolution and NaClO via oxidative etching. In contrast, H₂O₂ treatment resulted in a relatively higher modulus, reflecting its more limited impact on mechanical integrity.

Poisson’s ratio decreased in the order: HCl + HF (0.465) > HCl (0.448) > PAM (0.316) > H₂O₂ (0.290) > NaClO (0.232). A higher Poisson’s ratio indicates increased plasticity and reduced brittleness. Acid treatments likely dissolved brittle minerals, leaving more ductile components, while the generation of micro-fractures promoted greater lateral expansion under compression. The lower Poisson’s ratio observed for oxidant treated samples suggests enhanced brittleness, probably due to sharp defects created by oxidative dissolution, making the coal more prone to brittle fracture.

Compressive strength increased as follows: HCl + HF (6.98 MPa) < NaClO (7.47 MPa) < HCl (10.41 MPa) < PAM (22.30 MPa) < H₂O₂ (25.01 MPa). The HCl + HF and NaClO caused the most pronounced strength reduction, followed by HCl. These fluids penetrated the coal matrix, enlarging existing pores and micro-fractures while creating new fracture networks, thereby significantly compromising structural integrity. In comparison, H₂O₂ had the smallest effect on compressive strength.

Brazilian split testing

The tensile strength of the coal samples, calculated from the failure load, is summarized in Fig. 15. Both acidizing and oxidizing treatments significantly reduced tensile strength, with values increasing in the following order: HCl + HF (0.94 MPa) < HCl (1.72 MPa) < NaClO (2.33 MPa) < H₂O₂ (3.92 MPa) < PAM (4.98 MPa). The HCl + HF and HCl alone caused the greatest reduction in tensile strength. This is attributed to their strong erosive and dissolving action on the coal matrix, which reduces particle cementation and weakens intergranular bonding. Chemical enlargement of pre-existing micro-fractures and pores further diminishes overall tensile resistance.

In comparison, NaClO and H₂O₂ had less pronounced effects. These oxidants primarily degrade aromatic organic macromolecules that contribute to coal’s structural integrity. Due to the dense nature of coal, oxidation proceeds more slowly than acid-mineral reactions, leading to a comparatively smaller decrease in tensile strength.

The lowered tensile strength implies that, during hydraulic fracturing, treated coal becomes more prone to tensile failure, requires lower fracture initiation pressure, allows easier fracture propagation, and ultimately promotes the formation of more complex fracture networks.

Fig. 15

Tensile strength of coal rock under different pre-fracturing fluid treatments.

Correlation analysis

To explore the relationship between changes in mechanical properties and pore-fracture structure, a Kendall correlation analysis was conducted between the mechanical parameters and pore-fracture metrics. The results are shown in Fig. 16. The analyzed variables include Young’s modulus (E), Poisson’s ratio (ν), compressive strength (R_c), tensile strength (R_t), total porosity (φ_t), irreducible porosity (φ_i), free porosity (φ_p), permeability (k), and average roughness (R_a).

Fig. 16

Correlation Analysis of Coal and Rock Mechanical Parameters and Pore-Fracture Parameters.

Young’s modulus (E) of coal exhibits a significant negative correlation with total porosity (φ_t) and average roughness (R_a). Increased porosity reduces the effective load-bearing skeleton and introduces stress concentrations, while higher roughness exacerbates local stress concentration and promotes crack initiation, thereby weakening interfacial strength. Together, these factors make the coal more deformable under load, resulting in larger strain for a given stress and, consequently, a lower Young’s modulus (E = σ/ε). This relationship underscores how macroscopic mechanical behavior is governed by microstructural defects.

Poisson’s ratio (ν) is positively correlated with total porosity (φ_t). A higher porosity implies a greater proportion of weak, compliant pore structures. Under axial compression, the collapse and closure of these pores not only contribute to axial strain but also force the solid framework to expand laterally, thereby increasing the measured Poisson’s ratio.

In contrast, compressive strength (R_c) shows no significant correlation with the pore-fracture parameters examined here.

Tensile strength (R_t) displays a significant negative correlation with average roughness (R_a). Higher surface roughness implies the presence of more and sharper stress concentration points, a smaller effective load-bearing area, and a higher density of initial defects. Under tensile loading, these features promote earlier micro-crack initiation and propagation at lower macroscopic stress, thereby reducing the overall tensile strength.

Limitations

Despite the clear trends observed, this study has several limitations that should be considered.

(1) Due to the complexity of sample preparation and experimental protocols, a limited number of core samples were used. Notably, only a single sample per treatment group underwent NMR analysis, and the number of parallel samples for mechanical testing was small. This limits the statistical power of the findings and necessitates caution in generalizing the conclusions. Furthermore, the laboratory core-scale experiments may not fully replicate in-situ reservoir conditions involving stress, temperature, and fluid flow.

(2) The determination of the NMR T₂ cutoff value is sensitive to centrifugation standards. The use of the empirical Timur-Coates model for permeability estimation, which carries an inherent order-of-magnitude uncertainty.

(3) The sample size in this study limits the ability to perform more complex multivariate modeling. The findings from the current variable correlation analysis should therefore be validated in the future using multivariate statistical methods on a larger sample base, with other potential factors controlled for.

Conclusions

Based on NMR, SEM, AFM, and mechanical testing, this study systematically investigated the effects of five pre-fracturing fluids (PAM, HCl, HCl + HF, NaClO, H₂O₂) on the pore-fracture structure and mechanical properties of deep coal. The key findings are as follows:

(1) Porosity enhancement varied among the fluids: PAM showed the highest total porosity increase (36.70%) with broad-spectrum pore growth. HCl mainly expanded macropores and mesopores but performed poorly in micropores. HCl + HF uniformly enlarged pores of all sizes. NaClO excelled in creating micropores, whereas H₂O₂ had the weakest effect on meso- and macropores.

(2) Permeability improvement followed the order: HCl + HF (+ 13,561.54%) > NaClO (+ 4,104.30%) > PAM (+ 1,461.94%) > H₂O₂ (+ 1,043.55%) > HCl (-78.01%). HCl alone reduced permeability, likely due to pore-throat blockage by insoluble residues.

(3) Microstructural changes observed via SEM included: smooth dissolution pores after PAM; significant corrosion with debris after HCl; deep, varied dissolution cavities after HCl + HF; swelling cracks induced by NaClO; and mild surface etching after H₂O₂.

(4) Mechanical properties were significantly altered: HCl + HF caused the greatest reduction in compressive strength, Young’s modulus, and tensile strength, but produced the highest Poisson’s ratio. H₂O₂ best preserved mechanical strength and stiffness. Acidizing fluids generally weakened coal more than oxidants.

(5) Correlation analysis indicated that Young’s modulus (E) decreased with higher φ_t and R_a; Poisson’s ratio (ν) increased with φ_t; tensile strength (R_t) decreased with R_a; and compressive strength (R_c) showed no significant correlation with pore-fracture parameters.