Synergistic Modification of Static Chemical Agent with Fly Ash and Hemihydrate Gypsum for Mechanically Compatible Gas Drainage Borehole Sealing

Abstract

Borehole sealing quality plays a critical role in coal seam gas extraction efficiency, while the mechanical mismatch between conventional cement-based sealing materials and surrounding coal masses often leads to interface damage and sealing failure under mining-induced stress conditions. To address this issue, a modified static chemical agent (MSCA) was developed by incorporating fly ash (FA) and hemihydrate gypsum (HG) into a static chemical agent (SCA) system. The hydration behavior, expansion characteristics, mechanical properties, and engineering applicability of MSCA were systematically investigated through laboratory experiments, microscopic analyses, numerical simulations, and field industrial tests. The results indicate that HG significantly accelerates the hydration reaction and promotes the early development of expansion pressure of MSCA. The maximum expansion pressure of MSCA increases linearly with SCA content and quadratically with HG content. XRD and SEM analyses reveal that the hydration products mainly consist of Ca­(OH)₂, calcium silicate hydrate (C–S–H), calcium aluminate hydrate (C–A–H), dihydrate gypsum (DG), and ettringite (AFt). Increasing FA and HG contents facilitates the formation of C–S–H gel and interwoven AFt crystal networks, resulting in a denser microstructure and improved mechanical performance. The splitting strength of MSCA increases significantly with decreasing SCA content and increasing HG content, whereas the expansion pressure exhibits an opposite trend. Among all tested formulations, the D1MSCA achieves the optimal balance between expansion pressure and mechanical strength, with a splitting strength of 1018 kPa and an expansion pressure of 5.7 MPa, which are close to the mechanical properties of coal seams and the field grouting pressure conditions. Numerical simulation results demonstrate that the mechanical compatibility between the sealing material and the coal seam is a key factor controlling interface damage under mining-induced stress. A sealing material with an elastic modulus close to that of the coal seam can effectively alleviate stress concentration around the borehole and suppress crack propagation. Field industrial tests further confirm that MSCA-sealed boreholes maintain significantly higher gas extraction concentrations than conventional cement-sealed boreholes during the entire mining process. Under strong mining-induced stress conditions, the gas extraction concentration of MSCA-sealed boreholes remains 31%–50% higher than that of conventional cement-sealed boreholes. The proposed MSCA provides an effective approach for improving borehole sealing quality and gas extraction performance in underground coal mines under mining disturbance conditions.

1. Introduction

Coal mining in China has entered the stage of deep mining, where low permeability and high gas pressure in coal seams have become increasingly pronounced, making methane control a critical constraint on mining efficiency and safety. − At present, gas extraction is the primary technical approach for preventing and controlling coal mine gas disasters. Pre-extraction of gas from coal seams effectively reduces gas concentrations in roadways and working faces during mining, thereby mitigating the risk of coal and gas outbursts. Borehole sealing is a crucial step in coal seam gas extraction, and its quality directly determines extraction efficiency. Poor sealing quality can lead to the formation of gas leakage pathways around boreholes, resulting in reduced gas concentration and a diminished extraction radius, which poses significant risks to safe coal mine production.

At present, cement grouting is the predominant technique for sealing gas extraction boreholes in coal mines. Owing to its advantages, including readily available materials, low cost, and a mature construction process, this method has been widely adopted in engineering practice. , However, extensive engineering applications and studies have demonstrated that cement-based sealing suffers from two inherent limitations that significantly compromise seal stability and durability. First, there exists a pronounced mismatch in mechanical properties between coal seams and the solidified cement matrix. , Consequently, stress concentration readily develops at the coal–cement interface under stress disturbances, leading to interfacial debonding and eventual sealing failure, as illustrated in Figure . Second, cement slurry is prone to volumetric shrinkage during solidification, leading to the formation of micro-gaps between the cement seal and the borehole wall. , These gaps can progressively enlarge under stress disturbances and eventually develop into continuous gas leakage pathways, resulting in sealing failure. To address this issue, considerable efforts have been devoted to the development of improved sealing materials. Zou et al. modified sulfate cements with nano-magnesium oxide to enhance the compactness of the solidified matrix. Liu et al. developed a viscoelastic self-healing suspension sealing material (VEHS) incorporating latex emulsion, superabsorbent polymer, fly ash, and nano-montmorillonite, which improved the flowability of the grouting material. Liu et al. enhanced the mechanical strength of cement-based materials by incorporating polypropylene fiber (PPF), poly­(vinyl alcohol) fiber (PVAF), and nano-silica (NS). Lou et al. proposed a novel sealing material based on xanthan gum, Na-CMC, polyacrylamide, and bentonite, and field application in the Tangshan Coal Mine demonstrated a 20% increase in gas extraction concentration. Despite these improvements, existing materials still fail to resolve the fundamental issue of mechanical mismatch between the cement matrix and coal seams. As a result, interfacial cracking may still occur under stress disturbances. In addition, many of these materials suffer from high costs and complex construction procedures, which limit their large-scale application in underground coal mines. − Therefore, there is an urgent need to develop a novel sealing material that exhibits mechanical compatibility with coal seams, eliminates volumetric shrinkage, and maintains cost-effectiveness, thereby overcoming the limitations of conventional technologies and improving the reliability of gas extraction.

1.

Sealing failure of gas extraction boreholes under mining-induced influence.

Static chemical agent (SCA) is an inorganic material primarily composed of calcium oxide (CaO). Its key characteristic is its ability to undergo significant volumetric expansion and release heat upon hydration with water, generating expansion pressures of up to tens of megapascals. , Owing to its high expansibility, SCA has been widely applied in fields such as rock breaking and concrete demolition. , In addition, its hydration process produces no harmful gases and the final reaction products are environmentally benign, indicating its potential for use in sealing underground coal mine boreholes. , However, SCA cannot be directly applied for sealing purposes. First, its hydration products are mainly in the form of powders or loosely bound particles with negligible cementitious strength, making it incapable of forming a stable sealing structure. Second, the expansion pressure generated by SCA is excessively high and far exceeds the bearing capacity of coal seams. Direct application may damage the borehole wall and thereby aggravate gas leakage. Therefore, composite modification of SCA is necessary to optimize its performance. While retaining its expansion capability, the modified material should also develop sufficient cementitious strength to satisfy the engineering requirements of borehole sealing in underground coal mines.

Based on the above background, this study proposes a composite modification strategy for SCA using fly ash (FA) and hemihydrate gypsum (HG), aiming to develop a novel modified static chemical agent (MSCA) for gas extraction borehole sealing. FA is an industrial by-product generated in large quantities from coal-fired power plants. Its primary components include silicon dioxide (SiO₂), mullite (2Al₂O₃·SiO₂), and a small amount of aluminum oxide (Al₂O₃). It is not only abundant and low-cost but also facilitates waste utilization, aligning with the principles of energy conservation and environmental protection. , HG forms calcium sulfate dihydrate (CaSO₄·2H₂O) upon hydration, exhibiting certain binding properties. Moreover, it can react with calcium hydroxide (Ca­(OH)₂), a key hydration product of SCA, as well as with the reactive components in FA, to produce hydration products with cementitious strength. − By optimizing the proportions of SCA, FA, and HG, the inert filling effect of FA and the reaction-regulating role of HG effectively reduce the expansion pressure generated during SCA hydration, thereby preventing damage to the borehole wall. Through the synergistic hydration of the three components, a composite material with considerable strength is formed, whose mechanical properties are comparable to those of coal seams, thus addressing the deformation incompatibility associated with conventional materials. In addition, the hydration of this composite system does not induce volumetric shrinkage; instead, the expansion of SCA enables a tight fit with the borehole wall, thereby eliminating sealing gaps at the source. Furthermore, all constituent materials are inorganic minerals characterized by low cost and simple construction procedures, without requiring additional complex equipment, which facilitates large-scale application in underground coal mines.

This study focuses on the performance optimization and working mechanism of MSCA. By designing a series of mix ratios for SCA, FA, and HG, we systematically investigated the hydration temperature evolution, expansion pressure development, and splitting strength of the MSCA reaction products to elucidate the influence of each component on the macroscopic performance of the system. Combined with microscopic characterization techniques, including X-ray diffraction (XRD) and scanning electron microscopy (SEM), the mineral composition and microstructural features of the hydration products are analyzed to clarify the synergistic interaction mechanisms among the three components. Furthermore, the quantitative relationships among the MSCA component ratios, expansion pressure, and splitting strength are established, providing a theoretical basis and technical support for the development of sealing materials for gas extraction boreholes in underground coal mines. This study not only effectively addresses the sealing failure associated with conventional cement-based technologies and improves the safety and efficiency of gas extraction, but also enables the resource utilization of industrial waste, thereby offering significant engineering value and environmental benefits.

2. Methods

2.1. Materials Preparation

The FA used in this study was obtained from Xinmi Power Plant, Henan, China; the HG was sourced from Xingtai Dongsheng Building Materials Co., Ltd., China; and the SCA was supplied by Blasting Technology Co., Ltd. To determine the mineral composition of each material, XRD analysis was conducted. The XRD measurements were performed using a SmartLab diffractometer (Rigaku Corporation, Japan). The instrument operated at an output power of 3 kW with a focal spot size of 1 mm × 10 mm. The scanning range was set from 5° to 90° (2θ) with a scanning rate of 10°/min. The obtained diffraction data were analyzed using Jade 6.0 software, and the results are presented in Figure . The XRD patterns indicate that SCA is primarily composed of CaO; FA mainly consists of SiO₂ and 2Al₂O₃·SiO₂, along with a small amount of Al₂O₃; and HG is predominantly composed of calcium sulfate hemihydrate (CSH).

2.

XRD test results of SCA, FA, and HG.

2.2. Experimental Scheme

To investigate the effects of different component ratios on the properties of MSCA, four groups of experiments (Groups A–D) were designed, as summarized in Table . Each group was further divided into three subgroups (e.g., Group A included subgroups A1–A3). The values listed in Table represent the mass fractions of each component relative to the total mass of MSCA. In subgroups A1–A3, the SCA content was fixed at 50 wt % of the total MSCA mass, while the mass fractions of FA and HG were varied. Similarly, in Groups B–D, the SCA content remained constant within each group, whereas the proportions of FA and HG differed among the subgroups. The mass of MSCA dry powder was fixed at 200 g for all test groups, and the water dosage was uniformly 70 g across each group, corresponding to a water-to-MSCA powder mass ratio of 35%.

1. Experimental Schemes for Different Material Ratios.

	Group A			Group B			Group C			Group D
Proportion	A1	A2	A3	B1	B2	B3	C1	C2	C3	D1	D2	D3
SCA	0.5	0.5	0.5	0.4	0.4	0.4	0.3	0.3	0.3	0.2	0.2	0.2
FA	0.4	0.3	0.2	0.3	0.2	0.1	0.4	0.3	0.2	0.4	0.3	0.2
HG	0.1	0.2	0.3	0.3	0.4	0.5	0.3	0.4	0.5	0.4	0.5	0.6

2.3. Experimental Setup

Both SCA and HG release heat during hydration, which influences the reaction process. Therefore, the temperature evolution of MSCA during hydration was first investigated. Subsequently, the expansion pressure generated during MSCA hydration was measured using a steel tube strain test. Finally, the hydration products of MSCA were cut into Brazilian disc specimens to determine their splitting strength.

2.3.1. Test for Hydration Temperature of MSCA at Room Temperature

Figure illustrates the experimental setup for measuring the hydration temperature evolution of MSCA. An Agilent 34970A data acquisition system (Keysight Technologies, USA) equipped with K-type thermocouples was used to record temperature changes during hydration. The system has a temperature measurement accuracy of ±1 °C and a sampling interval of 10 s. The thermocouples (model WRNK-191, Tianjin Hongrun Instrument Technology Co., Ltd., China) have a measurement error within ±1.5 °C. The experimental procedure was as follows:

• (1)A total of 200 g of MSCA mixture was prepared according to the mass ratios listed in Table and placed into a 250 mL plastic beaker.
• (2)The components were thoroughly mixed using a mechanical stirrer.
• (3)A total of 70 g of deionized water was added, and the mixture was stirred until homogeneous.
• (4)The K-type thermocouple was inserted into the center of the slurry and fixed in place using adhesive tape.
• (5)The data acquisition system was activated to continuously record the temperature evolution of the MSCA slurry during hydration.

3.

Experimental system for monitoring temperature during the reaction process of MSCA.

2.3.2. Test for Expansion Pressure of MSCA

The steel tube strain method has been widely used to measure the expansion pressure of SCA; therefore, it was also adopted in this study to determine the expansion pressure of MSCA. The pressure-measuring steel tube was designed to satisfy the thin-walled cylinder condition, as described by eq ().

$$\frac{r_{\mathrm{o}\mathrm{u}\mathrm{t}}-r_{\mathrm{i}\mathrm{n}}}{r_{\mathrm{i}\mathrm{n}}}>0.05$$ 1

Where r _out and r _in denote the outer and inner diameters of the steel tube, mm, respectively. The steel tube used in the experiment had an inner diameter of 40 mm and a wall thickness of 2 mm, meeting the thin-walled requirement. A schematic diagram of the steel tube strain test system is shown in Figure . Considering that MSCA releases a large amount of heat during hydration and that the hydration temperature varies with different mix ratios, a water bath was employed to minimize temperature-induced effects on the measurement results. A steel base plate with a diameter of 100 mm and a thickness of 50 mm was welded to the bottom of the pressure-measuring steel tube to ensure vertical stability in the water bath. Prior to testing, the leak-tightness of the steel tube was verified to ensure weld integrity and to prevent water infiltration due to welding defects. After the leak-tightness test, three strain gauges were attached to the outer surface of the steel tube at 20 mm intervals starting from the bottom. The strain gauges (model BX120-20AA) had a resistance of 120.2 ± 0.3 Ω and a nominal gauge factor of 2.15 ± 1%. After installation, hot-melt adhesive was applied to waterproof the strain gauges. Once the adhesive had cooled and solidified, the instrumented steel tube was placed in the water bath for testing.

4.

Steel tube strain system for testing expansion pressure of MSCA.

After assembling the experimental system, 270 g of MSCA slurry was prepared according to the mix proportions listed in Table and poured into the pressure-measuring steel tube. A DH3812 static strain acquisition system (Donghua Testing Technology Co., Ltd., China) was used to record the strain data at a sampling frequency of 2 Hz. The expansion pressure of MSCA (P _in) was calculated based on the measured strain using eq ().

$$P_{\mathrm{i}\mathrm{n}}=\frac{({r_{\mathrm{o}\mathrm{u}\mathrm{t}}}^2-{r_{\mathrm{i}\mathrm{n}}}^2)E{\epsilon }_{\mathrm{t}}}{2{r_{\mathrm{i}\mathrm{n}}}^2}$$ 2

Where r _in and r _out denote the inner and outer diameters of the steel tube, mm; E and ε_t are the elastic modulus and expansion strain of the steel tube, respectively.

3. Results and Analysis

3.1. Temperature Change during MSCA’s Hydration Process

Figure presents the temperature evolution of MSCA during hydration, along with the morphologies of the post-reaction products under laboratory conditions. As shown in Figure (a), the temperature evolution trends of the three subgroups (A1–A3) were generally consistent despite differences in mix proportions. Within the first 1.5 h after slurry preparation, the temperature gradually increased from the ambient temperature of 34 °C to approximately 40 °C. Thereafter, the hydration temperature rose rapidly and reached a peak of about 110 °C at approximately 2.5 h. When the temperature exceeded 100 °C, the unreacted free water in the slurry vaporized, leading to a continuous increase in internal pressure and ultimately causing blowout of the MSCA slurry in Group A from the beaker. This phenomenon is consistent with the blowout behavior observed in conventional SCA systems. , After blowout, the thermocouple became exposed to air, resulting in an abrupt drop in the recorded temperature to ambient levels. Based on the timing of this sudden temperature decrease, the blowout times of the three subgroups were found to be similar. A magnified view of the temperature curves between 2.4 and 2.5 h indicates that the hydration rates followed the order A3 > A2 > A1. This difference is attributed to variations in FA and HG contents among the subgroups. Specifically, subgroup A1 contained a higher proportion of FA and a lower proportion of HG, whereas subgroup A3 had the highest HG content, indicating that increasing HG content accelerates the hydration reaction.

5.

Temperature changes of MSCA during hydration reaction under room temperature condition.

The images in Figure also show the morphologies of the post-reaction MSCA products. The residual materials in the beaker were mainly in the form of fine powders or loosely aggregated particles, with no intact rock-like structure formed. This behavior can be explained as follows. First, the rapid temperature rise during hydration in Group A induced blowout, which disrupted the subsequent reaction process. Second, due to the high SCA content in Group A, a large amount of CaO underwent hydration to form Ca­(OH)₂, generating substantial expansion pressure. However, the formation of cementitious products, such as calcium silicate hydrate (C–S–H) and calcium aluminate hydrate (C–A–H), was limited and insufficient to resist the expansion pressure. As a result, the final products remained in a powdery or loosely particulate state.

The temperature evolution of Group B MSCA during hydration is shown in Figure (b). Significant differences were observed among the three subgroups (B1–B3). The temperature evolution of each subgroup can be divided into three stages: a slow temperature rise stage, a rapid temperature rise stage, and a temperature decline stage. During the initial stage, the temperature variations of the three subgroups were similar. In the rapid temperature rise stage, subgroup B3 exhibited the earliest temperature increase and reached a maximum temperature of 97 °C at 1.87 h. Subgroup B2 reached its peak temperature of 89.4 °C at 2.12 h, while subgroup B1 attained a maximum temperature of 81.4 °C at 2.52 h. These results indicate that subgroup B3 had a shorter heating duration and a higher peak hydration temperature.

The results of Group B further demonstrate that, under a constant SCA content, increasing the HG content (i.e., decreasing the FA content) shortens the heating duration and elevates the peak hydration temperature. In addition, the temperature evolution curves exhibit a clear inflection point, indicating that the heating process is nonlinear. This behavior is attributed to the different exothermic characteristics of CaO in SCA and CaSO₄·0.5H₂O in HG. Specifically, the hydration of CaO is highly exothermic and rapid, whereas CaSO₄·0.5H₂O undergoes a relatively mild crystallization reaction with lower heat release over a longer duration. The post-reaction morphologies of MSCA indicate that no blowout occurred in Group B. Under the expansion effect of MSCA, the plastic beakers underwent significant deformation, and the final products formed intact, rock-like masses.

The temperature evolution of Group C MSCA during hydration is shown in Figure (c). Compared with Group B, the hydration initiation time of Group C was further shortened, while maintaining the same trend that higher HG content corresponds to a shorter initiation time and a higher peak hydration temperature. However, significant differences were observed in the temperature decline stage among the subgroups. For subgroup C3, the temperature reached a maximum of 54.1 °C at 0.96 h, followed by a decrease. After 1.5 h, the rate of temperature decline slowed markedly, with only a 1 °C decrease between 1.5 and 2 h, after which the temperature gradually decreased. In contrast, subgroup C2 exhibited a more complex temperature evolution: the temperature increased during 0–1.12 h, decreased during 1.12–1.5 h, and then increased again between 1.5 and 2.45 h, reaching a maximum of 51 °C, before gradually declining to ambient temperature. A similar trend was observed for subgroup C1. This multi-stage temperature behavior can be attributed to the coupled hydration reactions of CaO and CaSO₄·0.5H₂O. Before reaching the peak temperature, both components hydrated rapidly, with the exothermic reaction of CaO dominating the temperature rise. After the peak, the hydration rate of CaO decreased, and its heat release was no longer sufficient to sustain the temperature increase, leading to a temperature decline. Subsequently, CaSO₄·0.5H₂O continued to hydrate and release heat, resulting in a slower temperature decrease and even a secondary temperature rise in subgroups C1 and C2. At the later stage, the exothermic effect of CaSO₄·0.5H₂O diminished, causing the temperature to decrease again. As shown in Figure (d), Group D exhibited a further reduction in hydration duration and peak temperature, with the temperature curves displaying a single unimodal pattern. The MSCA in this group formed intact rock-like masses after hydration.

The temperature evolution characteristics of MSCA with different compositions were further analyzed based on the temperature change. According to the temperature–time curves shown in Figure , the moment at which the temperature increased by more than 0.1 °C within 2 min was defined as the initial time (t ₀). The time corresponding to the first peak temperature was defined as the peak time (t _p). The time interval between t ₀ and t _p, denoted as Δt, represents the duration of the rapid reaction stage of MSCA. Since the peak temperature of Group A exceeded 100 °C and resulted in blowout, this group was excluded from further analysis. As shown in Figure (a), both t ₀ and t _p followed the order B3 < B2 < B1, indicating that increasing the HG content advanced the onset of hydration and shortened the time to peak temperature. A similar trend was observed for Group C. The statistical analysis of Δt for Groups B and C shows that the duration of the rapid reaction stage decreases with increasing HG content. When the mass fraction of HG (φ_HG) increased from 0.3 to 0.4, the duration of the rapid reaction stage shortened more notably. In contrast, Group D exhibited nearly identical reaction processes among its subgroups, which can be attributed to its low SCA content and the corresponding reduced heat release during hydration. A comparison among Groups B, C, and D indicates that the duration of the rapid reaction stage follows the order: Group B > Group C > Group D. Overall, at a constant SCA content, increasing the HG proportion accelerates the onset of hydration heat release and shortens the duration of the rapid reaction stage.

6.

Statistics of reaction times and peak temperatures of MSCA.

Figure (b) presents the statistical results of peak temperatures for MSCA with different compositions, along with the linear relationships between HG mass fraction and peak temperature. For all three groups, the peak temperature increased linearly with increasing HG content at a constant SCA fraction. The slopes of the fitted lines reflect the sensitivity of peak temperature to HG content. For Group B, the slope was 78, indicating that when the SCA mass fraction was fixed at 0.4, the peak temperature increased by approximately 7.8 °C for every 0.1 increase in HG mass fraction. The corresponding slopes for Groups C and D were 40.5 and 35, respectively, demonstrating that the influence of HG content on peak temperature weakened as the SCA content decreased. Overall, the hydration heat release of CaO in SCA is the dominant factor governing the rapid temperature rise of MSCA. Increasing the HG content promotes the hydration process of SCA, which is reflected by a higher peak temperature and a shorter duration of the rapid reaction stage.

3.2. Expansion Characteristics of MSCA

3.2.1. Expansion Pressure Evolution of MSCA

The expansion pressure of MSCA measured using the steel tube strain method is presented in Figure . Three strain gauges were installed along the height of the steel tube, and the expansion pressures derived from the bottom, middle, and upper positions are denoted as 1, 2, and 3, respectively. For example, A1-1 represents the expansion pressure of subgroup A1 calculated from the bottom strain gauge.

7.

Expansion pressure curves of MSCAs for (a) D1 MSCA, (b) D2 MSCA, (c) D3 MSCA, (d) D4 MSCA.

Figure (a) shows the evolution of expansion pressure with time for Group A. No data were obtained for A1-2 and A3-3 due to data acquisition issues. The expansion pressures measured at different positions along the steel tube were generally consistent within each subgroup, indicating a relatively uniform stress distribution. However, significant differences were observed among the subgroups. The expansion pressure in subgroups A1, A2, and A3 began to develop at approximately 7.2, 5.1, and 4 h, respectively, indicating that increasing HG content significantly shortened the onset time of expansion pressure development. Compared with the temperature results in Figure (a), the water bath conditions delayed the initiation of the expansion process. The evolution curves show that the expansion pressure increased rapidly at the early stage of hydration and then tended to stabilize with time. After 30 h of monitoring, the expansion pressure of all subgroups continued to increase, although at a reduced rate. The average expansion pressures of subgroups A1, A2, and A3 at 30 h were 11.5, 11.9, and 12.3 MPa, respectively. At the later stage (after 25 h), subgroup A1 exhibited a relatively higher growth rate compared to the others.

To further compare the evolution characteristics, the initial segments prior to pressure development were removed and the curves were replotted, as shown in the inset of Figure (a). In the early stage after the onset of expansion, subgroup A3 showed the fastest increase in expansion pressure, reaching 3.6 MPa within 0.5 h. A clear inflection point was observed, after which the rate of pressure increase decreased significantly. Similar inflection behavior was observed in subgroups A2 and A1, with turning points occurring at approximately 0.5 h after the onset of expansion. However, the expansion pressures at these turning points differed, with values of 2.2 and 1.5 MPa for subgroups A2 and A1, respectively. These results indicate that increasing HG content accelerates the initial rate of expansion pressure development and enhances the ultimate expansion capacity of MSCA. The rapid increase in expansion pressure before this point is primarily attributed to the promotion of SCA hydration by Ca²⁺ ions, which accelerates the early-stage reaction kinetics.

Figure (b) shows the evolution of expansion pressure for Group B. With a reduction in SCA content, the onset of expansion pressure development was significantly advanced, and the expansion pressure tended to stabilize after approximately 30 h. The maximum expansion pressures of subgroups B1, B2, and B3 were 10.2, 11.1, and 12.9 MPa, respectively. As shown in Figure (c,d), the expansion pressures of Groups C and D after 30 h reached 8.9, 9.8, and 10.4 MPa for Group C, and 5.7, 7.0, and 7.8 MPa for Group D, respectively. These results indicate that a decrease in SCA content leads to a reduction in expansion pressure. Under a constant SCA content, increasing the HG fraction results in higher expansion pressure. By comparing the temperature evolution in Figure , it can be observed that the water bath condition influenced the onset of MSCA hydration. However, as the SCA content decreased, the influence of the water bath on the reaction process gradually weakened.

3.2.2. Effect of Material Proportion on Maximum Expansion Pressures of MSCAs

Variations in material composition led to significant differences in the expansion pressure of MSCA. To quantify the effects of individual components, both linear and nonlinear regression analyses were performed using Origin software to examine the relationships between SCA content (φ_SCA), HG content (φ_HG), and maximum expansion pressure (P _max).

The fitting results are summarized in Table . Both SCA and HG contents showed positive correlations with the maximum expansion pressure. Compared with the linear model, the nonlinear model exhibited improved statistical performance, as indicated by a lower P-value and reduced multicollinearity. The adjusted R ² values for the linear and nonlinear models were 0.955 and 0.969, respectively. In addition, the reduced chi-square values were 0.223 and 0.152, respectively. These results demonstrate that the nonlinear model provides a significantly better fit to the experimental data. The relationship between MSCA expansion pressure and material composition is expressed in eq (), indicating that the maximum expansion pressure increases linearly with SCA content and shows a quadratic dependence on HG content.

$$P_{\max }=22.67\times {\varphi }_{\mathrm{SCA}}+14.68\times {\varphi }_{\mathrm{HG}}^2$$ 3

2. Fitting Results.

	Fitting factor	Value	Standard error	t-Value	P-value	Correlation
P max = a × ϕSCA + b × ϕHG	a	20.813	0.697	29.859	4.15 × 10–11	0.7168
	b	7.712	0.693	11.132	5.9 × 10–7	0.7168
P max = a × ϕSCA + b × ϕHG 2	a	22.671	0.460	49.067	2.98 × 10–13	0.55936
	b	14.678	1.077	13.634	8.72 × 10–8	0.55936

3.3. Reaction Products of MSCAs

3.3.1. Mineral Compositions of MSCA

Figure presents the XRD patterns of the hydration products of MSCA. The testing equipment and procedures were consistent with those described in Section . As shown in Figure (a), the main hydration products of Group A MSCA were Ca­(OH)₂ and CaSO₄·2H₂O. The diffraction peak intensities corresponding to the (020), (021), and (150) crystal planes of DG in subgroup A3 were significantly higher than those in subgroups A1 and A2. This is attributed to the higher HG content in subgroup A3, which promoted the formation of DG during hydration. DG can further react with Ca­(OH)₂ and the reactive Al₂O₃ in FA to form AFt. A comparison of the diffraction peak intensities of the (100) and (110) planes of AFt indicates that the AFt content follows the order A1 > A2 > A3. This trend is consistent with the FA content in Group A, which also follows the order A1 > A2 > A3, as FA serves as the primary source of reactive Al₂O₃ required for AFt formation.

8.

XRD test results of MSCA after reactions.

A comparison of the XRD patterns indicates that the Ca­(OH)₂ content in the hydration products of Group B is lower than that in Group A, as evidenced by the reduced diffraction peak intensities at the (011) and (001) crystal planes. In contrast, the diffraction peak intensities corresponding to the (020) and (021) planes of DG in Group B are higher than those in Group A, indicating enhanced formation of DG in Group B. In particular, at 2θ ≈ 34°, the diffraction peak is indistinct in Group A but clearly observable in Group B, corresponding to the characteristic peak of the (150) plane of DG. This further confirms the increased generation of DG in Group B. Furthermore, a comparison of the diffraction peak intensities of the (100) and (110) planes of AFt shows that the AFt content in Group B is higher than that in Group A.

Figure (b) presents the XRD patterns of the hydration products of Groups C and D. Both groups produced DG and AFt. However, a notable difference is observed in the Ca­(OH)₂-related peaks. In Group D, no obvious diffraction peaks are detected at 2θ ≈ 18° and 2θ ≈ 34°, indicating a relatively low Ca­(OH)₂ content in the reaction products. This can be attributed to the low SCA content in Group D (φ_SCA = 0.2), which limits the formation of Ca­(OH)₂ from the hydration of CaO. In addition, the generated Ca­(OH)₂ is further consumed through reactions with DG and reactive Al₂O₃ to form AFt, leading to an additional reduction in its residual content.

3.3.2. Microscopic Morphology of MSCA

The micromorphological characteristics of the hydration products were further investigated using scanning electron microscopy. The observations were conducted using a GAIA3 system (TESCAN, Czech Republic) at the Modern Testing and Analysis Technology Center of China University of Mining and Technology. Figure presents SEM images of the hydration products of subgroups B1, C1, C3, and D1, with magnifications of 1500× for the main images and 5000× for the enlarged views. For subgroup B1, the hydration products exhibit lamellar crystalline structures and irregular flocculent gel phases. The lamellar crystals are identified as Ca­(OH)₂ formed from the hydration of CaO, while the flocculent gels correspond to C–S–H, generated through the reaction between reactive SiO₂ in FA and Ca­(OH)₂ under strongly alkaline conditions. In subgroup C1, well-defined hexagonal prismatic crystals are observed, which are characteristic of AFt formed through the interaction of Ca­(OH)₂, reactive Al₂O₃, and CaSO₄·2H₂O. , In subgroup C3, a large amount of calcium aluminate hydrate (C–A–H) is formed. The SEM images show that C–A–H gel is extensively distributed on the surfaces of spherical FA particles, indicating that the reactive components of FA actively participate in the hydration reactions. For subgroup D1, the hydration products mainly consist of interwoven rod-like C–A–H structures, forming a relatively dense microstructure.

9.

SEM images of MSCAs after reactions.

3.3.3. MSCA Reaction Process

MSCA underwent complex chemical reactions upon contact with water, which mainly consisted of three core stages, as shown in Figure . The first stage involved the hydration reactions of CaO and HG. CaO hydrated to form Ca­(OH)₂, accompanied by a significant increase in molecular volume (with a volume expansion rate of 300%). , This reaction serves as the primary source of expansion pressure in MSCA, while the associated exothermic heat release accelerates subsequent reactions. Simultaneously, HG undergoes hydration, generating Ca²⁺ and SO₄ ^2– ions in solution. These ions promote the hydration of CaO, thereby influencing the evolution of expansion pressure. This mechanism explains the experimental observation in Figure that subgroup A1 exhibits no distinct inflection point in the expansion pressure curve, which can be attributed to its low HG content (φ_HG = 0.1), and consequently insufficient Ca²⁺ concentration. As the ionic concentration increases to saturation, calcium sulfate dihydrate precipitates. The crystallization of DG induces a slight volumetric expansion; however, its contribution to the overall expansion pressure is limited. In addition, the hydration of HG also releases heat, further facilitating the progression of the overall reaction system. During this stage, the generated Ca­(OH)₂ provides negligible cementitious strength, whereas the acicular network structure formed by DG contributes to the initial structural integrity of MSCA.

10.

Diagram of reaction process of MSCA.

The second stage corresponds to the pozzolanic reaction of fly ash (FA). The strongly alkaline environment generated by Ca­(OH)₂ provides the necessary conditions for this reaction. Under such alkaline activation, the amorphous glassy network structures of reactive silica and alumina in FA gradually dissolve. The released SiO₄ ^2– and AlO₂ ^– subsequently react with Ca²⁺ in the pore solution to form C–S–H gel and C–A–H.

$$x\mathrm{Ca}{(\mathrm{OH})}_2+{\mathrm{SiO}}_2+(y-x){\mathrm{H}}_2\mathrm{O}\to x\mathrm{CaO}·{\mathrm{SiO}}_2·{y\mathrm{H}}_2\mathrm{O}(\mathrm{C}-\mathrm{S}-\mathrm{H})$$ 4

$$x\mathrm{Ca}{(\mathrm{OH})}_2+{\mathrm{Al}}_2{\mathrm{O}}_3+(y-x){\mathrm{H}}_2\mathrm{O}\to x\mathrm{CaO}·{\mathrm{Al}}_2{\mathrm{O}}_3·{y\mathrm{H}}_2\mathrm{O}(\mathrm{C}\text{−}\mathrm{A}\text{−}\mathrm{H})$$ 5

C–S–H gel was the most critical strength phase and core cementitious phase of MSCA after reaction. It presented an amorphous flocculent morphology and filled the gaps between the previously formed DG crystal skeleton. Quartz and mullite in FA were inert components that barely participated in chemical reactions, serving as filler aggregates. The C–S–H gel and C–A–H bound the DG crystal skeleton to the inert components in FA, making the overall structure denser and thus improving the medium-term mechanical strength of MSCA.

The third stage was primarily dominated by the formation of AFt. Ca­(OH)₂ generated from the hydration of CaO provided a strongly alkaline environment. The C–A–H gel was further reacted with SO₄ ^2– to form AFt, as shown in the following chemical reaction.

$$3\mathrm{Ca}{(\mathrm{OH})}_2+3{\mathrm{CaSO}}_4·2{\mathrm{H}}_2\mathrm{O}+{\mathrm{Al}}_2{\mathrm{O}}_3+23{\mathrm{H}}_2\mathrm{O}\to 3\mathrm{CaO}·{\mathrm{Al}}_2{\mathrm{O}}_3·3{\mathrm{CaSO}}_4·32{\mathrm{H}}_2\mathrm{O}$$ 6

AFt existed as slender acicular crystals that are interlaced and interconnected, accompanied by a certain degree of volume expansion effect, which further enhanced the strength of the reaction products. The formation of AFt consumed a substantial amount of water. If the water content in the initially prepared MSCA slurry was insufficient, the duration of the third stage would be prolonged. During the natural curing process, the MSCA products absorbed moisture from the surrounding environment, leading to a gradual increase in strength.

3.4. Brazilian Splitting Strengths of MSCA Reaction Products

3.4.1. Brazilian Splitting Strengths for Different MSCAs

As observed in the hydration temperature experiments described in Section , Group A MSCA with a high SCA content failed to form consolidated materials with sufficient mechanical strength after hydration. In contrast, Groups B, C, and D formed intact, well-bonded solid blocks. To evaluate the mechanical properties of these cured MSCA materials, the Brazilian splitting test was employed to determine their tensile strength after hydration. After the test of expansion pressures, the MSCA specimens were placed in a standard curing chamber intended for cement-based materials for 30 days of curing, with a constant temperature of 20 ± 2 °C and a relative humidity of no less than 95%. Subsequently, they were machined into standard cylindrical disc specimens with a diameter of 50 mm and a thickness of 25 mm, as shown in Figure . The Brazilian splitting tests were conducted using an electro-hydraulic servo testing machine. A displacement-controlled loading mode was applied, with a loading rate of 0.005 mm/s.

11.

Splitting strength tests for MSCA reaction products

Figure presents the stress–displacement curves of the MSCA reaction products obtained from the Brazilian splitting tests. As shown in Figure (a), the splitting behavior of Group B specimens was generally consistent. During the loading stage, the stress increased approximately linearly with displacement. The similar slopes of the stress–displacement curves for the three subgroups indicate comparable deformation characteristics and stiffness. However, their peak splitting strengths differed. Subgroups B1, B2, and B3 reached their peak strengths at compression displacements of 0.63, 0.68, and 0.72 mm, respectively, with corresponding values of 130, 143, and 154 kPa. These results indicate that a higher HG content leads to improved splitting strength. As shown in Figure (b), the stress of subgroup C1 increased approximately linearly with displacement throughout the splitting process, whereas the C2 and C3 specimens exhibited a distinct plastic compaction stage prior to failure. The peak splitting strengths of subgroups C1, C2, and C3 were 185, 243, and 285 kPa, occurring at compression displacements of 0.86, 0.96, and 1.00 mm, respectively. Compared with Group B, the variation in splitting strength among Group C specimens was more pronounced. Nevertheless, the same trend was observed, namely that increasing HG content resulted in higher splitting strength.

12.

Stress-strain curves and splitting strength statistics of MSCA reaction products.

The stress–displacement curves of Group D specimens are shown in Figure (c). The stress evolution during the splitting process exhibited a nonlinear and stepwise increasing pattern, with consistent trends observed among all three subgroups. Specifically, all specimens showed a noticeable change in slope when the stress reached approximately 230 kPa, followed by a second transition at around 540 kPa with an increased slope, and a third transition at approximately 690 kPa, where the slope decreased again. These features indicate a multi-stage deformation process during loading. The main differences among the subgroups lie in their deformation capacity and peak strength. Subgroups D1, D2, and D3 failed at compression displacements of 1.24, 1.42, and 1.59 mm, respectively, with corresponding splitting strengths of 1018, 1110, and 1224 kPa. These results demonstrate that increasing HG content not only enhances the splitting strength but also improves the deformation capacity of the MSCA specimens.

Compared with Groups B and C, Group D exhibits significantly higher splitting strength, indicating that the optimized material composition in this group leads to superior mechanical performance. Figure (d) summarizes the splitting strengths of MSCA with different compositions. At a constant SCA content, the splitting strength of MSCA increases with increasing HG content. Notably, Group C shows a more pronounced sensitivity to HG content: the splitting strength of subgroup C2 is 31.4% higher than that of C1, while that of C3 is 17.3% higher than that of C2. At a constant HG content, the SCA content has a significant influence on the splitting strength. Specifically, for HG mass fractions of 0.3, 0.4, and 0.5, the splitting strengths increase by 42.3%, 69.9%, and 85.1%, respectively, when the SCA content decreases from 0.4 to 0.3. When the SCA content is further reduced to 0.2, the splitting strengths of MSCA with HG contents of 0.4 and 0.5 increase dramatically by 318.9% and 289.5%, respectively. These results demonstrate a substantial enhancement in strength for Group D. Combined with the XRD and SEM analyses, the superior mechanical performance of Group D can be attributed to the near-complete consumption of Ca­(OH)₂, which is extensively converted into ettringite. This transformation promotes the formation of a denser and more interconnected microstructure, thereby significantly improving the splitting strength.

3.4.2. Statistics of MSCA Expansion Pressures and Splitting Strengths

Expansion pressure and splitting strength are the two key indicators for evaluating the engineering applicability of MSCA, and their coupled regulation is essential for targeted mix design optimization. Figure illustrates the relationship between expansion pressure and splitting strength for Groups B, C, and D. As the splitting strength increases markedly from 130–154 kPa in Group B to 1018–1224 kPa in Group D, the expansion pressure correspondingly decreases from 10.2–12.9 MPa to 5.7–7.8 MPa, revealing a clear negative correlation between the two parameters. MSCA in Groups B (φ_SCA = 0.4) and Group C (φ_SCA = 0.3) exhibits relatively high expansion pressure but low strength, whereas Group D shows the opposite trend, characterized by high strength and reduced expansion pressure. This negative correlation originates from the competitive balance between the expansion effect of SCA and the cementitious strengthening effects of FA and HG. On the one hand, expansion pressure is primarily governed by the volumetric expansion of Ca­(OH)₂ generated from CaO hydration; therefore, SCA content serves as the dominant factor controlling expansion pressure. On the other hand, splitting strength depends on the cementitious capacity and microstructural compactness of the hydration products. As the SCA content decreases, the proportions of FA and HG increase, promoting the formation of strength-bearing phases. The reactive SiO₂ and Al₂O₃ in FA undergo pozzolanic reactions with Ca­(OH)₂ under strongly alkaline conditions, generating C–S–H gel and C–A–H. These products fill pore spaces and bond inert particles such as mullite, leading to enhanced structural densification. Meanwhile, HG reacts with Ca­(OH)₂ and reactive Al₂O₃ to form interlaced needle-like AFt crystals, which construct a rigid skeleton and significantly improve load-bearing capacity. Therefore, the optimization of MSCA composition is essentially a process of balancing expansion pressure and cementitious strengthening. By reducing SCA content while increasing FA and HG contents, the excessive expansion pressure can be effectively controlled, while simultaneously enhancing the mechanical strength and structural integrity of the material. This synergistic regulation mechanism provides a theoretical basis for designing MSCA with both suitable expansion characteristics and high mechanical performance for borehole sealing applications.

13.

(a) Statistics of splitting strengths and maximum expansion pressures, (b) comparison between D1MSCA and coal sample.

The relationship between expansion pressure and splitting strength of all MSCA samples shown in Figure (a) provides a theoretical basis for the engineering application of MSCA. At present, borehole sealing for gas extraction in underground coal mines is commonly implemented using the “two-plug, one-injection” technique, with a typical grouting pressure of approximately 5 MPa. Meanwhile, the splitting strength of coal seams is generally around 1000 kPa. − The overlapping region in Figure (a) indicates that when the expansion pressure of MSCA is comparable to the grouting pressure, the material can simultaneously maintain mechanical properties similar to those of the coal seam. This region therefore represents the optimal performance window for practical application. In this study, the composition of subgroup D falls within this range, demonstrating its suitability for borehole sealing applications in underground coal mines. Furthermore, the stress-displacement curves of coal sample in the splitting process were tested and compared with those of MSCA specimens. Coal blocks were sampled from the 5304 mining face of Changping Coal Mine for specimen preparation. The coal blocks were machined into standard Brazilian disc specimens. As illustrated in Figure (b), the stress-displacement curves of the coal sample and D1MSCA during the splitting process exhibit good consistency. The Brazilian splitting tensile strength of the coal sample from laboratory tests and that of D1MSCA from numerical simulation are 1.07 and 1.06 MPa, respectively. This result demonstrates that D1MSCA possesses mechanical properties close to those of the coal seam, making it suitable for use as a borehole sealing material.

4. Discussions

4.1. Numerical Model

To further investigate the influence mechanism of the mechanical compatibility between sealing materials and coal seams on interface damage under mining-induced stress, a two-dimensional numerical model was established using the discrete element method, − as illustrated in Figure (a). The overall dimensions of the model were 1000 mm × 1000 mm, with a centrally located borehole of 120 mm in diameter. The linear parallel bond (LPB) contact model was adopted to characterize the mechanical behavior between particles. The LPB contact model is capable of accurately simulating the bonding and fracture behavior of brittle materials, as shown in Figure (b). Figure (c) presents the force–displacement law of the LPB model. When the tensile stress or shear stress at a contact exceeds the prescribed tensile or shear strength, bond breakage occurs and the contact fails. To simulate the increase in mining-induced stress during the mining process, the boundary conditions of the model were defined as follows: (1) initial horizontal and vertical stresses of 15 MPa were applied to the model using the servo-control mechanism in PFC; (2) the left, right, and bottom boundaries were fixed; and (3) the stress applied to the top boundary was gradually increased from 15 to 30 MPa in increments of 1 MPa, as illustrated in Figure (d). Coal has an inherent bedding structure. Since the bedding structure is not the focus of this numerical simulation, it is not included in the model.

14.

(a) Numerical model, (b) contact behavior of LPB model, (c) force–displacement law, and (d) stress boundary conditions.

To ensure the reliability of the numerical simulation, the microscopic parameters of both the coal model and the D1MSCA model were calibrated based on the results of the Brazilian splitting tests. , The calibrated microscopic parameters are listed in Table . Figure compares the stress–displacement curves obtained from the numerical simulations and laboratory experiments for the Brazilian splitting tests. For the coal model, the experimental splitting strength and the corresponding displacement at peak stress were 1.03 MPa and 1.159 mm, respectively, while the simulated values were 1.052 MPa and 1.172 mm, corresponding to relative errors of 2.14% and 1.12%, respectively. For the MSCA model, the experimental splitting strength and the corresponding displacement were 1.019 MPa and 1.241 mm, respectively, whereas the simulated values were 1.028 MPa and 1.229 mm, with relative errors of 0.88% and 0.1%, respectively. The simulated curves exhibit excellent agreement with the experimental results for both coal and MSCA specimens, indicating that the calibrated microscopic parameters can accurately reproduce the macroscopic mechanical behavior of the two materials.

3. Parameters Setting for Coal and MSAC Models.

Parameter	Coal model	MSCA model (D1)
Emod	2.1 × 108 Pa	1.95 × 108 Pa
Pb_emod	2.1 × 108 Pa	1.95 × 108 Pa
kraion	1	1
Pb_ten	6.4 × 106 Pa	6.2 × 106 Pa
Pb_coh	13.5 × 106 Pa	13.2 × 106 Pa

15.

Comparison of split strengths between numerical models and samples.

Furthermore, to independently investigate the effects of elastic modulus and material strength on sealing interface damage, four comparative cases were designed, as summarized in Table . In cases E1, E2, and E3, the strength parameters of the coal model were kept unchanged, while the elastic modulus was increased to 2, 3, and 4 times the original value, respectively, in order to examine the influence of deformation-property mismatch on sealing performance. In contrast, case T maintained the same elastic modulus as the coal model, while the tensile strength and bond strength were both increased by 1.5 times to evaluate the influence of strength-property differences. The stress–displacement curves of the Brazilian splitting tests under different conditions are presented in Figure . With increasing elastic modulus, the peak strength remained nearly unchanged, whereas the peak displacement decreased significantly and the slope of the stress–displacement curve increased markedly, indicating enhanced brittleness. By comparison, increasing the strength parameters significantly improved both the peak strength and the peak deformation capacity, while the overall deformation characteristics remained essentially unchanged.

4. Parameter Setting for Different Sealing Materials.

	E1 Group	E2 Group	E3 Group	T Group
Emod	4.2 × 108 Pa	6.3 × 108 Pa	8.4 × 108 Pa	2.1 × 108 Pa
Pb_emod	4.2 × 108 Pa	6.3 × 108 Pa	8.4 × 108 Pa	2.1 × 108 Pa
Pb_ten	6.4 × 106 Pa	6.4 × 106 Pa	6.4 × 106 Pa	9.6 × 106 Pa
Pb_coh	13.5 × 106 Pa	13.5 × 106 Pa	13.5 × 106 Pa	20.25 × 106 Pa

16.

Stress-displacement curves for different model parameters.

4.2. Stress and Crack Distributions around Borehole

The initiation and propagation of cracks around boreholes under mining-induced stress are the fundamental causes of sealing failure, while the mechanical compatibility between the sealing material and the coal seam directly governs the extent and distribution of crack development. Figure summarizes the numbers of tensile and shear cracks around the borehole under different conditions when the vertical stress in the y-direction reached 25 and 30 MPa, respectively. When the y-direction stress was 25 MPa, only 5 and 3 cracks were generated around the boreholes sealed with D1MSCA and the T-group material, respectively, both of which possessed elastic moduli close to that of the coal seam. In contrast, the numbers of cracks in groups E1–E3, where the elastic modulus was progressively increased, were 4, 21, and 24, respectively. These results indicate that elastic modulus is a critical factor affecting sealing performance, and that greater differences in elastic modulus between the sealing material and the coal seam lead to more severe damage. When the y-direction stress further increased to 30 MPa, the numbers of cracks in the D1 and T groups increased to 55 and 53, respectively, whereas those in groups E1–E3 increased markedly to 78, 126, and 135, respectively. Analysis of the crack distributions reveals that, in the D1 and T groups, cracks were mainly generated within weak-contact regions of the coal matrix. By contrast, in groups E1–E3, crack development was concentrated primarily around the borehole. These results clearly demonstrate that the mismatch in mechanical properties between the sealing material and the coal seam significantly deteriorates sealing performance.

17.

Crack distribution under different conditions.

Stress concentration is the primary driving factor for damage development; therefore, the stress distribution characteristics around the borehole were further analyzed. A vertical measuring line and a horizontal measuring line were arranged through the center of the model to investigate the distributions of the y-direction stress and x-direction stress, respectively, under a vertical stress of 30 MPa. The results are presented in Figure . As shown in Figure (a), the stress distributions around the boreholes sealed with D1MSCA and the T-group material were relatively uniform under a y-direction stress of 30 MPa, indicating that no obvious stress concentration occurred around the borehole. In contrast, distinct stress concentration zones developed around the central borehole in groups E1–E3. The maximum stress values indicate that the degree of stress concentration followed the order E3 > E2 > E1. The maximum stresses around the borehole in these groups all exceeded 42 MPa, which is approximately 10 MPa higher than those in the D1 and T groups, corresponding to an increase of approximately 31.3%. The y-direction stress distribution for group E3 is also presented in Figure , clearly illustrating the pronounced stress concentration around the borehole. As shown in Figure (b), the horizontal stress distributions similarly demonstrate evident stress concentration around the boreholes in groups E1–E3, with the degree of concentration also following the order E3 > E2 > E1. The differences in stress distribution further highlight the importance of mechanical compatibility. Since the elastic modulus of D1MSCA is close to that of the coal seam, the sealing material can deform coordinately with the surrounding coal mass, thereby preventing stress concentration around the borehole and reducing the risk of sealing failure. However, when a large discrepancy exists between the elastic moduli of the sealing material and the coal seam, pronounced stress concentration zones are formed around the borehole, which subsequently induce extensive crack generation and ultimately lead to sealing failure.

18.

Stress distributions at measuring lines.

4.3. Engineering Test

To verify the engineering applicability of the laboratory experiments and numerical simulation results, a field industrial test was conducted at the 5304 mining face of Changping Coal Mine. The test area was located 170 m ahead of the mining face, where the coal seam had an average thickness of 5.6 m and a dip angle ranging from 3° to 5°, and five test boreholes were arranged. Among them, boreholes 1# and 2# were sealed using conventional cement-based materials as the control group, whereas boreholes 3#, 4#, and 5# were sealed using the optimal D1MSCA formulation as the experimental group. All boreholes had a depth of 80 m and a spacing of 4 m, as shown in Figure (a). All boreholes were sealed using the “two-plug, one-injection” sealing technique. The specific construction procedure was as follows: (1) a φ153 mm reaming bit was used to enlarge the first 1 m of the borehole opening, after which a φ113 mm drill bit was employed to drill to the designed depth of 80 m; (2) after drilling, flushing was continuously performed for 30 min until the coal debris inside the borehole was essentially removed; (3) a φ32 mm slotted screen pipe was installed to provide full-length borehole protection; (4) the sealing pipe assembly was installed; and (5) boreholes 1# and 2# were sealed using conventional cement-grouting materials under a grouting pressure of 5 MPa, whereas boreholes 3#–5# were sealed by pressure-free injection of MSCA slurry. A schematic diagram of the sealing structure is presented in Figure (b).

19.

(a) Engineering trail area, (b) schematic diagram of borehole sealing, (c) test process for gas concentration in borehole, (d) Gas concentration measurement equipment, and (e) On-site testing in underground coal mine.

After sealing was completed, a gas concentration monitoring system was employed to continuously monitor the gas concentration of the boreholes. Figure (c) illustrates the schematic diagram of the field gas concentration measurement system, and the measurement equipment is shown in Figure (d). During the measurement process, the detection device was first connected, and a suction pump was used to extract gas from the borehole gas drainage pipe. The extracted gas then passed through a desiccant filter before entering the portable gas concentration detector. A CD4 portable gas concentration detector was used in the field test. Figure (e) shows the on-site gas concentration measurement process. All five boreholes were connected to the main gas drainage pipeline, and the negative pressure of gas drainage was maintained stably at 22 kPa.

The variation in borehole gas concentration with the advance of the mining face is shown in Figure , where the horizontal axis represents the distance between the mining face and the boreholes. As shown in Figure (a), when the distance between the mining face and the boreholes exceeded 45 m, the gas concentrations of all boreholes gradually decreased. For example, the gas concentration of borehole 5# decreased from 14% to 10%. This phenomenon can be attributed to the relatively low mining-induced abutment stress, which reduced the permeability of the coal seam. During this process, the mining-induced stress in the test area gradually increased. When the mining face advanced to within 45 m of the test area, the borehole gas concentration began to increase. This is because the mining-induced abutment stress caused damage to the coal mass, thereby enhancing gas desorption and resulting in increased borehole gas concentration. Comparison of the gas concentration variations among the five boreholes indicates that the gas extraction concentrations of boreholes 3#–5#, sealed using MSCA, were consistently higher than those of boreholes 1# and 2#, which were sealed using conventional cement slurry. These results demonstrate that the sealing performance of MSCA is superior to that of traditional cement-based sealing materials.

20.

(a) Statistics of gas concentrations, (b) average of gas concentrations for different borehole.

To further evaluate the differences in gas extraction performance at different mining stages, the average gas concentrations under different sealing methods were analyzed, as shown in Figure (b). Throughout the entire mining process, the gas concentrations of the MSCA-sealed boreholes remained consistently higher than those of the cement-sealed boreholes. When the mining face advanced to within 40 m of the test boreholes, the coal seam experienced increased damage under mining-induced abutment pressure, leading to a gradual increase in borehole gas concentration. A comparison between the two sealing methods shows that the gas concentration of boreholes sealed with conventional cement slurry increased from 7.6% to 9.6% and then remained relatively stable. In contrast, the gas concentration of the MSCA-sealed boreholes continuously increased from 10.2% to 14.4%, indicating that MSCA could still maintain excellent sealing performance under mining-induced stress conditions. The dashed line in Figure (b) represents the relative increase rate of the average gas concentration of MSCA-sealed boreholes compared with that of conventional cement-sealed boreholes. When the distance between the mining face and the boreholes exceeded 50 m, the gas concentration of MSCA-sealed boreholes increased by 52%–112% relative to that of conventional cement-sealed boreholes, indicating that under relatively weak mining influence, the gas extraction concentration of MSCA-sealed boreholes was significantly higher than that achieved using conventional cement slurry. In regions strongly affected by mining-induced stress, the gas concentration of MSCA-sealed boreholes remained 31%–50% higher than that of cement-sealed boreholes, demonstrating that MSCA-sealed boreholes could still maintain high gas extraction efficiency under severe mining disturbance conditions. The field test results demonstrate that, compared with conventional cement-based sealing materials, MSCAwith mechanical properties closer to those of the coal seamcan enhance the coordinated deformation capacity between the sealing material and the surrounding coal mass, thereby reducing crack generation around the borehole under stress conditions and improving borehole gas extraction concentration.

5. Conclusions

This study developed a modified static chemical agent by incorporating fly ash and hemihydrate gypsum into conventional SCA to overcome the inherent limitations of cement-based sealing materials in coal seam gas extraction. Through combined macroscopic experiments and microscopic characterization, the effects of component proportions on hydration behavior, expansion pressure, and mechanical performance were systematically investigated, and the underlying mechanisms were elucidated. The main conclusions are as follows:

• (1)The hydration exotherm of CaO in SCA is the dominant factor governing the temperature evolution of MSCA. Increasing HG content significantly shortens both the hydration initiation time and the duration of the rapid reaction stage. At a constant SCA content, the peak hydration temperature increases linearly with HG content, whereas its influence diminishes as SCA content decreases. When the SCA mass fraction exceeds 0.5, excessive heat release leads to blowout and prevents the formation of intact solid structures. Reducing the SCA content from 0.5 to 0.2 lowers the peak temperature from ∼110 °C to 36–54 °C, effectively suppressing blowout and improving hydration stability.
• (2)Both SCA and HG contents are positively correlated with the maximum expansion pressure, with SCA playing the dominant role. As the SCA mass fraction decreases from 0.5 to 0.2, the maximum expansion pressure after 30 h reduces from 11.5–12.9 to 5.7–7.8 MPa. At a fixed SCA content, increasing HG content enhances both the initial growth rate and peak value of expansion pressure, primarily due to the promotion of SCA hydration by Ca²⁺ released during HG hydration.
• (3)The pozzolanic reaction of FA under strongly alkaline conditions generates C–S–H and C–A–H gels, which fill pores and densify the microstructure. Meanwhile, HG reacts with Ca­(OH)₂ and reactive Al₂O₃ to form interlaced acicular AFt, establishing a rigid load-bearing framework. As a result, expansion pressure exhibits a negative correlation with splitting strength. Among all investigated mix proportions, when the mass fractions of SCA and HG are 0.2 and 0.4, respectively, MSCA achieves a splitting strength of 1018 kPa and an expansion pressure of 5.7 MPa, which are well matched with the mechanical properties of coal seams and typical grouting pressures.
• (4)Numerical simulation results demonstrated that the mechanical compatibility between the sealing material and the coal seam plays a critical role in controlling interface damage under mining-induced stress. When the elastic modulus of the sealing material was close to that of the coal seam, stress concentration around the borehole was effectively alleviated, and crack development was significantly reduced. In contrast, a large mismatch in elastic modulus caused severe stress concentration around the borehole, resulting in extensive crack propagation and sealing failure. Compared with strength parameters, elastic modulus matching had a more significant influence on sealing performance.
• (5)Field industrial tests conducted at the 5304 mining face of Changping Coal Mine verified the engineering applicability of MSCA. Compared with conventional cement-based sealing materials, MSCA-sealed boreholes exhibited significantly higher gas extraction concentrations throughout the mining process. Under strong mining-induced stress conditions, the gas concentration of MSCA-sealed boreholes remained 31%–50% higher than that of conventional cement-sealed boreholes. These results demonstrate that MSCA can effectively improve the coordinated deformation capacity between the sealing material and the coal seam, reduce crack generation around the borehole, and enhance gas extraction performance under mining disturbance conditions.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

W.T.: Writing original draft, visualization, data curation, conceptualization, methodology, and funding acquisition. C.Z.: Supervision, funding acquisition. J.X.: Formal analysis. Y.Z.: Validation. J.Z.: Investigation. X.F.: Validation. W.Y.: Writing original draft, validation.

The authors declare no competing financial interest.