Characteristics of Grouting-Induced Microfractures in Fractured Rock Masses: Numerical Simulation, Microseismic Monitoring, and Laboratory Tests

Abstract

In deep mining engineering, grouting operations, although designed for reinforcement, inevitably induce microfracturing and associated microseismicity. Investigating the characteristics of grouting-induced microfractures in fractured rock masses is crucial for evaluating the grouting process and its effectiveness. Using the Wutongzhuang Mine as a case study, this paper first establishes mechanical criteria covering three stages—fracture filling, coupled permeation, and fracturing propagation—to analyze the process characteristics of grouting-induced microfractures. It reveals the mechanisms by which grouting pressure, in situ stress, and rock mass strength control fracture initiation and propagation. Furthermore, a grouting simulation method based on the Particle Flow Code (PFC) is proposed and summarized, constructing a “pipe-domain” fluid network considering fluid–solid coupling, thereby achieving a refined numerical reproduction of the entire grouting process. Addressing the complex geological conditions of the mine, three typical grouting modes are simulated and analyzed: grouting under conventional geological conditions, grouting under densely fractured conditions, and grouting near fault structures. The simulation results unveil their core influencing factors and behavioral characteristics: under conventional conditions, microfractures exhibit a “three-stage” evolution with the grouting process; under densely fractured conditions, the density of pre-existing fractures dominates the formation of complex fracture networks; and finally, fault structures guide fracture propagation, causing microfractures to cluster nearby. Subsequently, the development trends of microfractures under different grouting effects are clarified: after effective reinforcement, the rock mass strength increases, and the scope and quantity of fractures induced by subsequent grouting significantly decrease. The behavioral patterns under these different grouting modes are effectively validated through field microseismic monitoring, confirming the intrinsic relationship between the spatio-temporal evolution of grouting-induced microfractures and geological structures/grouting techniques. Finally, laboratory tests are conducted using a self-developed experimental apparatus, selecting grouting pressure, pore water pressure in the rock mass, and matrix grain size as variables. The mapping relationships between these variables and microseismic waveform characteristics, amplitude, etc., under different schemes are obtained, providing a basis for inverting the microfracturing process and evaluating grouting effectiveness. The research results provide multi-faceted references for characterizing the stability of fractured rock masses via microseismic monitoring and for optimizing grouting effectiveness.

1. Introduction

As coal mining progressively transitions into deep mining stages, rock masses are subjected to complex environments characterized by high in situ stress and high seepage pressure, making floor water inrush increasingly a major disaster threatening mine safety production [1,2]. Grouting reinforcement is a key technical measure for modifying the floor aquifuge and sealing water-conducting channels [3,4]. However, the grouting process itself, as an active disturbance to the rock mass, induces numerous microfractures accompanied by the release of microseismic signals while the high-pressure grout permeates and fractures the rock mass [5,6,7]. These grouting-induced microseismic activities are direct manifestations of rock mass fracturing and potentially influence the stability of the surrounding rock. Therefore, a profound understanding of the mechanical mechanisms, spatio-temporal evolution patterns, and response relationships with geological structures of grouting-induced microfractures is of great significance for scientifically evaluating grouting effectiveness, dynamically controlling the grouting process, and ensuring mine safety [8].

Currently, researchers have conducted extensive studies on grouting-induced microfractures. For instance, Liu et al. [9] performed grouting tests on materials similar to coal rock, revealing that larger pre-existing fracture sizes make the surrounding rock more prone to rupture, facilitate the connection between grouting-induced splitting fractures and pre-existing fractures, and lead to more complex secondary splitting modes. Cao et al. [10] analyzed the diffusion patterns of grout under different working conditions, concluding that moderately increasing the grouting rate can accelerate the process, but reducing the spacing between grouting holes is more efficient than increasing the grouting rate. Wang et al. [11] studied the influence of grouting on mesoscopic failure mechanisms, revealing that when the defect inclination angle is ≤30°, grouting significantly increases the proportion of shear microfractures during the large crack propagation stage, shifting the failure mode towards shear. When the defect inclination angle is 45° and 60°, grouting increases the shear microfracture ratio throughout the failure process; when it is 75° and 90°, grouting has little effect on subsequent crack propagation. Fan et al. [12] systematically studied the diffusion process of fracture grouting in homogeneous and heterogeneous strata, finding that weak zones experience tensile fractures and horizontal deflection before grout solidification, and local stress rotation in the soil can induce inclination and deflection. Liu et al. [13] established a Critical Load Angle Model (C-LAM) to predict fracture development and clarify the mechanism by which grouting alters fracture mode transitions in grouted fractured rock masses. Yan et al. [14] proposed a three-dimensional grouting model based on the Combined Finite-Discrete Element Method (FDEM), clarifying grout transport, pressure distribution, grout-rock interaction, rock deformation, and fracture initiation and propagation laws. They also proposed a 2D grouting model considering hydro-mechanical coupling and fracturing to analyze the interaction between grouting-induced fractures and pre-existing fractures [15]. Yu et al. [16] used FLAC^3D to conclude that the grout diffusion range affects the relationship between mining-induced deformation and the retained coal pillar behind the working face, proposing that timely grouting can strengthen the support system. Liu et al. [17] developed a grouting process simulator for hydro-mechanically coupled grouting problems, linking FDEM with a grout flow simulator to address fracture initiation, propagation, and grout solidification. Wu et al. [18] constructed a 3D discrete element model to quantify the directional effect of fracture loading on the mechanical properties of grouted rock masses and analyzed the failure mechanism of grouted rock masses under different loading directions and fracture angles. Ding et al. [19] proposed an Acoustic Emission (AE) energy-based Crack Classification Energy Accumulation Curve (CCEAC) method to analyze the initiation and propagation of tensile and shear cracks in grouted sandstone under different inclination angles. Additionally, Liu et al. [20] studied grouted fractured sandstone reinforcement bodies, analyzing the progressive damage law leading to macroscopic defects within the rock mass during freeze–thaw cycles. Li et al. [21] improved grouting technology for water inrush from working faces after grouting reinforcement, proposing a new sleeve model to prevent water inrush by controlling floor strata deformation.

Although the aforementioned studies have made significant progress in revealing grouting fracture mechanisms [22] and optimizing processes [23], there remain shortcomings in research angle or scale. For example, they fail to effectively correlate and mutually verify microscopic fracture mechanisms, macroscopic field responses, and signal physical characteristics, resulting in a disconnect in the chain from “fracture mechanism” to “interpretable field signals.” Moreover, there is a lack of systematic experimental research on how microseismic signal characteristics, such as waveform and amplitude, are controlled by specific grouting parameters and rock mass structures. This lack makes it difficult to use microseismic signals as a physical basis for inverting the grouting process and evaluating reinforcement effectiveness. In summary, this paper first analyzes the process characteristics of grouting-induced microfractures and establishes mechanical criteria covering the three stages of fracture filling, coupled permeation, and fracturing propagation. Then, a PFC hydro-mechanical coupling model considering a “pipe-domain” fluid network is generated to analyze the initiation, propagation, and aggregation behavior of microfractures during the entire grouting process under different conditions, validated against field microseismic monitoring at the 182602 working face of Wutongzhuang Mine. Finally, through self-developed laboratory grouting test equipment, key variables such as grouting pressure, pore water pressure, and matrix grain size are actively controlled to systematically investigate the quantitative mapping relationships between these parameters and microseismic waveform, amplitude, and their coefficient of variation. The research results reveal the characteristics and patterns of grouting-induced microfractures and provide multi-faceted references for characterizing the stability of fractured rock masses via microseismic monitoring and for optimizing grouting effectiveness.

2. Engineering Background

2.1. Mine Geological Conditions

Wutongzhuang Mine is located in China’s Hanxing Mining Area. Its 182602 working face mines the #2 coal seam, with average burial depth, dip angle, and thickness of 950 m, 14°, and 3.2 m, respectively. The inclined length of the working face is 256 m, and the strike length is 640 m. During mining, the working face is affected by the underlying Yeqing limestone, Shanfuqing limestone, and Ordovician limestone aquifers, with average distances from the coal seam floor of 38, 72, and 148 m, and maximum water head pressures of 1.8, 6.25, and 10.7 MPa, respectively. Geophysical exploration and drilling revealed 6 faults within the working face area, along with areas of abnormally dense fractures, all requiring verification and reinforcement via grouting drilling. Therefore, according to the water inrush coefficient method and water prevention and control requirements [24,25], comprehensive grouting modification of the coal seam floor aquifuge and the top of the Ordovician limestone is necessary, divided into surface grouting and underground grouting.

As shown in Figure 1, surface horizontal multi-branch directional grouting, taking the 182602 working face mining range as a unit, involved 6 main holes, 21 Ordovician limestone branch holes, and 4 Shanfuqing limestone branch holes, with a cumulative grouting volume of 7066.5 tons. The spacing between adjacent branch holes was 50 and 60 m, and the final grouting pressure at the hole bottom was 24.24 MPa. Underground grouting within the 182602 working face mining range involved 106 boreholes with a cumulative grouting volume of 998.77 tons.

2.2. Microseismic Monitoring Layout

The KJ1073 microseismic monitoring system (Hebei Coal Science Research Institute, Xingtai City, China) was employed. Five uniaxial geophones and one triaxial geophone were arranged in the haulage roadway, and six uniaxial geophones and one triaxial geophone were arranged in the return airway, with a spacing of 120 m between geophones. As shown in Figure 2, a ring-shaped array structure was formed around the 182602 working face, focusing on monitoring the working face area while also covering microseismic signals in the surrounding mining areas.

3. Mechanical Mechanism of Grouting-Induced Microfracture Process

First, we analyze the mechanical mechanism of grouting induced microfracture process. The grouting process is essentially a fluid–solid coupling process where grout, as a fluid medium, interacts with the fractured rock mass under pressure-driven conditions. The induced microfracturing behavior can be divided into stages with different dominant mechanical mechanisms based on the changes in the rock mass stress state caused by the grout.

(1)

Fracture Filling Process

In this process, the grouting pump pressure is very low. The grout primarily relies on gravity and fluidity to fill existing open fractures and larger pores in the rock mass under hydrostatic pressure. Therefore, the rock mass can be treated as an elastic semi-infinite body, with the grouting hole and connected open fractures simplified as boundaries. The grouting pump pressure $P_h$ ≈ 0 in this stage, and the pressure acting on the fracture boundaries is mainly the grout’s gravitational pressure $P_0$.

According to elasticity theory, modeling the stress concentration around a perforation in an infinite plate to simulate the grouting hole. The circumferential stress ${\sigma }_{\theta }$ around the hole is:

$${\sigma }_{\theta }={\displaystyle \frac{{\sigma }_1+{\sigma }_3}{2}}\left(1+{\displaystyle \frac{r^2}{{\delta }^2}}\right)-{\displaystyle \frac{{\sigma }_1-{\sigma }_3}{2}}\left(1+{\displaystyle \frac{3r^4}{{\delta }^4}}\right)cos2\theta -P_0{\displaystyle \frac{r^2}{{\delta }^2}}$$

(1)

where $r$ is the hole radius, $\delta$ is the radial distance from the point of interest to the hole center, ${\sigma }_1$ and ${\sigma }_3$ are the far-field maximum and minimum principal stresses, respectively. At the hole wall where $\delta =r$, the equation simplifies to:

$${\sigma }_{\theta }=({\sigma }_1+{\sigma }_3)-2({\sigma }_1-{\sigma }_3)cos2\theta -P_0$$

(2)

The circumferential stress reaches its minimum value when $\theta =0°$ or 180°, perpendicular to the direction of ${\sigma }_3$:

$${\sigma }_{\theta min}=3{\sigma }_3-{\sigma }_1-P_0$$

(3)

Therefore, the criterion for fracture initiation during the grout filling process is that the circumferential stress overcomes the rock mass tensile strength ${\sigma }_t$:

$${\sigma }_{\theta min}\le {-\sigma }_t$$

(4)

Combining Equations (3) and (4), the fracture criterion for the filling process is:

$$3{\sigma }_3-{\sigma }_1-P_0\le {-\sigma }_t$$

(5)

Since $P_0$ is very small in this stage and ($3{\sigma }_3-{\sigma }_1$) is often relatively large, this process typically only fills inherent voids and hardly generates new fractures.

(2)

Coupled Permeation Process

As the grouting pump pressure increases, grout permeates into micro-pores, triggering fluid–solid coupling effects. This stage considers the radial seepage of grout in the porous rock mass, analyzed using the effective stress principle for porous media [26]. The grout pressure $P_f$ is considered the sum of $P_h$ and $P_0$, and the pore water pressure $P_p$ within the rock mass changes significantly.

According to Biot’s effective stress theory, the effective stress experienced by the rock matrix is:

$${\sigma }^{\omega }=\sigma -\alpha P_p$$

(6)

where $\alpha$ is Biot’s effective stress coefficient.

The increment in pore pressure due to grout seepage directly reduces the effective stress on the rock matrix. Taking the critical location for fracture initiation, the hole wall circumferential stress, as an example, the effective circumferential stress ${\sigma }_{\theta }^{\omega }$ becomes:

$${\sigma }_{\theta }^{\omega }={\sigma }_{\theta }-\alpha P_p$$

(7)

Substituting the hole wall circumferential stress formula and considering the additional tangential stress increment $∆{\sigma }_{\theta }$ caused by seepage, the minimum effective circumferential stress at $\theta =0°$ or 180°, considering seepage effects, is:

$${\sigma }_{\theta min}^{\omega }=\left(3{\sigma }_3-{\sigma }_1-P_f\right)-\alpha \left(P_p-P_{p0}\right)-∆{\sigma }_{\theta }$$

(8)

where $P_{p0}$ is the initial pore water pressure in the rock mass; the additional tangential stress increment $∆{\sigma }_{\theta }$ can be expressed in terms of Poisson’s ratio $\mu$ and pore water pressure $P_p$ as:

$$∆{\sigma }_{\theta }={\displaystyle \frac{\alpha \left(1-2\mu \right)}{1-\mu }}\left(P_p-P_{p0}\right){\displaystyle \frac{r^2}{{\delta }^2}}$$

(9)

After simplification, when the effective circumferential tensile stress generated by the grouting pump pressure reaches the rock mass tensile strength, satisfying ${\sigma }_{\theta min}^{\omega }\le {-\sigma }_t$, the fracture initiation criterion for the permeation process is obtained:

$$P_h\ge 3{\sigma }_3-{\sigma }_1+{\sigma }_t-\beta (P_p-P_{p0})$$

(10)

where $\beta$ is a comprehensive seepage influence coefficient greater than 0.

From Equation (10), since $\beta \left(P_p-P_{p0}\right)>0$, the seepage effect reduces the threshold grouting pressure required for fracture initiation. In this stage, microfractures initiate at pre-existing defects due to the reduction in effective stress.

(3)

Fracturing Propagation Process

After fractures initiate due to grouting pressure, their further propagation can be analyzed using fracture mechanics theory. Firstly, the initiated fracture is simplified as a Griffith crack of length $2a$, subjected to internal pressure $P_f$ and constrained by the far-field stress ${\sigma }_3$. The Mode I stress intensity factor $K_{\mathrm{I}}$ driving fracture propagation is expressed as:

$$K_{\mathrm{I}}=(P_f-{\sigma }_3)\sqrt{\pi a}$$

(11)

This equation indicates that the net pressure ($P_f-{\sigma }_3$) acting on the fracture surface and the fracture half-length $a$ jointly determine the stress field intensity at the fracture tip.

The critical condition for unstable fracture propagation is when the stress intensity factor reaches the material’s fracture toughness $K_{\mathrm{I}\mathrm{C}}$:

$$K_{\mathrm{I}}\ge K_{\mathrm{I}\mathrm{C}}$$

(12)

Substituting the expression for $K_{\mathrm{I}}$, the core criterion for fracturing grouting is obtained:

$$(P_f-{\sigma }_3)\sqrt{\pi a}\ge K_{\mathrm{I}\mathrm{C}}$$

(13)

Equation (13) shows that maintaining a sufficiently high grouting net pressure ($P_f-{\sigma }_3$) is the direct driving force for fracture propagation, and the fracture will preferentially propagate in the direction perpendicular to the minimum principal stress ${\sigma }_3$.

4. Characteristics of Microfracture Induced by Different Grouting Conditions

4.1. Simulation Method

On the basis of the microfracture mechanics mechanism mentioned above, we further analyze its micro mechanism using numerical modeling. The Particle Flow Code (PFC, Version 5.0.21) was used to simulate the microseismic response characteristics induced by different grouting modes, assuming grout flow follows Darcy’s law [27]. For porous rock masses, the inter-particle voids can be considered as a structure of “pipes” and “domains”. This fluid network composed of fluid “pipes” and fluid “domains” can simulate complex fluid–solid coupling mechanisms such as grout seepage and fracturing. The parallel bond model was used for the rock constitutive model in this numerical simulation, and the smooth joint model was used for discontinuities such as faults and folds. The modeling process involved first generating a rock particle assembly. For cases considering discontinuities, smooth joints with specified dip angles and orientations were inserted at specified locations. Then, a fluid network was generated within these particles. The convex polygon formed by the centers of several particles constitutes a fluid domain, as shown in Figure 3, where the fluid domain boundary is the side length of this convex polygon.

Fluid exchange between fluid domains is achieved through flow through pipes at particle contacts. Contacts compress or extend under the effective stress of particles, decreasing or increasing the permeability of the flow pipes, thereby affecting fluid flow. Fluid flow causes changes in pore pressure, leading to changes in the effective stress acting on particles, thus achieving fluid–solid coupling. It is important to note that fluid domains are virtual entities, and fluid flow under pressure gradients is also virtual, but the pressure changes are real. When fluid exchanges between fluid domains at contacts, it is treated as flow between parallel plates, and the flow rate can be solved using the cubic law [28]:

$$Q=k{b}^3{\displaystyle \frac{\mathsf{\Delta }P}{L}}$$

(14)

where $b$ is the aperture, $k$ is the permeability coefficient, $∆P$ and L are the pressure difference and contact length, respectively, constituting the pressure gradient.

The pipe or contact length L is the average of the radii of the two particles, and since particles are rigid, this is fixed. From the above equation, the aperture, i.e., the contact spacing, significantly influences the fluid–solid coupling process [29]. Determining the aperture is crucial. Its change is primarily controlled by the inter-particle force. When the contact is compressed, the aperture b can be calculated by:

$$b={\displaystyle \frac{b_0{F}_0}{F+F_0}}$$

(15)

where $b_0$ is the initial aperture when no force acts on the contact, and $F_0$ is the pressure when the aperture is reduced to half. For contacts under tension or after bond breakage, the aperture can be calculated by:

$$b=b_0+\lambda \left(d-r_1-r_2\right)$$

(16)

where $\lambda$ is a dimensionless coefficient, and $r_1$ and $r_2$ are the radii of the particles. The update of pore pressure $\mathsf{\Delta }{P}_k$ can be determined by the change in fluid domain volume and the fluid bulk modulus:

$$\mathsf{\Delta }{P}_k={\displaystyle \frac{K_f}{V_d}}\left(\mathsf{\Delta }t\sum Q-\mathsf{\Delta }{V}_d\right)$$

(17)

where $K_f$ is the bulk modulus, $V_d$ is the fluid domain volume, $∆V_d$ is the change in fluid domain volume caused by deformation due to effective stress, $∆t\sum Q$ is the change in fluid domain volume caused by fluid exchange between fluid domains, and $∆t$ is the time step for each calculation. As shown in Figure 4, the length of the green line segment is taken as the contact area, and the normal direction is the resultant force direction. To apply pore pressure to particles, the contact area (length) needs to be determined. Since fluid pressure is isotropic, the force acting on a particle is:

$$f_i=P_k{l}_i{n}_i$$

(18)

Thus, the basic framework for fluid–solid coupling numerical simulation using PFC is achieved. However, a problem remains: when a bonded contact breaks, it is immediately deleted. How fluid flows and pressure is determined at that moment becomes an issue. To solve this, the following assumption is adopted: at the instant the contact is deleted, the fluid domains on both sides instantly achieve pressure balance, and fluid flow is completed instantaneously. Therefore, the fluid pressure at that instant becomes the average of the pressures in the two fluid domains before contact breakage.

$$P_k^{\prime }={\displaystyle \frac{P_{k1}+P_{k2}}{2}}$$

(19)

For materials like rock and soil with non-uniformly distributed pores and large variations, Zhou [30] proposed a new calculation formula:

$$P_k^{\prime }=\left[{\displaystyle \frac{V_{k1}+V_{k2}}{\left(V_{r1}+V_{r2}\right)\phi }}-1\right]K_f$$

(20)

Unlike traditional approaches that treat rock mass as a continuous porous medium, the “pipe-domain” model in this study directly embeds the fluid network into a discrete particle assembly. This enables the natural simulation of the entire process where fractures spontaneously initiate, propagate, and dynamically alter seepage paths along with grout flow. Compared to earlier simplified PFC fluid models, this method physically captures pore pressure gradients and local seepage forces, accurately realizing the mechanical mechanism from seepage-induced weakening to fracturing propagation as analyzed theoretically. It provides a more fundamental numerical representation for revealing the evolution of grouting-induced microfractures.

4.2. Model Construction and Research Scheme

The microseismic monitoring in this study covers the entire working face scale, which is quite large. Modeling the entire floor with current computer capabilities is impractical. Therefore, to study the influencing factors of primary microseismicity during grouting, a unit body of size 160 × 160 cm surrounding a grouting hole was selected. Using the distribute command in PFC, a particle geometry was generated with particle sizes following a normal distribution, with minimum and maximum particle sizes of 0.005 m and 0.0083 m, respectively. The meso-contact constitutive model used was the parallel bond model. The final model is shown in Figure 5.

It is worth noting that the 160 cm × 160 cm numerical model represents a Representative Elementary Volume around a single grouting borehole, a scale justified by the field-derived influence radius (0.8–1.2 m) of grouting operations at the Wutongzhuang Mine. To mitigate boundary effects and simulate an infinite rock mass domain, servo-controlled stress boundaries were implemented, maintaining constant far-field stresses and ensuring realistic stress fields and fracture propagation around the borehole. This approach, coupled with direct inputs of field-measured in situ stresses and grouting pressure, ensures the model’s validity for investigating the fundamental mechanisms of grouting-induced microfracturing.

In the above modeling process, fluid domains were established first, followed by the addition of pore water pressure. The application method is based on the servo-control principle: applying a velocity to the walls, and after a certain time, the reaction force on the walls can be calculated. The difference between this reaction force and the target confining pressure is iteratively reduced by adjusting the wall speed via a servo coefficient until the difference falls within a set tolerance. After reaching the target in situ stress, the particle force chain diagram is shown in Figure 6. The force chain values at the boundary are significantly larger than in other areas because there are no fluid domains between the walls and particles, so the stress at the boundary is the effective stress. In contrast, in the internal areas, the presence of pore pressure reduces the effective stress on the particles, so the boundary values are higher compared to the internal values. During subsequent grouting, servo control is maintained, meaning the stress boundary conditions remain constant. The grouting process first changes the initial water pressure to free flow, then changes the grouting hole boundary condition.

It is worth noting that, the Biot effective stress coefficient (α), which in continuum theory describes the influence of pore pressure on effective stress, is realized in the PFC model through the direct mechanical application of fluid domain pressure on particles. The force exerted by pore pressure on particles directly alters the contact forces between them, and this network of contact forces represents the microscopic manifestation of “effective stress” in the model. Consequently, the physical essence of the Biot coefficient is naturally represented through the direct mechanical balance between particles and fluid, with its macroscopic effect determined by the stiffness and density of the particle assembly. Furthermore, the comprehensive seepage influence coefficient (β), which theoretically reflects the additional stress effects induced by seepage, is dynamically simulated in PFC through the “pipe-domain” system. Fluid flows through the pipes (contact gaps) according to the aperture and pressure gradient; the pipe aperture, in turn, dynamically changes with the effective stress between particles. This closed-loop feedback automatically and completely reproduces the influence of seepage on the stress field, effectively realizing the physical process summarized by the coefficient β.

The grouting process is accompanied by crack development. Each micro-rupture event releases an elastic wave, forming a microseismic event. In the simulation results, each microfracture represents one microseismic event. Based on the field final hole grouting pressure settings and parameter calibration results, this simulation set a fixed grouting pressure of 24.24 MPa. The relevant parameters are shown in

Table 1. Comprehensive mesoscopic parameters for grouting simulation in PFC.

Parameter Category	Parameter Name	Value	Unit
Material Properties	Particle Density	2300	kg/m3
	Fluid Density	1200	kg/m3
	Particle Friction Coefficient	0.5	--
	Joint Friction Coefficient	1.0	--
Stiffness Parameters	Particle Contact Modulus	2.0	GPa
	Parallel Bond Modulus	2.0	GPa
	Particle Stiffness Ratio	0.8	--
	Parallel Bond Stiffness Ratio	0.8	--
	Joint Normal Stiffness	100	GPa/m
	Joint Shear Stiffness	50	GPa/m
Geometric Parameters	Particle Size Range	0.5–0.8	mm
	Initial Domain Aperture	1 × 10 −5	m
Strength Parameters	Parallel Bond Normal Strength	9.0	MPa
	Parallel Bond Shear Strength	6.5	MPa
	Fracture Normal Strength	6.0	MPa
	Fracture Cohesion	12.0	MPa
	Fracture Friction Angle	40	°
Fluid Parameters	Fluid Bulk Modulus	2.0	GPa
	Fluid Dynamic Viscosity	0.1	Pa·s
	Aperture-Pressure Coefficient	1.0	N

. Among them, the PFC microscopic parameters adopted in this paper are mainly determined by the macroscopic-mesoscopic parameter calibration method. Firstly, core sampling was carried out on-site at the Wutongzhuang mine, and the results of indoor mechanical tests of the rock mass (such as uniaxial compressive strength, elastic modulus, etc.) were obtained. Subsequently, taking it as the calibration target, the calibration was carried out by iteratively running numerical experiments and adjusting the microscopic parameters. In addition, the fluid parameters are determined based on the slurry rheological test and the reverse analysis of the rock mass permeability coefficient.

Finally, based on the field grouting environment, the following three typical grouting modes were systematically analyzed, and the influence patterns of different grouting effects were studied:

(1)

Grouting under conventional geological conditions: Simulating crack development and microseismic response during the three stages of filling, permeation, and fracturing.

(2)

Grouting under densely fractured conditions: Setting different fracture densities ($f_d$ = 6, 8, 10) to study their impact on grouting-induced fracturing.

(3)

Grouting near fault structures: Setting different distances between the grouting hole and the fault (0.5, 0.6, 0.7 m) to study the guiding effect of faults on cracks and the clustering of microseismic events.

(4)

Influence mode of grouting effectiveness: Simulating the inhibitory effect of increased rock mass strength after grout solidification on crack propagation during subsequent grouting.

4.3. Simulation Results and Field Verification

4.3.1. Microseismic Response Characteristics Induced by Grouting Under Conventional Geological Conditions

(1)

Analysis of numerical simulation results

Grouting fracturing is different from hydraulic fracturing, and the whole process is more complex. Grouting can be subdivided into three stages. The initial grouting stage is the first stage, mainly involving grout filling, characterized by self-priming grouting without pump pressure. The middle grouting stage is the second stage, mainly involving permeation grouting, where the pump pressure begins to rise closely related to grouting pressure. The final grouting stage is the third stage, mainly involving permeation and fracturing grouting, where the grouting pressure remains at a high level and fracturing easily occurs. The transition boundary between permeation grouting and fracturing grouting is difficult to define, so microfractures during grouting are mainly concentrated in these two stages.

As shown in Figure 7a,b, in the crack morphology diagrams, magenta represents bond breakages (generated cracks), and green represents the grout diffusion zone. When the running step is < 200,000 steps, this corresponds to the initial grouting stage, dominated by filling. During the stage, grout primarily fills existing pores and fractures near the borehole, resulting in limited diffusion range and minimal development of microfractures (dip angles: 0–140°). The tensile stress at the crack tips is low [31], and the total number of cracks is 24 and 27, corresponding to 24 and 27 microseismic events, respectively.

As shown in Figure 7c, when the running step is between 200,000 and 300,000 steps, it enters the middle grouting stage, dominated by permeation grouting. Since the grout pressure is higher than the surrounding pore pressure, under the pressure difference, on one hand, the grout diffuses outward, and on the other hand, the pore pressure increases, reducing the effective stress on the rock particle skeleton, causing the aperture of micro-fractures to increase, further facilitating grout flow. At this time, the grout diffusion range and microfractures begin to show an increasing trend, the tensile stress at crack tips begins to increase, the total number of cracks is 39, corresponding to 39 microseismic events, but the crack propagation angle shows no significant change.

As shown in Figure 7d–f, when the running step is >300,000 steps, it enters the final grouting stage, dominated by permeation and fracturing grouting. The pressure difference not only drives grout flow but also causes stress concentration at the throats of fractures [32,33]. The fracture aperture and total length increase significantly, forming dominant channels locally. At this time, the grout diffusion range and cracks first increase sharply, then the growth rate slows down, and obvious tensile stress concentration appears at the crack tips. The total number of cracks is 84, 121, and 136, corresponding to 84, 121, and 136 microfractures, respectively, but the crack propagation angle shows no significant change. The above results can be reasonably inferred that when the grouting pressure varied by ±15%, the changes in both the b-value and the number of primary cracks were less than 10%, indicating that the system maintained a stable three-stage evolutionary pattern.

(2)

Field verification of microseismic response

Taking a single-hole grouting process as an example, grouting lasted 23 days, with a hole bottom pressure of 24.24 MPa, hole orifice pressure of 12 MPa, total grouting volume of 780 tons, and cement grout density controlled at 1350 kg/m³, as shown in Figure 8.

In the early stage of single-hole grouting, the hole orifice pressure remained at a low level, not exceeding 3.1 MPa, while the grouting volume continued to increase. During this period, grouting was mainly filling, meaning the grout filled the fractures in the surrounding rock mass primarily by gravity, inducing few microfractures, with a maximum daily microseismic frequency not exceeding 18 events. This is consistent with the simulation results for running steps < 200,000.

In the middle stage of single-hole grouting, which lasted longer, the hole orifice pressure began to increase, fluctuating significantly with a maximum value of 6.1 MPa. The grouting volume continued to increase, but the grouting rate decreased in the latter part of this period. This period was mainly permeation grouting, meaning the grout interacted with the surrounding rock mass, accompanied by rock fracturing. The induced microseismic events showed an increasing trend, with a maximum daily microseismic frequency of 28 events. This is consistent with the simulation results for running steps between 200,000 and 300,000.

In the final stage of single-hole grouting, which was short, the hole orifice pressure rapidly increased to 12.4 MPa and stabilized for over 30 min. The pump rate was below 35 L/min, and the grouting rate was low. This period was mainly fracturing grouting, meaning high-pressure grout caused stress concentration at the throats of pores and fractures, leading to fracture propagation and extension. The induced microseismic events were at a high level, with a maximum daily microseismic frequency of 38 event.

In summary, combining simulation analysis and microseismic response results during grouting, it can be found that the crack distribution patterns caused by grout permeation and fracturing correspond well with microseismic monitoring. That is, the increase in the number of microfractures during the high-pressure grout permeation and fracturing stages corresponds to an increasing trend in microseismicity. To quantitatively validate the consistency between the simulation and the field response, we calculated the sum of the squares of the total crack lengths per unit area (ΣL²) for models at different grouting stages, using it as a proxy indicator for cumulative microseismic energy release. The results (Figure 8) show that ΣL² was 1.8 × 10⁴ mm²/m² during the filling stage, increased to 4.5 × 10⁴ mm²/m² during the permeation stage, and rose sharply to 2.1 × 10⁵ mm²/m² during the fracturing stage. This order-of-magnitude growth trend aligns well with the significant enhancement observed in the daily frequency and energy density of field microseismic events from the early stage (<18 events/day) to the late stage (38 events/day), quantitatively confirming that the later grouting stage is the primary phase for energy release and rock mass modification.

4.3.2. Microseismic Response Characteristics Under Densely Fractured Conditions

(1)

Analysis of numerical simulation results

This section investigates the influence of the development degree of pre-existing fractures on grouting fracturing behavior by controlling the fracture density parameter $f_d$. Under the premise of keeping the physical and mechanical properties of the rock mass and the in situ stress conditions constant, fracture networks with fracture sizes following a power-law distribution and spatial locations following a uniform distribution were constructed in the numerical model. Three levels of $f_d$ = 6, 8, 10 were set to quantitatively characterize the severity of rock mass cutting by fractures. The simulation results show that the crack distribution pattern induced by grouting is significantly controlled by the density of pre-existing fractures. As shown in Figure 9, as the $f_d$ value increases, the total number of cracks in the system shows a non-linear growth trend. Specifically, when $f_d$ = 6, the rock mass structure is relatively intact. The grout mainly propagates by fracturing along the direction of the maximum principal stress, forming 3 dominant cracks. Their morphology is similar to the propagation mode in a homogeneous rock mass but accompanied by numerous secondary microfractures. The total number of cracks is 285, an increase of about 109.6% compared to the homogeneous model ($f_d$ = 0) with 136 cracks. When $f_d$ increases to 8, the interaction between fractures enhances, and the crack morphology changes significantly. The number of dominant cracks increases from 3 to 6, and the length of individual cracks significantly extends. The total number of cracks reaches 683, an increase of 139.7%, showing the strengthening effect of fracture density on the complexity of the fracture network. When $f_d$ = 10, the rock mass is highly fragmented. The grouting pressure drives the grout to migrate along the dense pre-existing fracture network, inducing an order-of-magnitude increase in the number of cracks. At this point, new cracks extensively intersect with pre-existing fractures, forming a highly complex, interconnected fracture structure, and tensile stress concentration at crack tips is more significant, reflecting that grouting in extremely fractured rock masses is more likely to trigger dense, distributed fracturing.

In summary, the density of pre-existing fractures not only determines the final scale of grouting-induced fracturing but also controls the morphological structure and evolution path of the fracture network. It is a key geological factor for evaluating the disturbance effect of grouting and the rock mass modification effect. In addition, a 15% increase in fracture density led to a decrease in the b-value by approximately 0.3 and an increase in the number of primary cracks by about 40%, confirming its stable control over fracture complexity.

(2)

Field Verification of the Effect of Pre-existing Fracture Aggregation Degree on Microseismic Response

As shown in Figure 10, pre-mining network parallel electrical method detection identified a low-resistivity zone in the haulage roadway between 600 m and 830 m. The apparent resistivity value in this area is below 20 Ω·m, indicating high rock fragmentation, high fracture density, and strong water abundance [34]. During the floor grouting reinforcement engineering in the Shanfuqing limestone for the working face, a dense cluster ofmicrofractures formed within the Shanfuqing limestone range inside the electrical low-resistivity zone, while microfractures outside the low-resistivity zone were discretely distributed without clustering tendency. Simultaneously, the microseismic energy kernel density value within the electrical low-resistivity zone was significantly higher than in the surrounding normal rock mass areas. This is consistent with the simulation results of the generalized model considering the influence of pre-existing fracture development.

In summary, areas with more developed pre-existing fractures exhibit more significant interaction between grout and the surrounding rock mass, a higher probability of inducing fractures, and are more likely to trigger microfractures. That is, primary microseismicity formed by grouting disturbance can reveal hydrogeologically abnormal zones. By quantifying the energy proxy indicator (ΣL²) for models with different fracture densities, we found that it increases sharply with the degree of pre-existing fracture development: when the fracture density ρ increased from 6 to 10, ΣL² jumped from 5.2 × 10⁴ mm²/m² to 3.5 × 10⁶ mm²/m², an increase of nearly two orders of magnitude. This quantitative result clearly explains why the low-resistivity zone identified by electrical surveying (a zone of high fracture density) in the field corresponds to an area of high microseismic energy kernel density—not only due to an increase in the number of microseismic events but also because the fractures are larger in scale and release more energy.

4.3.3. Microseismic Response Characteristics Induced by Grouting near Fault Structures

(1)

Analysis of numerical simulation results

To deeply investigate the controlling effect of fault structures on grouting-induced fracture behavior, this study designed three sets of numerical simulation schemes with distances between the grouting hole and the fault of 0.7 m, 0.6 m, and 0.5 m, respectively, to systematically analyze the variation mechanism of crack propagation laws and microseismic response characteristics near faults. The simulation results shown in Figure 11 reveal that the spatial distribution and propagation paths of grouting-induced cracks are strictly controlled by their distance from the fault. As the distance decreases, the crack system evolves from directional propagation to fault-oriented clustering: At a distance of 0.7 m, although the crack propagation maintains a pattern dominated by 3 main cracks, their extension direction shows significant deflection towards the fault, indicating the remote disturbance effect of the fault on the stress field. When the distance reduces to 0.6 m, the attractive effect of the fault on cracks further enhances. New cracks are more inclined to develop towards the fault. The number of main cracks increases to 5, and the total number of cracks reaches 447, showing that the fault begins to dominate the fracture direction as a guiding plane. Under the extremely close condition (0.5 m), the crack propagation behavior undergoes a qualitative change. The number increases by an order of magnitude to 1265 (the crack number increased by 283%) and is highly concentrated in the narrow area between the grouting hole and the fault, forming a typical “grouting-hole-fault” interconnected fracture zone.

We statistically analyzed the length of all tensile cracks. The proportion of long cracks (>10 mm) significantly increased from 18.2% at 0.7 m to 41.5% at 0.5 m. This indicates that proximity to the fault not only generates more cracks but also promotes the propagation and coalescence of longer, more influential fractures. Moreover, In the PFC model, each microfracture represents a microseismic event. We used the square of the crack length (L²) as a proxy for the released seismic energy. The equivalent microseismic energy density (total L² per unit area) was calculated. The energy density increased dramatically by 605% when the distance reduced from 0.7 m to 0.5 m, highlighting the intense energy release and higher risk associated with grouting in close proximity to the fault. In addition, when the fault distance increased to 0.575 m, the b-value recovered by about 0.4 and the number of primary cracks decreased by approximately 60%, strictly validating the critical distance effect.

To transform the analysis of grouting near faults into practical guidance, this study defines a critical distance $Dcr$ based on simulation results, which is the ratio of the distance between the grouting hole and the fault to the diameter of the borehole. Based on simulation analysis of indicators such as fracture quantity, length, and microseismic energy, it was found that when the distance was as small as about 0.6 m, the fracture behavior underwent a qualitative change, shifting from being controlled by far-field stress to being dominated by faults. Therefore, when the distance between the grouting hole and the fault is less than about 0.6 m (the on-site drilling diameter is about 0.152 m, and the ratio is about 3.9), the fracture propagation mode undergoes a fundamental change: (1) In the critical zone ($Dcr$ < 3.9): The rupture is dominated by the fault, and the fractures are strongly attracted towards the fault direction. Microseismic events are highly concentrated, energy release increases dramatically, and the risk of grouting activating the fault is extremely high. (2) Outside the critical zone ($Dcr$ > 3.9): the fracture propagation is mainly controlled by the regional stress field, and the influence of faults manifests as limited directional deviation. In addition, the analysis of the stress transmission path indicates that the fault, as a mechanical weak plane, effectively blocks the continuous transmission of force chains, resulting in a “blank” characteristic in the force chain distribution at its contact surface. Meanwhile, the grouting pressure cannot dissipate along the fault but becomes highly concentrated in the hole-fault interval, forming a thick, high-density force chain network and a wide tensile stress zone, thereby driving the dense initiation and propagation of cracks in this area. These patterns indicate that when grouting near a fault, the fracture behavior transitions from “stress-field-controlled directional propagation” to “structural-weak-plane-controlled guided clustering.” This conclusion provides an important theoretical basis for assessing fault activation risks and optimizing grouting design in structural zones.

(2)

Field monitoring of microseismic influence in fault structure areas

As shown in Figure 12, there is an F602-4 fault in the haulage roadway between 370 m and 480 m. During the floor grouting reinforcement engineering in the Shanfuqing limestone for the working face, a dense cluster of microfractures formed within the Shanfuzi limestone range near the F602-4 fault area. In contrast, microfractures far from the fault area were fewer, discretely distributed, and showed no clustering tendency. Simultaneously, the microseismic energy kernel density value near the F602-4 fault area was significantly higher than in the surrounding normal rock mass areas. This is consistent with the simulation results of the generalized model considering the influence of fault structures.

In summary, in relatively fractured areas near fault structures, the interaction between grout and the surrounding rock mass is more significant, the probability of inducing cracks is higher, and microfractures are more easily triggered. Primary microseismicity formed by grouting disturbance can reveal structural abnormal zones such as faults. The simulation shows that when the distance from the grouting hole to the fault decreased from 0.7 m to the critical distance of 0.5 m, ΣL² increased non-linearly and drastically from 8.7 × 10⁴ mm²/m² to 2.7 × 10⁶ mm²/m², representing a growth rate exceeding 3000%. This data quantitatively reveals the high-risk nature of grouting in close proximity to the fault: not only does the fracture pattern change, but the intensity of energy release increases exponentially. This result provides a direct numerical correspondence with the extremely high values of microseismic energy kernel density observed near the F602-4 fault in the field, strongly confirming the spatial aggregation effect of fault structures on grouting-induced microseismic energy. Therefore, the following specific safety measures have been proposed, including: (1) maintaining a wellbore diameter of at least four times (about 0.6 m) between the grouting hole and the large fault to prevent fault fracturing, (2) implementing a gradual pressurization plan in the affected area of the fault to maintain the injection pressure below the critical level; otherwise, unstable cracks may spread towards the fault; and (3) implementing a real-time monitoring strategy which involves utilizing real-time microseismic monitoring to dynamically adjust grouting parameters.

4.3.4. Differences in Microseismic Response Under Different Grouting Effectiveness

(1)

Analysis of numerical simulation results

As the grouting process proceeds, fractures in the fractured rock mass are filled, increasing the compactness and reducing the connectivity of the rock mass. To study the impact of grouting reinforcement effectiveness on microfractures, a generalized model of grouting effectiveness influence was constructed, as shown in Figure 13. Figure 13a shows the result at a grouting pressure of 24.4 MPa after 250,000 running steps. In Figure 13b, the green and yellow represent solidified grout particle entities. Bonding was added after flow stopped to simulate the solidification process. It can be observed that when two grouting holes are close together, the crack propagation direction changes, converging horizontally. From Figure 13c, it can be seen that after the grout solidifies, a third grouting hole is created between the two original holes for grouting. Since the grout particles are small, absolute numbers are not ideal for comparison; therefore, the crack propagation range is used to judge the number of microfractures. From Figure 13c, it is found that due to grout solidification, which enhances the strength of the surrounding rock mass, the crack propagation range significantly decreases. Moreover, the three main cracks at 120° change to one nearly horizontal main crack, and the length of the main crack significantly decreases. Therefore, when grouting is effective, after grout solidification forms bonds, the rock mass strength increases, and the number, range, and length of cracks generated by subsequent grouting are significantly reduced.

(2)

Field Verification

As shown in Figure 14a, in the early stage of the regional treatment project, the microseismic kernel density area covered most of the outer section of the 182602 working face, with a maximum value of 0.00024 events/m². As the grouting project proceeded, only a very small kernel density area of microseismic events existed in the F602-4 fault area. In the later stage of the regional treatment project, there was no microseismic kernel density area coverage within the working face, only scattered microseismic events at a low level (see Figure 14b). This is consistent with the simulation results of the generalized model of grouting effectiveness influence. In summary, due to the continuous progress of grouting, the grout fills and reinforces the pre-existing fractures, the rock mass strength improves, the grouting fracturing effect weakens, resulting in areas with good grouting reinforcement effectiveness being less prone to produce microseismic effects.

5. Laboratory Test on Signal Characteristics of Grouting Induced Microfracture

5.1. Test Device

Based on the analysis of mechanics and numerical micro mechanisms, laboratory tests were conducted. This test used an inverse investigation approach to study the vibration signal characteristics generated when a porous rock mass is subjected to high-pressure grout action, thereby approximately simulating the microseismic response characteristics during the grouting process. A self-developed test apparatus was used. The main model dimensions are φ150 × 1000 mm, and the diameters of the water injection hole and outlet hole are designed as 13 mm. The main test model consists of two 500 mm long organic glass tubes with flanges. Uniformly sized limestone crushed stone is placed inside the tubes to simulate the porous rock mass matrix, and geophones are buried within the crushed stone. One side of the main model is connected to a water injection pressurization device. A pressure gauge is connected to the middle of the tube on the water injection side (to monitor changes in pore water pressure in the rock mass), and the microseismic signal acquisition system is located in the middle of the tube on the outlet side (to monitor microseismic signals). Flowing water enters from the injection hole through the pressurization device and flows out from the outlet. During the test, one microseismic substation and one microseismic geophone from the KJ1073 microseismic system were used to complete the acquisition, recording, and analysis of vibration signals. The test apparatus and microseismic acquisition system are shown in Figure 15. During the test, microseismic signals were acquired with a primary frequency range of 1–500 Hz, and the STA/LTA was set to 2.5.

5.2. Scheme Design

The microseismic signals generated by grouting are the result of multiple factors. Ignoring secondary factors, grouting pressure, pore water pressure in the rock mass, and matrix grain size were primarily selected as variables. A comprehensive test was conducted using the control variable method to obtain the waveform characteristics of microseismic signals generated by grouting. This experiment determined 2 groups with 4 types of test schemes. The specific sequences are shown in

Table 2. The sequence of pilot scenarios.

Test Scheme	Grouting Pressure	Matrix Grain Diameter	Pore Rock Mass Water Pressure (MPa)
A	1 MPa	4 mm	0.02	0.04	0.06	0.08	0.1	0.12	0.14	0.16	0.18	0.2
B		10 mm
C	2 MPa	4 mm	0.02	0.04	0.06	0.08	0.1	0.12	0.14	0.16	0.18	0.2
D		10 mm

.

(1)

Grouting pressure is 1 MPa, divided into two Schemes A and B based on matrix grain size.

Pore rock mass water pressures were set to 0.02, 0.04, 0.06, 0.08, 0.1, 0.12, 0.14, 0.16, 0.18, and 0.2 MPa, respectively. Matrix grain sizes of 4 mm and 10 mm were selected for repeatability tests.

(2)

Grouting pressure is 2 MPa, divided into two Schemes C and D based on matrix grain size.

Pore rock mass water pressures were set to 0.02, 0.04, 0.06, 0.08, 0.1, 0.12, 0.14, 0.16, 0.18, and 0.2 MPa, respectively. Matrix grain sizes of 4 and 10 mm were selected for repeatability tests.

Uniformly sized limestone crushed stone (4/10 mm) was loaded into the organic glass tubes, and geophones were buried. Then, the pressure gauge, water injection device, and microseismic acquisition main station were connected to construct the porous rock mass system, water injection pressurization system, and microseismic acquisition system. The porous rock mass was first subjected to water injection pressurization to check system airtightness. The grouting pressure remained constant during the test (1/2 MPa). The pore water pressure in the rock mass gradually increased as the test proceeded. Microseismic signals were acquired every time the water pressure in the organic glass tube changed by 0.02 MPa.

5.3. Test Results

(1)

Microseismic signal waveform characteristics

As shown in Figure 16. When the pressure in the pore rock mass system, matrix grain size, and water injection pressure change, the waveforms are similar, being irregular simple harmonic waves. The frequency distribution ranges from 0 to 50 Hz, the average duration is between 300 and 500 ms, the time interval between upper and lower peaks ranges from 20 to 125 ms, there is a developing coda wave, the attenuation is relatively slow, and the amplitude ranges from 15 mV to 60 mV.

(2)

Influence of pore rock mass water pressure changes on microseismic amplitude

Under the same conditions, multiple repeated tests were conducted. The maximum amplitude of the microseismic signal showed a non-linear phased change with the increase in pore rock mass water pressure, as shown in Figure 17. The test results show that the amplitude of the acquired microseismic signals generally follows a trend of increase–decrease–increase, which can be divided into two stages. The first is the grouting channel penetration stage. During this stage, the pressure value in the pore rock mass system is generally in the range of 0–0.06 MPa, and the microseismic signal amplitude increases abnormally. Analysis suggests that in this stage, there are many dead-end fractures between the limestone crushed stones. The dynamic pressure grout acts on the porous rock mass, causing the dead-end pores between the crushed stones at the injection and outlet ports to rapidly extend and connect, forming a grouting channel nearly along the tube axis. Frequent increases, decreases, and even closures of fracture apertures lead to an abnormal increase in the amplitude of the acquired microseismic signals. The second stage is the continuous action stage of dynamic pressure grout. During this stage, the pressure value in the pore rock mass system ranges from 0.04 MPa to 0.2 MPa, and the microseismic signal amplitude shows a gradually increasing trend. Analysis suggests that as the grout pressure increases, its interaction with the porous rock mass becomes more intense, and the generated energy shows an increasing trend, so the microseismic signal amplitude also shows an increasing trend.

(3)

Influence of matrix grain size and grouting pressure on microseismic amplitude

Since multiple repeated tests were conducted, multiple sets of microseismic signal amplitudes were obtained under the same conditions. The coefficient of variation can represent the relative dispersion of the data, i.e., the degree of variation in amplitude [35,36]. The coefficient of variation is introduced here to quantitatively analyze the differential changes in microseismic signal amplitude under different matrix grain size conditions. Taking Schemes A and B as examples, the coefficient of variation of microseismic signal amplitude is shown in Figure 18a. When the water pressure is between 0.02 MPa and 0.08 MPa, the coefficient of variation increases abnormally, all exceeding 10%. When the water pressure is greater than 0.08 MPa, the coefficients of variation are all less than 10%. This indicates that when the pore rock mass water pressure is between 0.02 MPa and 0.08 MPa, grouting causes large-scale opening or closing of fractures, penetrating the near-axial grouting channel, and the amplitude increases abnormally. Furthermore, the average peak amplitude of the microseismic waves monitored in each group was calculated. Keeping the grouting pressure constant, under the same water pressure condition, the increase in matrix grain size led to an increase in the average amplitude ranging from 50.6% to 68.7% (comparing Schemes A and B) and 50.6% to 68.4% (comparing Schemes C and D). This indicates that when the matrix grain size increases (i.e., porosity increases), the degree of amplitude variation is greater, and signals with larger amplitudes are more easily generated. Analyze the reasons as follows: Rock masses with high porosity possess a weaker skeletal structure [37], characterized by reduced interparticle contacts and lower cementation strength. Under grouting pressure, fracture propagation is no longer confined to a single crack plane but tends to undergo larger-scale brittle failure, either trans-granular or inter-granular. This failure mode involves a larger rupture area and more significant particle displacement, effectively releasing a greater amount of accumulated strain energy in a single event. According to seismological principles, microseismic amplitude is closely related to the seismic moment, which is determined by the rupture area, average dislocation, and medium rigidity. Consequently, the large-scale fracturing occurring in high-porosity media generates a larger seismic moment, radiating stronger seismic wave energy, thereby ultimately manifesting as a significant increase in the amplitude of the collected signals. Furthermore, the instability of the force chain network in high-porosity media also leads to more intense and intermittent fracturing dynamics, which aligns with the observed phenomenon of an increased amplitude variation coefficient. Furthermore, we conducted spectral and b-value analyses on the collected microseismic signals. The results demonstrate that when the grouting pressure increases from 1 MPa to 2 MPa or the matrix grain size increases from 4 mm to 10 mm, the dominant frequency of the signals shifts significantly toward lower frequencies with a broadening spectral shape. Simultaneously, under conditions of high grouting pressure and large matrix grain size, the b-value shows a systematic decrease. These characteristics comprehensively elucidate the microseismic features of grouting-induced fracturing in different fractured rock masses from multiple perspectives.

As shown in Figure 18b, taking Schemes A and C as examples, under the test condition with a water injection pressure of 2 MPa, the coefficient of variation for each set of signal amplitudes is greater than that under the 1 MPa water injection pressure. This indicates that when the water injection pressure increases, the degree of amplitude variation is greater, signals with larger amplitudes are more easily generated, and the risk increases.

In summary, in actual grouting processes, facing fractured rock masses with different pore water pressures and different apertures, different grouting pressures will lead to significant differences in microseismic signals. Based on the microseismic waveform, amplitude, and other characteristics obtained from this experiment, the field can invert the integrity of the fractured rock mass and further optimize the grouting process. In addition, there are scale differences between laboratory research and field conditions, but their design follows similarities in key physical laws. The laboratory model utilized a φ150 mm acrylic tube filled with crushed limestone, simulating in situ grouting stress conditions by controlling grouting pressure (1–2 MPa) and pore water pressure (0.02–0.2 MPa). Although the absolute stress values in the laboratory were lower than those at kilometer-depth field conditions, we maintained mechanical similarity between experimental and field stress states by controlling the ratio of grouting pressure to rock mass strength. In terms of seepage characteristics, the slurry viscosity and injection rate were adjusted to ensure the Reynolds number in the laboratory was of the same order of magnitude as field conditions, thereby maintaining consistency in the grout flow regime. The amplitude (15–60 mV) and spectral characteristics (0–50 Hz) of the microseismic signals observed in the experiments exhibited consistent response patterns with field monitoring results, validating the applicability of the fracture mechanisms obtained under laboratory conditions to field-scale scenarios.

6. Conclusions

This study, through a combined research method of mechanical analysis, numerical simulation, field microseismic monitoring, and laboratory tests, multi-dimensionally reveals the evolution patterns and control mechanisms of microfractures during the grouting process in fractured rock masses. The main conclusions are as follows:

(1)

The process of grouting-induced microfracture undergoes three stages: fracture filling, coupled permeation, and fracturing propagation. Among them, the fracture filling stage hardly produces new fractures. The coupled permeation stage forms a mechanical mechanism where microfracture initiation is promoted by the reduction in effective stress. The propagation criterion in the fracturing propagation stage is that the stress intensity factor on the fracture surface reaches the rock mass fracture toughness, with the key being the grouting net pressure ($P_f-{\sigma }_3$).

(2)

The PFC fluid–solid coupling simulation reproduced the fracture modes under three typical geological conditions: Under conventional geological conditions, microfractures show a “three-stage” evolution, with cracks mainly propagating along the direction of the maximum principal stress. The number of cracks produced in the late grouting stage was 136, more than 5 times that in the early stage. Under densely fractured conditions, when the fracture density $f_d$ increased from 6 to 10, the total number of fractures increased non-linearly from 285, significantly controlling the morphological structure and evolution path of the fracture network. Under fault structure conditions, when the distance between the grouting hole and the fault decreased from 0.7 m to 0.5 m, the main cracks composed of 447 fractures transitioned from directional propagation to being highly concentrated in the narrow area between the grouting hole and the fault, showing the strong guiding effect of the fault on the fracture path. These patterns are highly consistent with the field microseismic monitoring results where microseismic events form dense clusters in the electrical low-resistivity zone (high fracture density) and near the F602-4 fault, confirming that microseismic activity is a direct response to the combined action of geological structures and grouting disturbance.

(3)

Numerical simulation shows that after grout solidification and re-grouting, the crack propagation range and number significantly decrease. Effective grouting reinforcement can significantly improve rock mass strength and inhibit fracture development during subsequent grouting. Field microseismic monitoring confirmed that with the advancement of the regional treatment project, the microseismic kernel density area decreased from a maximum of 0.00024 events/m² in the pre-treatment period to having no kernel density area coverage within the working face in the post-treatment period, proving effective improvement in rock mass integrity.

(4)

Laboratory test results show that microseismic amplitude is jointly influenced by grouting pressure and rock mass structure: when the pore water pressure is between 0.02 and 0.08 MPa, the coefficient of variation of microseismic amplitude exceeds 10%, reflecting the intense process of grouting channel penetration. Increasing the matrix grain size from 4 mm to 10 mm (i.e., increasing porosity) led to an average amplitude increase of 50.6–68.7%. The research results provide a direct basis for using microseismic waveform, amplitude, and other characteristics to invert fracture degree of underground rock mass and optimize the grouting process.