Abstract
Mine geological structures are the core elements controlling safe coal production and serve as the primary conduits inducing groundwater inrush accidents. To address the issues of incomplete evaluation indicators and single weighting methods in assessing mine structural complexity, this study takes Guobei Coal Mine as the research object and constructs a quantitative evaluation model for mine structural complexity based on the coupling of GIS, AHP, and the Independence Weight Coefficient Method. Four core indicators were selected: fault density, fault intensity, fracture fractal dimension, and fold planar deformation coefficient. The Analytic Hierarchy Process (AHP) was employed to determine subjective weights, while the Independence Weight Coefficient Method was used to determine objective weights, and the multiplicative synthesis normalization method was applied to achieve the fusion of subjective and objective weights. Spatial quantitative analysis and thematic map overlay were completed using ArcGIS, dividing the study area into four levels of structural complexity zones. The results indicate that: (1) fault intensity exerts the greatest influence on structural complexity (comprehensive weight of 0.4212), followed by the fold planar deformation coefficient (0.2894) and fracture fractal dimension (0.2516); (2) the structurally complex zones and relatively complex zones exhibit three concentrated distributions, located in the northeastern, central, and southern regions of the mine, respectively, the structurally complex and relatively complex zones accounting for 15.8% and 24.3% of the total mine area, respectively; (3) specifically in the Mining Areas 84 and 85, the structurally complex and relatively complex zones account for 21.1% and 41.0% of their respective district areas. Groundwater inrush points in the mining area are predominantly distributed within or adjacent to the structurally complex and relatively complex zones, with particular attention required for water inrush threats from roof fracture zones and fault intersections. The research findings accurately characterize the spatial differentiation of structural complexity in Guobei Coal Mine, providing targeted objectives and scientific geological basis for mine water hazard prevention and control. Simultaneously, the study offers methodological references for quantitative evaluation of structural complexity in mines with similar geological conditions, thereby enriching the theoretical and methodological framework of quantitative mine structural evaluation.
1 Introduction
Mine geological structures are the core geological elements controlling safe coal production. Fault and fold structures not only directly influence the occurrence morphology of coal seams and mining layout, but also serve as the primary conduits inducing mine water inrush accidents. Their development characteristics and spatial distribution directly determine the complexity of hydrogeological conditions within the mining field [–]. As one of the major disasters threatening the safe and efficient extraction of coal resources, water hazards are closely associated with the development of fault and fold structures in most water inrush accidents occurring in coal mines worldwide. Issues such as roof fracture connectivity and fault zone activation in structurally complex areas can readily trigger the sudden inflow of aquifer water into mining working faces, resulting in production interruptions and even casualties. Therefore, conducting quantitative evaluation of mine structural complexity, accurately characterizing the spatial differentiation of structural development within the mining field, and revealing the intrinsic relationship between structural development and water inrush hazards constitute important geological foundations for mine water hazard prediction and prevention. Such efforts hold significant theoretical research value and engineering practical significance for the rational layout of coal mining operations and the formulation of safe extraction technical measures.
In recent years, domestic and international scholars have conducted extensive research on the quantitative evaluation of mine structural complexity, with evaluation methods gradually evolving from traditional qualitative descriptions toward quantitative analyses based on multi-indicator and multi-method coupling [, ]. Regarding the selection of evaluation indicators, fault density and fault intensity have been widely employed to characterize the development degree of fault structures, while the introduction of fractal theory has compensated for the limitations of single density or intensity indicators, making fracture fractal dimension an effective indicator for quantitatively reflecting the complexity of fracture networks [–]. For fold structures, the application of the fold planar deformation coefficient has achieved quantitative characterization of fold deformation degree, promoting the development of dual-factor comprehensive evaluation integrating both faults and folds [–]. In terms of evaluation methodologies and technical approaches, the Analytic Hierarchy Process (AHP), by virtue of its advantages in multi-factor hierarchical decision-making, has become a classical method for determining subjective weights of evaluation indicators [–]. The Independence Weight Coefficient Method can effectively eliminate information redundancy among indicators, providing an important approach for determining objective weights [–]. Geographical Information System (GIS), with its powerful capabilities in spatial data processing and overlay analysis, has become the core technology for spatial visualization and quantitative analysis of mine geological structural elements, and has been widely combined with various mathematical evaluation methods for zoning studies of mine structural complexity [–].
Furthermore, with the rapid development of artificial intelligence, machine learning methods such as deep learning, deep belief networks, and Bayesian networks have been increasingly applied to water inrush prediction and causal analysis, providing new pathways from structural complexity evaluation to quantitative risk forecasting [–]. Meanwhile, three-dimensional geological modeling and discrete fracture network techniques have enabled more accurate characterization of fault vertical cutting depths and fracture connectivity, offering refined data foundations for structural complexity analysis []. Multi-method fusion models combining AHP, entropy weight, and grey system theory have also been widely employed in aquifer water-richness evaluation [].
Despite the substantial achievements in evaluation indicators and methodological systems attained by existing studies, numerous issues remain to be addressed in practical applications targeting specific mining fields. First, most studies employ either subjective or objective weighting methods alone, which readily leads to weight allocation bias, making it difficult to simultaneously account for expert experience and the inherent patterns within the data, thereby compromising the accuracy of evaluation results. Second, some studies provide insufficiently comprehensive characterization of structural elements, either neglecting the quantitative evaluation of fold structures or failing to introduce fractal theory to depict the complex characteristics of fracture networks, rendering the evaluation results unable to fully reflect the actual development status of structures within the mining field. Third, existing research predominantly focuses on zoning of structural complexity across the entire mining field, with insufficient integration of evaluation results with mine water inrush regularity, present-day in situ stress field characteristics, and structure-controlled water regularity. This disconnect between evaluation outcomes and practical engineering applications makes it difficult to provide targeted technical guidance for water hazard prevention and control in specific mining districts. This study specifically targets structure-controlled water inrush, which constitutes the majority of water hazards in Guobei Coal Mine. Other hydrological and mining factors will be incorporated in subsequent studies to develop a comprehensive water inrush early-warning system.
Guobei Coal Mine is located in the southeastern margin of the North China Plate, in the central section of the Xu-Huai arcuate structural belt, with a geotectonic position belonging to the northern margin influence zone of the collision-suture belt between the Yangtze Plate and the North China Plate. The regional structural evolution has undergone the following key stages as shown in Table 1.
TABLE 1
| Structural stage | Era | Tectonic events | Impact on the mining area |
|---|---|---|---|
| Basement formation period | Archean–Early proterozoic | Regional metamorphism and magmatic intrusion | Formation of crystalline basement |
| Cover development period | Late proterozoic–Early paleozoic | Stable subsidence, marine sedimentation | Formation of carbonate-clastic rock cover |
| Indosinian movement period | Triassic | Near E-W compression from N-S plate collision | Formation of near N-S trending fold basement structures |
| Yanshan movement period | Jurassic–Cretaceous | Intensive NW-SE compression | Formation of NE-NNE trending fault framework, superimposed modification of early folds |
| Himalayan period | Paleogene–Neogene | Regional extension, differential uplift | Fault reactivation, modern landform formation |
Geological structure background of Guobei Coal Mine.
The main structural framework of the mining area is characterized by the Yanshan Movement NE-NNE trending fault system superimposed on Indosinian near N-S trending folds, forming a “fault-fold composite” structural style. This structural evolution history determines the complexity features of intensive fault development and fold superposition in the mining area.
From 2005 to 2022, multiple water inrush accidents occurred cumulatively, with the majority closely related to structural development. Structural complexity has become a critical factor constraining safe mining operations at this mine. In particular, the Mining Areas 84 and 85, serving as the core excavation areas, urgently require investigation into their structural development characteristics and water control regularity. Based on this context, the present study takes Guobei Coal Mine as the research area and conducts quantitative evaluation of mine structural complexity grounded in multi-source information fusion theory. Four core indicators were selected—fault density, fault intensity, fracture fractal dimension, and fold planar deformation coefficient—to comprehensively characterize the development features of both fault and fold structures. The Analytic Hierarchy Process (AHP) was employed to determine subjective weights, while the Independence Weight Coefficient Method was used to determine objective weights; the multiplicative synthesis normalization method was then applied to achieve the organic integration of subjective and objective weights, yielding the comprehensive weight of each indicator.
GIS technology was utilized to complete spatial quantitative analysis and normalization processing of each indicator, and a quantitative evaluation model for mine structural complexity based on the coupling of GIS-AHP-Independence Weight Coefficient Method was constructed to conduct structural complexity zoning across the entire mining field, with particular emphasis on analyzing the structural development characteristics of the Mining Areas 84 and 85. Furthermore, by integrating historical water inrush records of the mine, the structure-controlled water regularity of the study area was revealed. The research findings of this study aim to accurately characterize the spatial distribution features of structural complexity in Guobei Coal Mine, provide scientific geological basis for water hazard prevention and control at this mine, and simultaneously offer methodological references for quantitative evaluation of structural complexity in mines with similar geological conditions, thereby enriching the theoretical and methodological framework of quantitative mine structural evaluation.
2 Construction of quantitative evaluation model for mine geological structural complexity
GIS is a spatial information technology developed in the 1960s, encompassing four fundamental components: computer hardware, software, data, and users. It possesses powerful capabilities in data acquisition, verification and editing, analysis, statistics and computation, spatial analysis, data reconstruction, transformation and generalization, as well as result display and output [].
With the continuous advancement of information technology, GIS has increasingly evolved toward diversified and multidisciplinary integration. It has been widely applied in numerous fields including geosciences, resource management, environmental protection, and disaster prediction, yielding favorable social and economic benefits while substantially broadening conventional approaches to geological work [–]. Given the complex spatial characteristics and strong concealment of mine structures, ArcGIS—with its robust functions for spatial data statistics, management, computation, and analysis, along with its capacity to generate result maps—enables quantitative evaluation of mine geological structural complexity through the fusion analysis of subjective and objective weights, thereby providing a more intuitive representation of structural complexity across the mining field.
2.1 Selection of evaluation indicators for mine structural complexity
The purpose of determining evaluation indicators is to express each structural factor affecting production in a quantitative form; therefore, each selected indicator should effectively reflect the complexity of geological structures. The nature of mine structural prediction and evaluation, combined with the characteristics of coal production, confers specific significance to each indicator.
Based on investigations into the development characteristics of mine structures, fault structures are intensely developed and fold structures are moderately developed within the mine scope. Accordingly, this chapter focuses on the quantitative evaluation of mine structural complexity in the study area by considering the influences of both faulting and folding. For fault structures, fault density, fault intensity, and fracture fractal dimension were selected as evaluation indicators; for fold structures, the fold planar deformation coefficient was chosen as the evaluation indicator.
2.1.1 Fault density
Fault density (number of faults per unit area) serves as an effective evaluation indicator that intuitively reflects the intensity and concentration of fault development when analyzing the distribution and development regularity of faults. To investigate the distribution characteristics of fault structures, the mining field area was divided into 173 rectangular calculation units of 800 m × 800 m based on the mining engineering plan of No. 82 Coal Seam in Guobei Coal Mine, as shown in Figure 1. The number of faults within each unit was quantitatively counted, and contour maps of fault density were compiled using grid subdivision interpolation.
FIGURE 1
2.1.2 Fault intensity
Although fault density reflects fault development to a certain extent, it cannot intuitively represent fault throw and extension length. In some areas, although faults are extensively distributed, their throws and extension lengths are relatively small, resulting in limited impact on mining operations; conversely, in other areas with fewer faults, the faults may be large in scale with substantial extension lengths and throws, exerting significant influence and rendering large areas of coal resources unmineable. Therefore, the fault intensity index (F) is required to evaluate the impact of faults. It is defined as the sum of the products of the extension length and throw of all faults within a unit area, as expressed in Equation 1.where l and h are the horizontal extension length and throw of faults, respectively, in meters; S is the statistical area, in m2; and n is the number of faults.
Its physical significance can be understood at three levels:
“Failure Volume” Potential: li·hi approximately represents the “influence area” of a single fault in the plane (fault trace length × vertical throw), and Σ(li·hi) is the cumulative influence area of all faults per unit area. The larger this value, the more severe the cutting and destruction of the coal seam by faults, and the higher the risk of encountering fault structures during mining.
Direct Correlation with Mining Engineering: In coal mining, the impact of faults on production depends not only on quantity (fault density) but more importantly on “scale.” A fault with length 500 m and throw 30 m versus five faults each with length 100 m and throw 5 m may have the same fault density (assuming the same area), but their impacts on mining are fundamentally different—the former may require complete rearrangement of the working face, while the latter may only need local adjustments. Fault intensity is an effective indicator for quantifying this difference.
Applicability argument for Guobei Coal Mine
Fault development in Guobei Coal Mine features “uneven density and large scale variation”: Mining Area 82 has high fault density but mostly small faults, while Mining Area 81 has moderate fault density but contains large faults such as F22 (throw >50 m). The fault intensity indicator can accurately identify the high risk of the latter, whereas fault density would underestimate its threat. This is precisely the fundamental reason why fault intensity (0.4212) is assigned a far higher comprehensive weight than fault density (0.0378) in this paper.
Since fault intensity considers both fault throw and horizontal extension length as well as fault quantity, it more effectively reflects the impact of faults on production than fault density alone.
The specific statistical procedure for fault intensity is as follows:
Following the statistical method for fault density, divide and number the units on the coal seam floor contour map.
For each unit, count the extension length of every fault (for faults crossing multiple units, only the length within that specific unit is counted) and the fault throw, then substitute these values into the above formula for calculation.
Sum all individual fault intensity values calculated in step (2) to obtain the total fault intensity for that unit.
2.1.3 Fracture fractal dimension
Commonly used indicators for evaluating the complexity of fault structures include fault density, fault length, and fault intensity. However, in many cases, these indicators cannot accurately reflect the actual complexity of fault structures. When the number of faults is identical but their lengths differ, or when lengths are identical but the number of faults differs, the complexity can be intuitively perceived as different; yet this distinction cannot be captured by any single indicator alone. Fractal dimension can overcome this limitation, as different configurations yield different calculated fractal dimension values. Therefore, fractal dimension effectively reflects the complexity of fault structures, aligns with intuitive perception, and quantifies this perception, making it an effective and accurate indicator.
In this study, the grid method was employed to calculate the fracture fractal dimension of the mining area. The principle of the grid method involves covering the fault base map of the study area with square grids of varying densities (side lengths). As the grid density and side length change, the number of grids N(r) covering the fault base map changes correspondingly. The self-similarity relationship is as expressed in Equation 2.where D is the fractal dimension, C is the proportionality constant, and r is the side length.
When the side lengths of the square grids are r1, r2, r3, …, rk, the corresponding numbers of grids covering the fault system are N(r1), N(r2), N(r3), …, N(rk). Taking the double logarithm of both sides yields Equation 3.where A is an undetermined constant.
The slope value D obtained by fitting the straight line using the least squares method represents the fracture fractal dimension of the square grid region. Simultaneously, the correlation coefficient R can be obtained. By analogy, the fracture fractal dimension of all grids in the mining area can be determined.
Given the relatively complex fault system in Guobei Coal Mine, with hundreds of faults documented in the statistics, this study implemented an automated statistical function for fracture fractal dimension through the AutoCAD platform by programming in AutoLISP language based on the principles of fractal dimension measurement. The statistical data (grouped in sets of four, corresponding to the grid numbers N(r) at scales of r = 1/2, 1/4, 1/8, and 1/16, respectively) were imported into Excel spreadsheets for regression calculation to obtain the optimal regression line slope, i.e., the fractal dimension Ds, and the correlation coefficient R was simultaneously determined, thereby enabling quantitative evaluation of the complexity of the fracture network in the study area. The specific operational procedures are as follows:
On the coal seam mining plan, close all layers except the upper or lower wall structural trace of faults. Based on the actual geographic coordinate length, divide Guobei Coal Mine into 173 square grid units using r0 × r0 (r0 = 800 m, flexibly adjusted according to the size of the mining area) as the square grid (Figure 1), and obtain the number of grids containing faults, N(r).
Set the similarity ratios as r1 = 400 m, r2 = 200 m, r3 = 100 m, and r4 = 50 m, and move at a step distance of 500 m until the grids of the entire study area have been counted (Figure 2). The corresponding N(r) values are obtained successively, as shown in Figure 2a. In the ln r—ln N(r) coordinate system, with ln r as the abscissa and ln N(r) as the ordinate, four points (ln r, ln N(r)) are formed.
In Excel, fit the straight line L using the least squares method (Figure 2b); the slope of L is the information dimension D of the unit sought. Through the above process, the dimension of each measurement unit is obtained, and the center value of each square unit is defined as the fractal dimension value of that unit.
FIGURE 2
2.1.4 Fold planar deformation coefficient
The fold planar deformation coefficient can quantitatively represent the deformation degree of folds under tectonic stress and effectively reflect the development characteristics of fold structures. In this study, the fold planar deformation coefficient was selected for quantitative evaluation of fold development degree. The statistical and calculation procedures are as follows: first, identify two contour lines near the unit point that are closest to it, measure the horizontal distance (L0) between adjacent contour lines, and measure the secant length (L2) and actual length (L1) of the contour line closest to the unit point. The fold planar deformation coefficient of the unit point is then calculated using Equation 4 below.where Kp is the fold planar deformation coefficient; h is the elevation difference between two adjacent contour lines, in m; L0 is the horizontal distance between two contour lines at the center of the calculation unit, in m; L1 is the actual length of the contour line closest to the center of the calculation unit within the unit, in m; L2 is the secant length of the contour line closest to the center of the calculation unit within the unit, in m [].
2.2 Method for determining comprehensive weights of main controlling factors
2.2.1 Method for determining objective weights of main controlling factors
Since the information among the selected main controlling factors affecting mine structural complexity contains overlapping components, directly employing these data for evaluating mine structural complexity would result in information redundancy and thereby reduce the accuracy of evaluation results. Therefore, it is necessary to conduct independence weight analysis on the data of each selected main controlling factor. The Independence Weight Coefficient Method is an approach for determining objective weights, applicable to studies where complex and intricate relationships exist among evaluation indicators and where repeated information among indicators needs to be eliminated [].
Existing literature (e.g., []) primarily applies the AHP-Independence Weight Coefficient Method coupling to aquifer water-richness evaluation, where the degree of information overlap among indicators is relatively low, and the “redundancy elimination” advantage of the Independence Weight Coefficient Method is not fully demonstrated. This paper addresses the core challenge of mine structural complexity evaluation—significant information overlap among indicators such as fault density, fracture fractal dimension, and fault intensity.
The core improvement of this paper in weight fusion strategy lies in: effectively identifying and reducing the weights of highly correlated indicators (fault density and fracture fractal dimension) through the Independence Weight Coefficient Method, while preserving expert cognitive preferences for fault intensity via AHP, ultimately achieving organic fusion of “data-driven redundancy elimination + experience-driven key factor preservation” through multiplicative synthesis normalization. This improvement enables the comprehensive weight of fault intensity to accurately reflect its dominant controlling effect on structural complexity, avoiding weight allocation bias caused by single methods.
Assuming there are m evaluation indicators X1, X2, X3, …, Xm, if the multiple correlation coefficient between a certain evaluation indicator Xⱼ and other evaluation indicators is larger, it indicates that the collinearity relationship between Xⱼ and other indicators is stronger, the repeated information is more abundant, and consequently the weight of this indicator is smaller. The multiple correlation coefficient Ri of each index is calculated using Equation 5.where is the residual matrix of with X j removed; = mean(X)。
Considering the inverse proportional relationship between R and the weight coefficient, the reciprocal of the multiple correlation coefficient is selected as the scoring basis. The multiple correlation coefficient R is calculated using Equation 6.
After normalization processing, the final weight Dᵢ is as expressed in Equation 7.
2.2.2 Method for determining subjective weights of main controlling factors
The Analytic Hierarchy Process (AHP) is a method for determining subjective weights. By employing this approach, the empirical judgments of decision-makers can be quantified, thereby achieving optimal decision-making. The fundamental principle of AHP involves decomposing a complex problem influenced by multiple factors into a hierarchical structure according to the subordinate relationships among the factors, forming an ordered, multi-level hierarchy. Professional judgment is then applied from top to bottom to conduct pairwise comparisons of factors within each subordinate relationship, determining the relative importance among factors and ultimately yielding a comprehensive weight ranking. AHP is capable of addressing numerous geological problems that cannot be tackled by traditional optimization techniques.
The application of AHP for decision-making generally comprises the following four steps:
Establishment of the hierarchical structure
First, based on the analysis of the research objective, several major aspects influencing the objective are identified. Subsequently, subordinate influencing factors affecting each aspect are determined. By analogy, factors that mutually influence and interact with each other are decomposed level by level, thereby forming a top-down hierarchical structure. When determining the factors influencing the research objective, the principles of highlighting main indicators, ensuring relative independence among main factors, and guaranteeing operational feasibility of each factor must be followed. According to the decision-making events, considered factors, and interrelationships among decision objects, the hierarchical structure is divided into the goal layer (Level A), criterion layer (Level B), and decision layer (Level C). Through decision-making on problems at the decision layer, the objective of solving the goal layer (Level A) is achieved. The hierarchical structure diagram is ultimately constructed, as shown in Figure 3.
2. Construction of the judgment matrix
FIGURE 3
The hierarchical structure model reflects the relationships among various influencing factors. Within the criterion layer, different factors exert varying degrees of influence on the goal layer, and in the mind of the decision-maker, they occupy certain proportions. Assuming that the influence magnitudes of n factors on the goal layer X (x1, x2, x3, … xn) need to be determined, pairwise comparisons can be conducted to establish a comparison matrix. That is, two factors xi and xj are selected for comparison each time to establish the comparison matrix K, as shown in Equation 8. Let aij represent the ratio of the influence magnitudes of xi and xj on the goal layer. The comparison results of all factors are K= (aij)n×n, where K is the judgment matrix between A and X.
If the ratio of the influence of xi and xj on the goal layer A is aij, then the ratio of the influence of xj and xi on the goal layer is aji = 1/aij. The values of aij are determined using the numbers 1–9 and their reciprocals as scales according to the suggestion by Saaty et al., as shown in Table 2.
3. Calculation of the judgment matrix
TABLE 2
| Scale | Meaning |
|---|---|
| 1 | The two factors are of equal importance |
| 3 | The former factor is slightly more important than the latter |
| 5 | The former factor is obviously more important than the latter |
| 7 | The former factor is strongly more important than the latter |
| 9 | The former factor is extremely more important than the latter |
| 2, 4, 6, 8 | Intermediate values between adjacent judgments |
| Reciprocal | If the importance ratio of factor i to factor j is aij, then the importance ratio of factor j to factor i is aji = 1/aij |
Scale and meaning of the judgment matrix.
The judgment matrix is calculated to obtain the weight Ci of each indicator, and the maximum eigenvalue λmax of the above matrix and its corresponding eigenvector are calculated.
4. Consistency test
According to the scales used for constructing the judgment matrix, all judgment matrices are positive reciprocal matrices. Whether the constructed positive reciprocal matrix is logically reasonable needs to be verified through consistency testing. Only judgment matrices that pass the consistency test can yield acceptable weight values.
The steps for consistency testing are as follows:
Step 1: Calculate C.I. (Equation 9).
Step 2: Determine the corresponding R.I. from the table (Table 3).
TABLE 3
| n | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| R.I. | 0.00 | 0.00 | 0.58 | 0.90 | 1.12 | 1.24 | 1.32 | 1.41 | 1.45 | 1.49 | 1.50 |
Standard values of average random consistency index R.I.
Step 3: Calculate C.R. (Equation 10).
If C.R. < 0.1, the consistency of the judgment matrix is considered acceptable; if C.R. > 0.1, it does not meet the requirement and needs to be revised and re-evaluated.
2.2.3 Method for determining comprehensive weights
Given that subjective and objective weights are derived from different weighting bases, and with the aim of integrating the advantages of both, it is necessary to combine subjective and objective weights through integrated weighting. When the number of evaluation indicators is relatively large and weight distribution is relatively uniform, the multiplicative synthesis normalization method is typically employed to organically integrate subjective and objective weights, thereby calculating the comprehensive weight Wᵢ of each evaluation indicator. The calculation formula is shown in Equation 11.where n is the number of evaluation indicators; Di is the objective weight of the ith evaluation indicator; Ci is the subjective weight of the ith evaluation indicator.
2.2.4 Method for constructing quantitative evaluation model of mine structural complexity
2.2.4.1 Data normalization processing
To eliminate the influence of data with different dimensions among the main controlling factors affecting structural complexity on the evaluation results, it is necessary to normalize the data of each main controlling factor. The purpose of normalization is to eliminate dimensions, convert different categories of data into the same category, and render the data comparable and statistically meaningful for systematic analysis. For example, the unit of fault density is faults/m2, whereas fracture fractal dimension is dimensionless. Numerous normalization methods exist, including formulas for positive and negative correlations. Since all influencing factors in this study have positive correlations with structural complexity, the positive correlation normalization formula (min-max standardization) is adopted for unified processing, as shown in Equation 12.where xᵢ* is the normalized value of the evaluation indicator data; xi is the original value of the evaluation indicator data; xmin is the minimum value of a certain evaluation indicator dataset; xma is the maximum value of a certain evaluation indicator dataset.
After normalization of the original data of each main controlling factor, the normalized data are imported into ArcGIS to establish normalized thematic maps.
2.2.4.2 Composite overlay of normalized thematic maps and construction of quantitative evaluation model for mine structural complexity
Based on the comprehensive weight values calculated by AHP and the Independence Weight Coefficient Method assigned to each main controlling factor, and following normalization of the original data of each main controlling factor, the normalized data are imported into ArcGIS to establish normalized thematic maps for the four indicators: fault density, fault intensity, fracture fractal dimension, and fold planar deformation coefficient. ArcGIS is then applied to conduct overlay calculations of the thematic maps of each main controlling factor, yielding an evaluation model that comprehensively reflects mine structural complexity, as shown in Equation 13. This model reflects the structural complexity at a specific location within a certain area resulting from the combined effects of all main controlling factors. The magnitude of the structural complexity evaluation index (Wt) at each location characterizes the structural complexity at that point, with larger evaluation index values indicating greater structural complexity.where Wt is the structural complexity evaluation index; i is the ith main controlling factor; N is the number of main controlling factors, taken as 4 in this study; Wi is the comprehensive weight value of the ith main controlling factor; Ai(x, y) is the normalized value of the ith main controlling factor, with (x, y) being the geographic coordinates.
Accordingly, the structural complexity evaluation model for Guobei Coal Mine is obtained as shown in Equation 14.
The technical workflow of this study is illustrated in Figure 4.
FIGURE 4
3 Results and analysis
3.1 Single-indicator analysis of mine structural complexity
3.1.1 Characteristic analysis of fault density
As shown in Figure 5, the fault density of the mine predominantly ranges from 12 to 60 faults/km2 A southeast-trending belt exists in the central part of the mine, where fault density is lower than that in the southern and northern parts, ranging from 0 to 36 faults/km2. In the central part of the Mining Area 82 in the northern mine, influenced by faults F5, F6, F15, F15-1, and surrounding minor faults, the maximum fault density exceeds 96 faults/km2. A region with relatively high fault density values also exists in the outer section of the Mining Area 81 in the northern mine, primarily affected by faults F50, F8, and the development of surrounding minor faults, with the maximum fault density exceeding 108 faults/km2. The overall fault density in the Mining Areas 84 and 85 is relatively small, mainly distributed between 24 and 72 faults/km2. Only in the southern part of the Mining Area 84, influenced by the superposition of faults F22, BF28, and surrounding minor faults, does a local area with values higher than 60 faults/km2 exist.
FIGURE 5
3.1.2 Characteristic analysis of fault intensity
Based on the method described in Section 2.1.2, the contour map of fault intensity for Guobei Coal Mine was generated, as shown in Figure 6.
FIGURE 6
Fault intensity comprehensively considers factors such as throw and extension length. As illustrated in Figure 6, several regions with relatively high fault intensity values are associated with faults of substantial extension length and throw, and are further influenced by the superposition of surrounding minor faults. In the middle-lower part of the Mining Area 82, the combined effects of faults F15, F15-1, F22-1, and F22 result in relatively high fault intensity values in this region, primarily distributed in the range of 0.1–0.35. The fault intensity values in the Mining Areas 84 and 85 are generally low, mainly distributed in the range of 0.05–0.15. However, an east-west trending belt exists in the central part of the mine; although the fault density in this area is relatively low, the fault intensity values are relatively high due to the influence of throws, primarily affected by fault zones F26, F26-1, F34, F25, F22, and SF023, with fault intensity mainly distributed between 0.2 and 0.45. In the Mining Area 81 located in the southern part of the mine, a region with relatively high fault intensity values also exists, influenced by faults F22, F22-1, F1, F3, F3-1, F18, F8, and DF2, with fault intensity values primarily distributed in the range of 0.1–0.25.
3.1.3 Characteristic analysis of fracture fractal dimension
Based on the method described in Section 3.3.3, the contour map of fracture fractal dimension for Guobei Coal Mine was generated, as shown in Figure 7.
FIGURE 7
The fracture fractal dimension exhibits a favorable correspondence with fault density. As shown in Figure 7, the overall distribution trend of fracture fractal dimension values in the mine is characterized by low values in the central part and high values in the southern and northern parts. Except for local areas within the mine, the overall fracture fractal dimension values are greater than 1.3. The fracture fractal dimension values in the Mining Areas 84 and 85 show relatively small variation, ranging from 1.4 to 1.9. The regions with relatively high fracture fractal dimension values in Guobei Coal Mine are mainly concentrated near faults F15 and F15-1 in the Mining Area 82 and near faults F8 and F50 in the Mining Area 81, with maximum values exceeding 1.9.
3.1.4 Characteristic analysis of fold planar deformation coefficient
The planar deformation coefficients of all unit points were calculated according to Equation 4, and the contour map of fold planar deformation coefficient within the mine area was generated, as shown in Figure 8.
FIGURE 8
As can be seen from Figure 8, the fold planar deformation coefficient values in the mine range from 0.2 to 0.8. A north-south trending belt exists in the central part of the mine, with fold planar deformation coefficient values between 0.26 and 0.38. Taking this belt as the boundary, the values show an increasing trend toward the western and eastern parts of the mine. A north-south trending belt also exists along the eastern boundary of the mine, with fold planar deformation coefficients between 0.56 and 0.62. In the southwestern part of the Mining Area 81, the fold planar deformation coefficient is relatively high, approximately between 0.62 and 0.8, where folds are relatively well developed. Folds in the Mining Area 85 are more developed than those in the Mining Area 84; the fold planar deformation coefficient in the Mining Area 85 is mainly distributed between 0.38 and 0.62, whereas that in the Mining Area 84 is mainly distributed between 0.26 and 0.44. The analysis demonstrates that a favorable correspondence exists between the fold planar deformation coefficient and the degree of fold development. The contour map of planar deformation coefficient can intuitively reflect the development degree of folds within the mining field.
3.2 Determination of comprehensive weights of main controlling factors
3.2.1 Determination of objective weights
Based on the principles of the Independence Weight Coefficient Method described above, the objective weights of the four main controlling factors affecting mine structural complexity were calculated, as shown in Table 4.
TABLE 4
| Main controlling factor | Multiple correlation coefficient R | Reciprocal of multiple correlation coefficient 1/R | Objective weight (%) |
|---|---|---|---|
| Fault density | 0.755 | 1.325 | 11.693 |
| Fault intensity | 0.203 | 4.928 | 43.491 |
| Fracture fractal dimension | 0.752 | 1.33 | 11.736 |
| Fold planar deformation coefficient | 0.267 | 3.748 | 33.080 |
Objective weights determined by the Independence Weight Coefficient Method.
3.2.2 Determination of subjective weights
3.2.2.1 Establishment of the hierarchical structure model
Through analysis of the main controlling factors affecting mine structural complexity, a hierarchical structure model was established with quantitative evaluation of mine structural complexity as the goal layer (Layer A); fault development characteristics and fold development characteristics as the criterion layer (Layer B); and fault density, fault intensity, fracture fractal dimension, and fold planar deformation coefficient constituting the decision layer (Layer C) of this structural model. Through decision-making on problems at this hierarchical level, the ultimate objective of evaluating the structural complexity of Guobei Coal Mine is achieved, that is, determining the influence weights of each main controlling factor in the decision layer (Layer C) on the evaluation of mine structural complexity (Layer A). The constructed hierarchical structure model is shown in Figure 9.
FIGURE 9
3.2.2.2 Construction of AHP judgment matrix and consistency test
Based on the roles played by each main controlling factor in the evaluation of structural complexity of Guobei Coal Mine, hierarchical judgment matrices were established for the main controlling factors. Using the method of “expert scoring solicitation,” quantitative scores between various factors were obtained.
The AHP judgment matrix in this study was constructed using the “expert scoring solicitation method,” with 5 experts possessing extensive experience in coal mine geology and hydrogeology invited to participate in scoring. The expert composition is as follows Table 5.
TABLE 5
| Expert ID | Title/Position | Specialization | Years of experience |
|---|---|---|---|
| E1 | Professor-level senior engineer | Mine geology | 28 years |
| E2 | Research fellow | Coalfield structural geology | 22 years |
| E3 | Senior engineer | Mine hydrogeology | 25 years |
| E4 | Professor | Mining engineering | 20 years |
| E5 | Senior engineer | Geological exploration | 18 years |
Expert information form.
Each expert conducted independent pairwise comparisons, and the comprehensive judgment matrix was constructed using geometric averaging. The consistency test result C.R. = 0.0077 < 0.1 satisfies the consistency requirement.
However, we also note the discrepancy between expert subjective preferences and objective data patterns: the objective weight calculated by the Independence Weight Coefficient Method shows fault intensity (43.491%) higher than fracture fractal dimension (11.736%), which precisely reflects the difference that expert experience emphasizes “morphological complexity” while data patterns highlight “scale effects.”
According to the final cumulative scores, total score comparisons between factors were conducted, forming the expert judgment set for each influencing factor. Accordingly, the judgment matrix among the main controlling factors (fault density, fault intensity, fracture fractal dimension, and fold planar deformation coefficient) for structural complexity evaluation of Guobei Coal Mine was established (Table 6).
TABLE 6
| Main controlling factor | Fault density | Fault intensity | Fracture fractal dimension | Fold planar deformation coefficient | Wᵢ |
|---|---|---|---|---|---|
| Fault density | 1 | 1/3 | 1/6 | 1/3 | 0.0749 |
| Fault intensity | 3 | 1 | 1/2 | 1 | 0.2247 |
| Fracture fractal dimension | 6 | 2 | 1 | 3 | 0.4974 |
| Fold planar deformation coefficient | 3 | 1 | 1/3 | 1 | 0.2030 |
Judgment matrix of main controlling factors for structural complexity evaluation.
After establishing the above judgment matrix, consistency testing is also required. As shown in Table 7, the weight values obtained after passing the consistency test can be considered acceptable.
TABLE 7
| Criterion layer | λmax | C.I. | R.I. | C.R. |
|---|---|---|---|---|
| B ∼ Ci (i = 1∼4) | 4.02060 | 0.00687 | 0.90 | 0.00770 |
Consistency test results of the judgment matrix.
As can be seen from the above table, the consistency ratio C.R. of the judgment matrix among the main controlling factors for structural complexity evaluation is less than 0.1, satisfying the consistency requirement.
3.2.2.3 Determination of subjective weights
The weights of the main controlling factors affecting mine structural complexity obtained through the above judgment matrix are shown in Table 8.
TABLE 8
| Main controlling factor | Fault density | Fault intensity | Fracture fractal dimension | Fold planar deformation coefficient |
|---|---|---|---|---|
| Subjective weight | 0.0749 | 0.2247 | 0.4974 | 0.2030 |
Subjective weights of each main controlling factor.
3.2.3 Construction of quantitative evaluation model for mine structural complexity based on comprehensive weighting
The objective and subjective weights of each evaluation indicator determined by the Independence Weight Coefficient Method and AHP, respectively, were substituted into the multiplicative synthesis normalization formula (Equation 11) to calculate the comprehensive weights of the four main controlling factors affecting the structural complexity of Guobei Coal Mine, as shown in Table 9.
TABLE 9
| Main controlling factor | Fault density | Fault intensity | Fracture fractal dimension | Fold planar deformation coefficient |
|---|---|---|---|---|
| Comprehensive weight | 0.0378 | 0.4212 | 0.2516 | 0.2894 |
Comprehensive weights of main controlling factors affecting mine structural complexity.
This paper achieves organic fusion of both through multiplicative synthesis normalization, with the final comprehensive weights ranking as: fault intensity (42.12%) > fold planar deformation coefficient (28.94%) > fracture fractal dimension (25.16%), respecting objective data patterns while preserving reasonable components of expert experience.
As can be seen from Table 9, fault intensity exerts the greatest influence on the structural complexity of Guobei Coal Mine, followed by fold planar deformation coefficient and fracture fractal dimension. Fault density only provides an intuitive reflection of structural development from the planar spatial perspective, without incorporating relevant parameters for the vertical spatial dimension of faults; therefore, its influence on structural complexity is relatively minor.
Substituting the comprehensive weights of each main controlling factor into Equation 14, the structural complexity evaluation model for Guobei Coal Mine is derived as Equation 15.
3.3 Study on mine structural development characteristics and structure-controlled water regularity
3.3.1 Evaluation of mine structural complexity
Based on the structural complexity evaluation model (Equation 15) for Guobei Coal Mine, statistical analysis was conducted on the structural complexity evaluation index of each independent unit after overlay. The natural breaks (Jenks) classification method was employed for four-level classification, and the comprehensive threshold values were determined as 0.3732, 0.4208, and 0.4718, respectively. A higher structural complexity evaluation index Wt indicates greater structural complexity. According to the classification thresholds, the mining area was divided into four zones: structurally complex zone, relatively complex zone, moderately complex zone, and simple zone, as shown in Table 10. The structural complexity zoning map of Guobei Coal Mine was generated, as shown in Figure 10.
TABLE 10
| Structural complexity evaluation index (Wt) | Structural complexity zone |
|---|---|
| Wt≥0.4718 | Structurally complex zone |
| 0.4718>Wt≥0.4208 | Relatively complex zone |
| 0.4208>Wt≥0.3732 | Moderately complex zone |
| Wt<0.3732 | Simple zone |
Zoning of structural complexity of No. 82 Coal Seam roof.
FIGURE 10
As shown in Figure 10, the structurally complex zones and relatively complex zones are distributed in three main areas within the mine. The first area is located in the northeastern part of the mine, radiating from the vicinities of faults F15, F15-1, and F22-1 toward the surrounding areas, with the structurally complex and relatively complex zones accounting for 4.4% and 6.7% of the total mine area, respectively. The second area is located in the central part of the mine, radiating from the vicinities of faults F26 and F34 toward the surrounding areas, with the structurally complex and relatively complex zones accounting for 5.7% and 4.5% of the total mine area, respectively. The third area is located in the southern part of the mine; the structurally complex–relatively complex zone in the southeastern part radiates from the vicinities of faults F22 and F48 toward the surrounding areas, while that in the southwestern part radiates from the vicinities of faults DF2, F8, and F3-1 toward the surrounding areas. The structurally complex and relatively complex zones in the southern part of the mine account for 5.7% and 13.2% of the total mine area, respectively. In the Mining Areas 84 and 85, the structurally complex and relatively complex zones account for 21.1% and 41.0% of their respective district areas.
3.3.2 Study on mine structure-controlled water regularity
According to the historical water inrush records of Guobei Coal Mine, a total of 14 groundwater inrush incidents occurred from December 2005 to December 2022 (excluding water inrushes in main and ventilation shafts). The locations of groundwater inrush points are shown in Figure 11. There were 10 water inrush (outflow) events with flow rates greater than or equal to 10 m3/h, including 9 sandstone fracture water inrushes (outflows) from the roof and floor of the main mining coal seam, with a maximum water inflow rate of 40 m3/h. Under mining-induced influence, a total of 5 groundwater inrush incidents were caused by roof fracture zones connected through faults that conducted roof sandstone fracture water, with a maximum water inflow rate of 15 m3/h. The large faults exposed during mining were all non-water-bearing with poor water conductivity, and no water inrush accidents caused by faults conducting floor limestone water occurred. As can be seen from the figure, most mine water inrush (outflow) events are strongly associated with geological structures; groundwater inrush points are mainly distributed in or near the structurally complex–relatively complex zones and at intersections of major mine faults and multiple faults. High degrees of fault development lead to well-developed fractures in the roof overlying rock of the coal seam, providing favorable water-conducting pathways and water storage spaces. After coal seam mining disturbance, the development height of the roof “two zones” (caved zone and fractured zone) is higher than that in intact rock strata sections, making it prone to affect overlying aquifers and trigger water inrush. Combined with the structural complexity zoning evaluation results of the mining area, particular attention should be paid to water inrush threats from roof fracture zones and fault intersections in or near the structurally complex–relatively complex zones of the Mining Areas 84 and 85.
FIGURE 11
4 Discussion
4.1 Scientific validity of the evaluation indicator system
The indicator system constructed in this study encompasses two dimensions—faulting and folding—with a multi-level complementary characteristic of “quantity–scale–morphology.” The comprehensive weights reveal that fault intensity (0.4212) is significantly higher than fault density (0.0378), which is consistent with the calculation results of the Independence Weight Coefficient Method. The high information overlap between fault density and fracture fractal dimension (multiple correlation coefficients of 0.755 and 0.752, respectively) led to a substantial reduction in their objective weights, indicating that fault scale parameters are more indicative than simple quantity statistics. The introduction of the fold planar deformation coefficient fills the methodological gap in quantitative evaluation of fold structures, accounting for 0.2894 in the comprehensive weights and forming favorable spatial complementarity with fault indicators. The limitation lies in the fact that factors such as fault attitude, nature, and filling characteristics have not been considered; future work could introduce parameters such as fault dip angle correction coefficient for further refinement.
4.2 Rationality of the combined weighting method
The multiple correlation coefficient of fault intensity R = 0.203 is far lower than that of fault density (0.755), fracture fractal dimension (0.752), and fold planar deformation coefficient (0.267). This feature has important methodological significance:
Information Complementarity: Fault intensity has low information overlap with fault density and fracture fractal dimension, meaning it provides unique structural information—the scale effect of faults, rather than simple quantity or morphological statistics.
Rationality of Weight Configuration: The Independence Weight Coefficient Method, based on the principle of “redundancy elimination,” assigns higher weights to more independent indicators. The high objective weight (43.49%) and high comprehensive weight (42.12%) of fault intensity are mathematical reflections of its strong independence.
Explanation of Geological Essence: The independence of fault intensity stems from the uniqueness of its physical connotation—it integrates two orthogonal dimensions of information: fault length (planar extension) and throw (vertical cutting), whereas fault density only reflects planar quantity and fracture fractal dimension only reflects planar morphological complexity, neither involving the vertical dimension. Therefore, fault intensity occupies a unique position in the information space.
Implications for Methodology: This result validates the rationality of the coupling method in this paper—if only AHP were used, expert preferences might overemphasize fracture fractal dimension; if only the Independence Weight Coefficient Method were used, fault intensity might be overemphasized while fold factors are neglected. Multiplicative synthesis normalization achieves a balance between the two.
The combined weighting of AHP and the Independence Weight Coefficient Method effectively balances subjective experience and objective regularities. A comparison of the results from the two methods shows that expert preference assigns the highest subjective weight to fracture fractal dimension (49.74%), whereas the inherent data regularity gives fault intensity the highest objective weight (43.49%). The comprehensive weights are dominated by data regularities, with fault intensity becoming the leading factor (42.12%), thereby avoiding the bias of a single method. The consistency of the AHP judgment matrix is satisfactory (C.R. = 0.00770); however, the limited sample size of experts suggests that group decision-making should be introduced in subsequent studies to improve the reliability of subjective weights.
4.3 Hydrogeological significance and accident cause analysis of zoning results
From December 2005 to December 2022, 14 groundwater inrush events occurred in Guobei Coal Mine (
Figure 11). The direct causes of these accidents can be categorized as:
Roof fracture zone water conduction (9 events): In structurally complex zones (e.g., near faults F15, F22), intense fault development created well-developed fractures in the roof overlying rock. Under mining-induced disturbance, the height of the caved zone and fractured zone significantly exceeded that in intact rock strata, connecting overlying sandstone aquifers and triggering inrush with a maximum flow rate of 40 m3/h.
Fault intersection conduction (5 events): At intersections of major faults (e.g., F22-F48, F26-F34), fault zones formed preferential water-conducting pathways. Although large faults themselves were non-water-bearing, mining-induced reactivation of fault zones connected roof fracture networks, conducting sandstone water with a maximum flow rate of 15 m3/h.
The indirect causes include:
High structural complexity: The spatial coincidence between inrush points and structurally complex/relatively complex zones (Figure 11) confirms that structural complexity is the fundamental geological precondition.
Mining-induced stress redistribution: Coal extraction in structurally complex areas exacerbated stress concentration and roof failure, increasing the probability of aquifer connectivity.
Insufficient targeted prevention: Prior to this zoning study, differentiated prevention strategies based on structural complexity were lacking, leading to inadequate water hazard control in high-risk zones.
The structural complexity zoning exhibits a significant spatial correspondence with the distribution of groundwater inrush points. The structurally complex–relatively complex zones and their transitional belts are characterized by high degrees of fault development and well-developed fractures in the roof overlying rock. Under mining-induced influence, the height of the roof “two zones” (caved zone and fractured zone) is significantly greater than that in intact rock strata sections, making it prone to affect overlying aquifers and trigger water inrush. The Mining Area 84 is generally structurally simple, where conventional groundwater hazard prevention and control measures can be adopted; the Mining Area 85 contains local structurally complex zones and should be prioritized for water hazard prevention and control [].
This zoning result enables the transformation from passive, uniform prevention to active, differentiated control. By aligning prevention intensity with structural complexity levels, both safety and economic efficiency can be improved.
4.4 Research limitations and prospects
Limitations in three-dimensional extension. The evaluation is based on two-dimensional planar data, and factors such as fault vertical cutting depth and fold profile morphology have not yet been incorporated. This can be improved in the next step based on three-dimensional geological modeling techniques.
Insufficient integration with water inrush prediction. Although the spatial correspondence between structurally complex zones and groundwater inrush points has been revealed, a quantitative statistical model linking the structural complexity index with water inrush probability has not yet been established. In the future, machine learning methods can be employed to achieve the transition from “status evaluation” to “risk prediction.”
While this study focuses on structural factors, mine water inrush is governed by multiple factors including aquifer water richness, mining-induced stress fields, and floor failure characteristics. Future research will integrate these factors with structural complexity indices to build a multi-factor coupling prediction model, potentially employing machine learning methods to achieve quantitative risk assessment [].
5 Conclusion
Based on multi-source information fusion theory, this study selected four evaluation indicators—fault density, fault intensity, fracture fractal dimension, and fold planar deformation coefficient—as the indicators affecting mine structural complexity. ArcGIS was applied for quantitative analysis of each evaluation indicator, and AHP and the Independence Weight Coefficient Method were employed to obtain the subjective and objective weights of each evaluation indicator, respectively, which were then synthesized and normalized. A quantitative evaluation model for mine structural complexity with comprehensive weighting was established, and comprehensive zoning of mine structural complexity in the study area was conducted. The specific conclusions are as follows:
The fault density of the mine predominantly ranges from 12 to 60 faults/km2. The overall fault density in the Mining Areas 84 and 85 is relatively small, mainly distributed between 24 and 72 faults/km2. Only in the southern part of the Mining Area 84, influenced by the superposition of faults F22, BF28, and surrounding minor faults, does a local area with values higher than 60 faults/km2 exist.
Several regions with relatively high fault intensity values are associated with faults of substantial extension length and throw, and are further influenced by the superposition of surrounding minor faults. The fault intensity values in the Mining Areas 84 and 85 are generally low, mainly distributed in the range of 0.05–0.15. However, an east-west trending belt exists in the central part of the mine; although the fault density in this area is relatively low, the fault intensity values are relatively high due to the influence of fault zones F26, F26-1, F34, F25, F22, and SF023, with fault intensity mainly distributed between 0.2 and 0.45.
The overall distribution trend of fracture fractal dimension values in the mine is characterized by low values in the central part and high values in the southern and northern parts. Except for local areas within the mine, the overall fracture fractal dimension values are greater than 1.3. The fracture fractal dimension values in the Mining Areas 84 and 85 show relatively small variation, ranging from 1.4 to 1.9. The regions with relatively high fracture fractal dimension values in Guobei Coal Mine are mainly concentrated near faults F15 and F15-1 in the Mining Area 82 and near faults F8 and F50 in the Mining Area 81, with maximum values exceeding 1.9.
Based on the comprehensive weight values calculated by AHP and the Independence Weight Coefficient Method assigned to each main controlling factor, ArcGIS was applied to conduct overlay calculations of the thematic maps of each main controlling factor, establishing an evaluation model that comprehensively reflects mine structural complexity. The natural breaks (Jenks) classification method was employed for four-level classification, and the comprehensive threshold values were determined as 0.3732, 0.4208, and 0.4718.
The structurally complex zones and relatively complex zones are distributed in three main areas within the mine. The first area is located in the northeastern part of the mine, the second in the central part, and the third in the southern part. The structurally complex and relatively complex zones account for 15.8% and 24.3% of the total mine area, respectively.
Groundwater inrush points in the mining area are predominantly distributed within the structurally complex–relatively complex zones or in transitional areas between zones. This study provides targeted objectives for water hazard prevention and control in future mining areas of the mine, with particular attention to be paid to water inrush risks in the structurally complex–relatively complex zones, at major mine faults, and at intersections of multiple faults.
Targeted recommendations for water inrush prevention and control:
For structurally complex zones (Wt ≥ 0.4718): Implement enhanced geophysical exploration (transient electromagnetic or seismic methods) before mining; reinforce roof support; establish real-time water pressure monitoring; and avoid high-intensity mining near major fault intersections.
For relatively complex zones (0.4208 ≤ Wt < 0.4718): Conduct pre-drainage of roof sandstone aquifers; optimize working face layout to bypass or obliquely cross major faults; and install early-warning systems for water inrush.
For moderately complex and simple zones: Apply routine water hazard prevention measures, but maintain vigilance during crossing of small faults.
Specifically for Mining Area 85: Prioritize water hazard prevention and control due to its high proportion (41.0%) of relatively complex zones. For Mining Area 84, conventional measures are generally sufficient, but attention should be paid to local complex zones near fault F22.
Statements
Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.
Author contributions
BX: Conceptualization, Data curation, Funding acquisition, Writing – review and editing. HL: Methodology, Software, Writing – review and editing. GZ: Validation, Writing – review and editing. RH: Writing – original draft, Writing – review and editing. HG: Conceptualization, Funding acquisition, Supervision, Writing – review and editing. YG: Validation, Writing – review and editing.
Funding
The author(s) declared that financial support was received for this work and/or its publication. This work was supported by the Scientific Research Project of China National Administration of Coal Geology (2025-II-5), University Natural Science Research Project of Anhui Province (2025AHGXZK40586), Doctoral Research Start-up Fund of Suzhou University (2022BSK011), Postdoctoral Research Start-up Fund of Suzhou University (2024BSH002), Key Laboratory of Mine Water Resource Utilization of Anhui Higher Education Institutes, Suzhou University (KMWRU202403), Fund Project of National Engineering Research Center for Coal Mine Water Hazard Prevention and Control (WBMDGCZXJJ(2025)2), Innovation Team for Key Technologies of Co-extraction of Coal, Water and Heat (2024TDPY02), Excellent Scientific Research and Innovation Team of Anhui Province for Coal Mine Water Hazard Prevention and Mine Water Resource Utilization (2022AH010081), Key Discipline of Geological Resources and Geological Engineering (2017xjzdxk2).
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fphy.2026.1927525/full#supplementary-material
References
1.
ZhaiPHLiuYXLiuTHShiLLvC. Quantitative evaluation of structural complexity in shengquan minefield based on improved local linear embedding algorithm. Energy Exploration and Exploitation (2023) 41(4):1439–56. 10.1177/01445987231157124
2.
ZhangGZGuoJZXuBXuLDaiZYinSet alQuantitative analysis and evaluation of coal mine geological structures based on fractal theory. Energies (2021) 14(7):1–14. 10.3390/en14071925
3.
PuYYangZH. Quantitative evaluation of complex degree of geological structure in yangquan mining area. Sustainability (2022) 14(19):1–20. 10.3390/su141912028
4.
XiaoLLLiFNiuCDaiGQiaoQLinC. Evaluation ofwater inrush hazard in coal seam roof based on the AHP-CRITIC composite weighted method. Energies (2023) 16(1):1–20. 10.3390/en16010114
5.
DuanHJHaoSJFengJWangYPengD. A prediction method for water enrichment in aquifer based on GIS and coupled AHP-Entropy model. Open Geosciences (2021) 13(1):1318–27. 10.1515/geo-2020-0277
6.
LiMJiangBLanFJ. Quantitative characterization of fractals and curvatures in complex geological structures of wugou coal mine, huaibei coalfield. Fractal and Fractional (2025) 9(10):1–25. 10.3390/fractalfract9100669
7.
ZhangHWangYJLiuCZ. Research on evaluation of degree of complexity of mining fault network based on GIS. J China Univ Mining and Technology (2007) 17(01):63–7. 10.1016/S1006-1266(07)60014-5
8.
WangXDZhangSPanRHZhaoYJiaZTursinhaliHet alCharacterization of mining fracture network development in overlying rock layers of deep mining in ordos basin. Front Earth Sci (2025) 13:1461180. 10.3389/feart.2025.1461180
9.
YuanCFZhangZXZhangGD. Discussion and application of deformation coefficients of fold. J China Coal Soc (1992) 17(03):11–8. 10.13225/j.cnki.jccs.1992.03.002
10.
ZhouYXZhouYZXieSYDaiyongCXiangrongQ. A grey fuzzy comprehensive model for evaluation of geological structure complexity. J China Univ Geosciences (2008) 19(04):436–40. 10.1016/S1002-0705(08)60078-3
11.
ZhangYTangSFShiK. Risk assessment of coal mine water inrush based on PCA-DBN. Scientific Rep (2022) 12(1):1370. 10.1038/s41598-022-05473-8
12.
DursunAE. Fatal accident analysis and hazard identification in Turkish coal-extracting industry using analytic hierarchy process. Mining Metall and Exploration (2024) 41(1):149–72. 10.1007/s42461-023-00905-z
13.
WeiDYDuCFLinYFChangBWangY. Thermal environment assessment of deep mine based on analytic hierarchy process and fuzzy comprehensive evaluation. Case Stud Therm Eng (2020) 19:100618. 10.1016/j.csite.2020.100618
14.
ZhaoJGSongSZhangKLiXZhengXWangYet alAn investigation into the disturbance effects of coal mining on groundwater and surface ecosystems. Environ Geochem Health (2023) 45(10):7011–31. 10.1007/s10653-023-01658-w
15.
XiaRHWangHLHuTCYuanSHuangBWangJet alA risk assessment of water inrush in deep mining in metal mines based on the coupling methods of the analytic hierarchy process and entropy weight method: a case study of the huize lead-zinc mine in northeastern Yunnan, China. Water (2025) 17(5):1–16. 10.3390/w17050643
16.
LiuYEckertCMEarlC. A review of fuzzy AHP methods for decision-making with subjective judgements. Expert Syst Appl (2020) 161(1):67. 10.1016/j.eswa.2020.113738
17.
TangFQinYPGuoMYWangPZhangFHouH. Thermal risk assessment in the mine of TOPSIS based on game theory subjective and objective comprehensive weighting. Therm Sci Eng Prog (2025) 59:103367. 10.1016/j.tsep.2025.103367
18.
LiQSuiWH. Risk evaluation of mine-water inrush based on principal component logistic regression analysis and an improved analytic hierarchy process. Hydrogeology J (2021) 29(3):1299–311. 10.1007/s10040-021-02305-3
19.
DongDLLiXLinGBianJLCaoCLWuH.Identification model of the independence right-fuzzy variabletheory of water inrush source. Coal Geology and Exploration (2019) 47(05):48–53. 10.3969/j.issn.1001-1986.2019.05.007
20.
ZhangSYHuYBXingSP. Discrimination of the mine water inrush source based on principalcomponent analyses-theory of gray relational degree. HYDROGEOLOGY and ENGINEERING GEOLOGY (2018) 45(06):36–41. 10.16030/j.cnki.issn.1000-3665.2018.06.06
21.
QiuMShaoZDZhangWQZhengYYinXGaiGet alWater-richness evaluation method and application of clastic rock aquifer in mining seam roof. Scientific Rep (2024) 14(1):1–23. 10.1038/s41598-024-57033-x
22.
QiuMYinXYShiLQZhaiPGaiGShaoZ. Multifactor prediction of the water richness of coal roof aquifers based on the combination weighting method and TOPSIS model: a case study in the changcheng no. 1 coal mine. Acs Omega (2022) 7:44984–5003. 10.1021/acsomega.2c05297
23.
ZhengQSWangCFZhuZH. Research on the prediction of mine water inrush disasters based on multi-factor spatial game reconstruction. Geomechanics Geophys Geo-Energy Geo-Resources (2024) 10(1):1–20. 10.1007/s40948-024-00761-1
24.
YangBYuanJDuanLLiuQ. Using GIS and fractal theory to evaluate degree of fault complexity and water yield. Mine Water Environ (2019) 38(2):261–7. 10.1007/s10230-018-0563-8
25.
ShiFLiXDLiMRBaiB. Site selection strategy for photovoltaic power plants construction on gangue hills: an integrated method based on GIS and AHP-TOPSIS. Energy Strategy Rev (2025) 59(1-20):101722. 10.1016/j.esr.2025.101722
26.
WangYPuZGGeQLiuJ. Study on the water-richness law and zoning assessment of mine water-bearing aquifers based on sedimentary characteristics. Scientific Rep (2022) 12(1):14107. 10.1038/s41598-022-18403-5
27.
LiHHaoXJJiJYYangHLinWZhangSet alQuantitative early warning of floor water inrush based on deep learning: joint response of microseismic three-dimensional kernel density and hydrodynamics. Rock Mech Rock Eng (2025) 59:7403–23. 10.1007/s00603-025-05152-w
28.
JuQHuY. Source identification of mine water inrush based on principal component analysis and grey situation decision. Environ Earth Sci (2021) 80(4):157. 10.1007/s12665-021-09459-z
29.
TongXZhengXZJinYFDongBLiuQLiY. Prevention and control strategy of coal mine water inrush accident based on case-driven and bow-tie-bayesian model. Energy (2025) 320:135312. 10.1016/j.energy.2025.135312
30.
RenQLiuTWanDWangJJiangRLiMet alAn innovative approach to discrete facture network modeling driven by geomechanics and multiple factors. Geoenergy Sci Eng (2025) 257(PartA):214200. 10.1016/j.geoen.2025.214200
31.
XueJShiLWangHJiZShangHXuFet alWater abundance evaluation of a burnt rock aquifer using the AHP and entropy weight method: a case study in the yongxin coal mine, China. Environ Earth Sci (2021) 80(11):417. 10.1007/s12665-021-09703-6
32.
WuQFanZLLiuSQZhangYWSunWJ.Water-richness evaluation method of water-filled aquifer based on theprinciple of information fusion with GIS: water-richness index method. J China Coal Soc (2011) 36(07):1124–8. 10.13225/j.cnki.jccs.2011.07.002
33.
RudkeAPde SouzaVASDos SantosAMFreitas XavierACRotunno FilhoOCMartinsJA. Impact of mining activities on areas of environmental protection in the southwest of the amazon: a GIS- and remote sensing-based assessment. J Environ Manage (2020) 263:110392. 10.1016/j.jenvman.2020.110392
34.
NatarajanSMohanK. Integrated assessment of groundwater quality and recharge potential for sustainable water resource management using hydrogeochemical analysis, GIS and AHP techniques. Water Air Soil Pollut (2026) 237(7):404. 10.1007/s11270-025-08966-0
35.
WangXDZhangCBWangCLiuGWangH. GIS-based for prediction and prevention of environmental geological disaster susceptibility: from a perspective of sustainable development. Ecotoxicology Environ Saf (2021) 226:112881. 10.1016/j.ecoenv.2021.112881
36.
SinghADashDKumarRSarkarNK. Multithematic geoscience data integration for identification of mineral potential areas in parts of central indian precambrian craton. J Geol Soc India (2022) 98(10):1466–74. 10.1007/s12594-022-2194-8
37.
BiYSWuJWZhaiXRShenSHHuRJuOD.Evaluation of coal mine aquifer water-richness based on AHP andi ndependent weight coefficient method. J China Hydrol (2020) 40(04):40–5. 10.19797/j.cnki.1000-0852.20190175
38.
ZengYFWuQZhaoSQMiaoYWZhangYMeiASet alCharacteristics, causes, and prevention measures of coal mine water hazard accidents in China. Coal Sci Technology (2023) 51(7):1–14. 10.13199/j.cnki.cst.2023-0500
Summary
Keywords
Analytic Hierarchy Process (AHP), Guobei Coal Mine, independence weight coefficient method, structural complexity zoning, structure-controlled groundwater inrush regularity
Citation
Xie B, Lin H, Zhao G, Hu R, Gui H and Guo Y (2026) Quantitative evaluation of mine structural complexity based on GIS-AHP-independence weight coefficient method: a case study of Guobei Coal Mine. Front. Phys. 14:1927525. doi: 10.3389/fphy.2026.1927525
Received
03 July 2026
Revised
05 August 2026
Accepted
11 August 2026
Published
09 September 2026
Volume
14 - 2026
Edited by
Zhengzheng Cao, Henan Polytechnic University, China
Reviewed by
Yanchun Yin, Shandong University of Science and Technology, China
Arif Emre Dursun, Konya Technical University, Türkiye
Xiaorong Zhai, Anhui University of Science and Technology, China
Updates
Copyright
© 2026 Xie, Lin, Zhao, Hu, Gui and Guo.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Ru Hu, 1595036962@qq.com
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.