1 Introduction

Coal, a significant component of global energy production, continues to play a crucial role in the world economy [1,2,3,4]. In 2022, global coal demand reached its highest level to date [5]. While this trend persists, there is a need to enhance efficiency and undertake improvement efforts in coal mining to contribute to the overarching goal of energy sustainability. The rise in global coal demand is causing a depletion of open-pit reserves where coal production takes place, leading to a transition towards deeper underground mining operations [6]. This transition involves creating galleries for underground coal production, with a particular focus on faster and more dependable support systems, to reinforce sustainability and operational efficiency within the coal mining industry.

In recent decades, underground coal mines have adopted mechanized excavation systems, particularly employing the longwall mining technique, to facilitate high-volume coal production [7,8,9,10]. However, this implementation has resulted in the formation of numerous mine galleries. The traditional roof support system utilizing steel sets in these galleries can adversely impact their daily advancement rates, unit costs, and safety, especially in rock mass environments characterized by very weak, weak, and moderately strong strengths [11, 12].

Over the past four decades, there have been notable advancements in rock bolting support systems, driven by improved insights into load transfer mechanisms and the evolution of rock bolting technology [13,14,15,16].

Distinguished from traditional support systems like steel sets, this system facilitates faster advancement rates, reduced unit costs, and enhanced safety through its active support capabilities. In contrast to passive support systems such as steel sets, rock bolts engage more rapidly with the rock mass to initiate their supporting function, resulting in less deformation and facilitating safer and swifter gallery advancement [17]. Consequently, it has been successfully implemented as the primary supporting element in underground coal mines across various regions globally [18, 19].

In underground coal production activities, especially in thick coal seams, detailed preliminary design studies should be conducted for the proper reinforcement of gateroads with rock bolting. The empirical approaches, commonly used in project works today, are developed by combining observations, measurements, experience, engineering intuition, and judgments. These approaches can provide predictions for support parameters such as rock bolt length and spacing, dimensions and intervals of steel sets, and thickness of shotcrete for the stability of underground openings. Eleven different empirical approaches can provide design outputs for rock bolting support. It is known that six of these approaches are based on the widely used RMR rock mass classification system [20,21,22,23,24,25], two are based on the widely used Q rock mass classification system [26, 27], and two are based on the RQD system [28, 29]. Additionally, there are two different design approaches based on Panek [30] and the number of discontinuiy sets and the dip angle of the discontinuities.

Mathematical-based and assumption-dependent numerical design approaches have been developed to determine stress and deformation behaviors in the opening zones of underground and surface rock engineering excavations. In today’s design practices, two- and three-dimensional numerical analysis programs developed in many significant studies are utilized [31,32,33,34,35]. In the two- and three-dimensional models presented in numerical analyses, alongside rock material and rock mass parameters, input parameters for support, such as rock bolt length, spacing, and shotcrete thickness, need to be initially defined. The necessity to perform hundreds of numerical simulations to determine unknown optimal support parameters is considered a serious problem. Additionally, the presentation of field stresses and displacement quantities, which cannot always be measured, using numerical analyses necessitates the careful use of numerical approaches, especially in complex field conditions [12].

This study aims to analyze the feasibility of rock bolting in an underground coal mine gallery with a thick coal seam, which is currently using steel supports, through numerical modeling. Within this scope, the Ömerler Underground Mine, located in the West Lignite Enterprise (GLI), which is a part of the Turkish Coal Enterprises (TKI) and situated in the Tavşanlı district of Kütahya province, has been selected as the study area. Coal production in the Ömerler underground coal mine, which possesses a thick coal seam, is carried out using mechanized mining methods. The mining operation, which utilizes the longwall top coal caving method (LTCC) as its production method, employs a self-advancing hydraulic powered roof support system (shields) within the coal face while using a steel arch support system in the main haulage galleries and the gates belonging to the panels.

To examine the applicability of rock bolting support in the mentioned headgate, in situ and laboratory studies were conducted to determine rock mass and material properties. Based on the obtained database, empirical and numerical analyses were performed to design an appropriate rock bolting system for the field. FLAC3D v6.0 finite difference method modeling software was used for numerical modeling. With the resulting design, the resin-grouted rebar rock bolting system (RBR) and steel arch support (SAS) were implemented in the pilot application area. In situ monitoring activities using various methods were conducted to assess the performance differences between the rock bolting systems and the steel arch support system.

2 Methodology

2.1 Study Site

TKI-GLI Ömerler Underground coal mine is located in the town of Tunçbilek (Tavşanlı district of Kütahya province) in Turkey (Fig. 1).

Fig. 1
figure 1

Location of study site

Full size image

The rock units within the Tunçbilek series are grouped into three main categories, namely clay stone, calcareous marl, and marl. The clay stone formation, which surrounds the coal seam, is also subdivided into three subgroups. These subgroups consist of the soft clay layer located immediately above the coal seam with a thickness ranging from 20 to 50 cm, the roof clay forming the main roof rock of this formation, and the floor clay formations situated beneath the coal seam (Fig. 2).

Fig. 2
figure 2

Lithology of geological structure

Full size image

In the GLI Tunçbilek coal basin, underground coal production has been carried out in the Ömerler-A section. In this underground mine, a fully mechanized mining system is used, and coal extraction is performed using the LTCC method. The thick coal seam, averaging 8 m in thickness, is excavated using a single-pass method for the lower 3.5 m, while the remaining approximately 5 m at the roof level is extracted through the caving process. The cross-sectional view of the mining method is presented in Fig. 2.

In the basin, strata have generally dip angles ranging from 5 to 20° toward the northeast. The coal reserve within the study area is estimated to be around 18 million tons. The coal seam thickness varies between 5 and 12 m, with an average thickness of 8 m [36,37,38]. The coal seam contains clay partings of approximately 15–30 cm thickness at various levels. The deepest working section in the underground mine is located at an elevation of + 469, and the thickness of the overlying strata is approximately 330 m.

The study area is the A6 longwall panel in the Ömerler underground coal mine (Fig. 3). Rock mass and rock material property determination studies for coal and surrounding rock were carried out in the A1, A2, and A6 longwall panels. Empirical design studies and numerical modeling studies, rock bolting applications, and monitoring activities were conducted in the headgate of the A6 panel (Fig. 3).

Fig. 3
figure 3

The view of the A6 panel which was considered in rock bolting support design studies on the mine layout

Full size image

2.2 Determination of Rock Mass and Rock Material Properties

In order to design and implement rock bolting support for the A6 longwall panel gateroads, a series of in situ and laboratory investigations were executed to ascertain rock mass and rock material properties [11, 12]. These studies not only encompassed the A6 longwall panel where the actual implementation occurred but also included examinations in the A1 and A2 longwall panels, where preparatory and production activities were concurrently in progress. Drilling operations and block extraction studies were carried out in specific underground zones to assess rock mass and material properties, as well as for classification studies.

Underground drilling activities were carried out in the production panel, encompassing four directions. Additionally, 50 blocks were extracted from the A1 and A2 panels and transported to the rock mechanics laboratory for subsequent sample preparation. Schmidt hammer rebound hardness tests (N-type), point load strength index tests (Is50), and plate loading tests were systematically performed in the A1, A2, and A6 panels.

All rock mechanics tests were executed on the rock material samples derived from the transported blocks and drilling cores. The resulting database is comprehensively presented in Table 1. Field-based Geological Strength Index (GSI) classification studies were conducted in the A1 preparatory gallery to classify the rock mass. The determined values, coupled with outcomes from other rock mass classification systems, were computed and are presented in Table 2.

Table 1 Rock material properties
Full size table
Table 2 Rock mass properties
Full size table

Accurate predictions regarding the stability of underground openings require an understanding of the mechanical properties of the rock mass and measurements of principal stresses in the environment [39]. In line with this, studies on principal stress analysis were undertaken within the A1 longwall panel of the Ömerler underground coal mine (Fig. 4).

Fig. 4
figure 4

An example of principal in situ stress measurement on a fault and definition and measurement of fault lines

Full size image

Following Aydan’s method [40] for determining principal in situ stresses through the fault slip approach, the analysis outcomes revealed that the maximum horizontal stress is predominantly aligned in the north–south direction. Additionally, it was established that at a depth of 300 m, the most significant horizontal principal stress (PH = 6.74 MPa) in the A1 panel aligns parallel to the gate axis.

2.3 Deriving Initial Design Outcomes via Empirical Methods

The data obtained from experimental studies, observations, and examinations (Table 3) have been utilized to ascertain the design parameters for rock bolts through empirical methodologies. Specific analyses tailored to empirical techniques were conducted to define crucial dimensions such as bolt length (L) and bolt spacing (S) [11, 12]. The summarized outcomes are delineated in Table 4.

Table 3 The empirical design utilizes input parameters
Full size table
Table 4 Design results for rock bolts derived through empirical methods
Full size table

The determination of the rock bolts’ quantity (N) involved separate empirical approaches. After reviewing the N values outlined in Table 4, the average N value was computed as 6. However, empirical methods also recommend incorporating shotcrete and/or steel mesh alongside rock bolting to enhance face stability.

In addition to the rock bolting design obtained through empirical approaches, following a comprehensive assessment of field inspections, observations, and engineering experiences, it has been determined that seven rock bolts will be applied for each line. The average values specified in Table 4, along with these assessments, have influenced the design outcome shown in Fig. 5. Additionally, Fig. 5 includes the layout for the coal seam.

Fig. 5
figure 5

Rock bolt design based on empirical design results

Full size image

As shown in Fig. 5, the cross-section of the gallery illustrates the utilization of seven rock bolts, featuring a bolt length (L) of 3.3 m, a spacing between bolts (S1) of 1.0 m at the gallery face, and an interval between bolt lines (S2) of 1.0 m along the gallery axis. Based on field observations, measurements, experience, and engineering insights, the plan includes the installation of three roof-anchored rock bolts on the gallery roof in the curved section (P1, M, T1). Likewise, two inclined rock bolts with inclinations of 70° and 50° (T2 and T3, respectively) have been positioned on the pillar (T) side, and on the face (P) side, two inclined rock bolts with inclinations of 70° and 50° (P2 and P3, respectively) have been similarly installed.

2.4 Numerical Modeling

The initial design results illustrated in Fig. 5 underwent three-dimensional numerical analyses using FLAC3D v6.0 for the designated A6 panel in this investigation [11]. The analyses conducted include:

2.4.1 Modeling Procedure

During the modeling of the A6 longwall panel in the Ömerler underground coal mine, the current state of the mine was considered. The solid model incorporates the previously mined and subsided A5 panel, the planned pilot application A6 panel, and the untouched A7 panel on the opposite side of the A6 panel, as depicted in the underground mine map (Fig. 3). In the model geometry (Fig. 6), the z-direction signifies depth, the y-direction signifies the length of the longwall panel, and the x-direction represents the length of the longwall face. The model dimensions were defined as + x direction 300 m, −z direction 200 m, and + y direction 500 m.

In the model, the longwall face length in the + x direction is considered to be 90 m with a pillar width of 20 m. In the −z direction, the main strata have a thickness of 11 m, followed by a 140-m claystone unit above the coal seam, 10 m of backfill material above the claystone, and a 39-m claystone unit below the coal seam. The + y direction is defined as 500 m.

To analyze the dynamic effects arising from production in the gallery (longwall panel), it is assumed that the point where the longwall panel eliminates the initial backfill effect is at 450 m. This point is regarded as the starting point of the longwall excavation, implying that the initial 50 m of the panel has been worked and is left as a subsided area (Fig. 6).

Fig. 6
figure 6

The geometry and details of the model created in FLAC 3D

Full size image

The model geometry utilizes rectangular and square-shaped brick elements. The longwall excavation is presumed to be conducted in 1-m-thick slices, and for analyzing the dynamic effects it induces, the longwall panel is subdivided into 1-m grids between 450 and 400 m. The remaining sections are further divided into 5-m and 10-m grids. Consequently, the model comprises a total of 361,665 zones and 376,320 nodes (Fig. 6).

Boundary conditions in a numerical model involve the predetermined values of field variables (such as stress and displacement) set at the grid’s boundaries. Boundaries fall into two categories: real and artificial. Real boundaries correspond to features present in the physical object being modeled (e.g., a tunnel surface or the ground surface). Artificial boundaries, although nonexistent in reality, must be introduced to enclose the chosen number of zones. For the Ömerler underground coal mine model, roller boundaries are established on the left, right, front, and rear boundaries of the grid, while the bottom of the grid remains fixed. The results from principal stress analyses conducted in situ are incorporated into the model, with an initial condition of K0 = 0.473 (σh /σv) assigned, and the gravitational effect is also defined. In the FLAC 3D program used to create the model for the A6 longwall panel, gob material properties are specified, incorporating equations found in the literature. The mechanical behavior of the gob is represented by the double-yield model implemented in FLAC 3D.

Pappas and Mark [41] investigated the behavior of longwall gob material through laboratory tests, concluding that the equation proposed by Salamon [42, 43] in the gob model yielded results closest to laboratory tests. In the Salamon gob model, the following equation is presented (Eq. 1).

$$\sigma= \frac{{{E}}_{{0}}{\varepsilon}}{{1-\varepsilon /}{{\varepsilon}}_{{m}}}$$
(1)

In Eq. (1), σ represents the uniaxial stress (MPa) on the material, ε denotes the unit deformation of the material under stresses, E0 stands for the initial tangent modulus (MPa), and εm represents the maximum unit deformation possible in the compacted rock material.

Equation (1) was employed in modeling studies to ascertain the mechanical behavior of the gob. With each 1-m advancement in the face, the A6 panel is associated with a double-yield constitutive model for the 1-m section behind the coal face. Consequently, in the model, during the early stages of face advancement, the gob region represents an area that is broken and collapsed, incapable of withstanding the pressure from the roof. In this region, the gob undergoes slow compression, resulting in increased roof stresses.

For the A6 longwall panel model, the equation governing volumetric unit deformation behavior, along with deformation change values, is outlined in Table 5 and expressed in Eq. (2) (Tables 6, 7).

$${\sigma}{ =}\frac{{29.2 \varepsilon}}{{1-6.25 \varepsilon}}$$
(2)
Table 5 Cap pressure for the double-yield model
Full size table
Table 6 The input parameters for the SAS in the model
Full size table

The FLAC 3D utilized beam structural elements to simulate the supports of the A6 longwall panel gates in the Ömerler underground mine, incorporating SAS. These structural elements are characterized by their geometric and material properties within the FLAC 3D program. For the modeling of RBR, pile structural elements were employed in FLAC 3D. In the model, shell structural elements with an elastic modulus (E) of 180 GPa, Poisson’s ratio (υ) of 0.3, and a thickness of 45 cm were implemented to represent self-advancing hydraulic roof support units.

Table 7 The input parameters for the RBR in the model
Full size table

2.4.2 Identification and Definition of Monitoring Zones in the Model Geometry

Two separate models have been crafted for three-dimensional analyses. The model reinforced with a steel arch on galleries featuring a horseshoe cross-section is identified as SAS, while the model strengthened with resin-grouted rebar rock bolts is labeled as RBR. In both models, successive activities of preparation (stage 1) and reversible production (stage 2) are presumed to occur. To monitor the stresses and deformations produced in the model, a total of 120 monitoring points have been established. In assessing the numerical analysis results, the focus has been given to two station points positioned above the material gallery (headgate) adjacent to panels A5 and A6. These station points are U9 at 300 m and U3 at 429 m along the material gallery (Fig. 7).

Fig. 7
figure 7

Placement of the monitoring points in the numerical model

Full size image

2.4.3 Assumptions and Limitations in the 3D Model

Throughout the modeling process, specific assumptions and constraints were considered. These include:

  1. (i)

    In the model studies, σ1 is presumed to be vertical (in the −z direction), while σ2 and σ3 are considered horizontal (in the x and y directions).

  2. (ii)

    The length of the A6 longwall panel, originally spanning between 400 and 450 m, was approximated as 500 m in the y direction within the model.

  3. (iii)

    To define the shield support units in the model, were represented using shell structural elements, the SAS with beam structural elements, and the RBR with pile structural elements.

  4. (iv)

    The dip angle of the coal seam where the A6 longwall panel is situated was assumed to be 0° in the model. Additionally, groundwater was disregarded in the modeling studies, as excavation works are conducted above the underground water table.

2.5 Pilot Application and Monitoring Studies

At the TKI-GLI Ömerler underground coal mine, pilot application and monitoring studies to performance analysis of the two different support system have been carried out [11, 12]. In the coal mine, a 45-m section located at the headgate of the A6 panel has been designated as a pilot area for testing and monitoring (Fig. 8). The commonly used SAS method in the mine has been employed as support in the initial 20-m section of this area. Convergence measurement stations (CO) have been established at three points to monitor displacements associated with coal production in the SAS zone.

Following the completion of the SAS zone, a rock bolting design specified in Fig. 5 has been implemented for the RBR support system in the 25-m section. To address potential safety issues arising from the RBR application in this 25-m zone, previously existing steel arches have been loosened and put into a passive state. Convergence measurement stations (CO) have been established at five points in the RBR zone to monitor displacements associated with coal production (Fig. 8).

Fig. 8
figure 8

The application zones in the headgate and the number of CO stations in the A6 panel

Full size image

During the RBR application, after scanning the roof, three rock bolts (P1, M, T1) were initially placed in the roof. Subsequently, angled rock bolts (P2, P3, T2, T3) were installed on both sides of the gallery at angles of 50° and 70°. The rock bolt length (L) and spacing (S) were taken as 3.3 m and 1 m, respectively, based on the design detailed in Fig. 5 (Fig. 9). Holes for rock bolts were drilled using a drilling machine with a 28-mm drill bit. Four resin cartridges with a diameter of 23 mm and a length of 60 cm were placed in each hole. The solidification time for the used resins is 180 s.

For the performance evaluation of the support systems in the pilot application area, monitoring systems have been implemented in both the 20-m SAS zone and the 25-m RBR zone. In each of these zones, sections equipped with convergence (CO) measurement stations at approximately 5-m intervals have been established (Fig. 8). Every shift, one measurement was taken, and measurements continued until the 45-m area was traversed and remained beneath the caving zone due to the in-seam production activities Measurements were continuously collected from these stations over time and in conjunction with the progress of the longwall excavation.

Fig. 9
figure 9

Typical images of the stages of RBR installation

Full size image

3 Results and Discussions

3.1 Numerical Modeling Results

The A6 panel, defined as a longwall panel with a headgate length of 500 m, was separately modeled for the SAS and RBR support systems. Each numerical model was run in two separate stages. In the first stage (stage 1), excavation and reinforcement of the gateroads for the A6 panel were performed. In the second stage (stage 2), the goaf behind the supports in the A6 panel was caved, the longwall face was prepared, and coal production was carried out at 1-m intervals [11].

The results of these two stages for both models were evaluated separately for the monitoring zone at the 300th meter of the headgate in the A6 panel (U9) and the monitoring point at the 429th meter (U3). Consequently, vertical displacements and changes in vertical secondary stresses were recorded at monitoring points U3 and U9, and the performances of SAS and RBR were assessed. Figure 10 presents the graphical representation of vertical displacement and changes in vertical secondary stresses for the SAS reinforcement model in stage 1. The data obtained from these graphs are specified in Table 8.

Fig. 10
figure 10

(a) Vertical displacements and (b) vertical secondary stresses in the gallery roof at the 300th m of the headgate (monitoring point U9), during gateroad excavation and SAS reinforcement in the A6 panel (Stage-1)

Full size image
Table 8 Vertical displacements and vertical secondary stress values of U9 for stage 1 (SAS)
Full size table

The data in Table 8 and Fig. 10 show that vertical displacements and vertical secondary stresses remain very low in the first 200 m as the excavation face approaches the monitoring point located at the 300th meter of the headgate, denoted as U9. Even with 100 m remaining to reach the U9, the values appear to stay close to the initial primary values in the field. However, as the excavation face approaches the U9 at the 300th meter, vertical displacements and secondary stresses start to change rapidly, and when the excavation face reaches the U9, these values are U = 57 mm and P=−40 kPa. After the excavation face passes the U9, vertical displacement continues to change rapidly up to the 400th meter. As the excavation face reaches from the 400th to the 500th meter in the gallery, it is understood that the vertical displacement values at the remaining U9 in the gallery change very little, reaching U = 96.2 mm. Similarly, it is understood that the vertical stress values undergo very little change up to the 500th meter after the excavation face passes the U9. When the excavation face reaches the 500th meter in the gallery, it is observed that the vertical stress values at the remaining U9 point in the gallery remain almost constant, reaching P=−10 kPa.

The vertical displacement and secondary stress values at the U3 monitoring point located at 429 m on the A6 longwall panel (Fig. 7) were determined using FLAC 3D. The model outputs showing the vertical displacement and secondary stress values at the U3 during the period from the start of excavation in the longwall face to the 18th meter of the advancing excavation face in the completed preparation panel (stage 2) are presented in Fig. 11. The vertical displacement and secondary stress values observed in the model outputs are presented in Table 9.

Table 9 and Fig. 11 illustrate the longwall advancement at 1-m intervals in the model. The monitored U3 point is located at the 429th meter of the headgate, initially positioned 18 m behind the excavation face. As excavation progresses in the longwall face, displacements and stresses at the U3 monitoring point begin to change. While displacements are relatively minor within the first 10 m, they rapidly escalate within the final 8 m (Fig. 11). Upon the longwall excavation advancing 18 m and reaching the U3 monitoring point, the total displacement amounts to U = 376 mm. Initially characterized by tensile stresses, stress variations subsequently transition to compressive stresses. Upon complete advancement of the longwall excavation to the U3 monitoring point, the stress value in the gallery roof amounts to P=−1.8 kPa.

Similar to the SAS support model, the gateroads of the A6 longwall panel were modeled using the RBR support system. Vertical displacement and secondary stress results at the monitoring point labeled U9 are presented in Fig. 12 and Table 10.

Fig. 11
figure 11

(a) Vertical displacements and (b) vertical secondary stresses in the gallery roof at the 429th m of the headgate (monitoring point U3), during gateroad excavation and SAS reinforcement in the A6 panel (Stage-2)

Full size image
Table 9 Vertical displacements and vertical secondary stress values of U3 for stage 2 (SAS)
Full size table
Fig. 12
figure 12

(a) Vertical displacements and (b) vertical secondary stresses in the gallery roof at the 300th m of the headgate (monitoring point U9), during gateroad excavation and RBR reinforcement in the A6 panel (Stage-1)

Full size image
Table 10 Vertical displacements and vertical secondary stress values of U9 for stage 1 (RBR)
Full size table

Table 10 and Fig. 12 reveal that within the initial 200 m, vertical displacement and secondary stress values remain relatively low as the excavation face progresses towards the monitoring point at the 300th meter of the headgate, identified as U9. Even with only 100 m left to reach U9, these values stay close to their initial measurements. However, as the excavation face nears U9, both vertical displacements and secondary stresses start to rise rapidly. Upon reaching U9, these values reach U = 14.7 mm and P=−3463 kPa. Beyond U9, vertical displacement continues to fluctuate rapidly until the 400th meter. From the 400th to the 500th meter, minimal changes are observed in vertical displacement at the remaining U9, stabilizing at U = 26.4 mm. Similarly, vertical stress values decrease rapidly until the 400th meter after surpassing U9, then stabilize until the 500th meter. At the 500th meter, vertical stress values at the remaining U9 remain almost constant, reaching P=−3344 kPa (Table 10; Fig. 12).

The vertical displacement and secondary stress results at the monitoring point labeled U3, up to the 18th excavation of the 1-m coal face (stage 2), for the RBR support model are presented in Fig. 13 and Table 11.

Fig. 13
figure 13

(a) Vertical displacements and (b) vertical secondary stresses in the gallery roof at the 429th m of the headgate (monitoring point U3), during gateroad excavation and RBR reinforcement in the A6 panel (Stage-2)

Full size image
Table 11 Vertical displacements and vertical secondary stress values of U3 for stage 2 (RBR)
Full size table

As seen in Table 11 and Fig. 13, the longwall advancement in the model was executed at 1-m intervals. The monitored U3 point is located at the 429th meter of the gateroad. This point is initially positioned 18 m behind the excavation face. As excavation progresses in the longwall face, displacements and stresses at the U3 monitoring point begin to change. Displacement values, which initiate within the first 10 m, rapidly escalate within the final 8 m (Fig. 13). Upon the longwall excavation advancing 18 m and reaching the U3 monitoring point, the total displacement amounts to U = 106 mm. While stress variations remain nearly constant up to the last 8 m, they subsequently decrease rapidly, culminating at a stress value of P=−1840 kPa when the excavation face reaches the U3 monitoring point (Table 11; Fig. 13).

3.2 In Situ Measurement Results

Continuous convergence measurements were taken from CO stations set up in the field to determine the performance of the existing SAS reinforcement in the 20-m zone in front of the coal face where production takes place and the subsequent 25-m zone where RBR is applied [12]. In the first 20-m section with SAS located just ahead of the coal excavation face, measurements were taken from three CO (CO-1, CO-2, and CO-3) stations, while in the following 25-m section with RBR, measurements were taken from five CO measurement stations (CO-4, CO-5, CO-6, CO-7, and CO-8). The measurement results obtained over time are presented in Fig. 14, and those related to the distance from the coal face are depicted in Fig. 15.

As depicted in Fig. 14, the blue curves representing the SAS zone have experienced faster and larger deformations compared to the red curves representing the RBR zone. Convergence values in the SAS zone ranged from 234.46 to 391.60 mm, while in the RBR zone, they ranged from 105.78 to 203.94 mm. Due to the distance of 10 m between the last station CO-3 within the SAS zone and the starting point of the RBR zone, it is considered that the RBR zone affects this point within the SAS zone, resulting in convergence values measured at the CO-3 station to be lower than those at CO-1 and CO-2 stations.

Fig. 14
figure 14

Time-dependent convergence curves obtained in the SAS and RBR zones of the A6 headgate

Full size image
Fig. 15
figure 15

Convergence curves as a function of distance from the coal face obtained in the SAS and RBR zones of the A6 headgate

Full size image

Figure 15 considers the progress distances in the coal face. As depicted in the presented graph, in the SAS-zone (blue curves), convergence values reach 391.602 mm at the CO-1 station, which is closest to the excavation face. Similarly, in the RBR zone, the convergence value measured at the CO-4 station, closest to the excavation face, is 203.94 mm. As the excavation face progresses in the RBR zone, the convergence values read at the other CO-5, CO-6, CO-7, and CO-8 stations in this zone reach a maximum of 105.78 mm. This situation is related to the liberation of the gallery roof from the influence of the SAS zone.

When considering the maximum displacement values observed under the same conditions for both zones, it is determined that the RBR-reinforced zone performs its duty of supporting the roof during coal production activities 52% more successfully than the SAS-reinforced zone.

3.3 Numerical Modeling and In Situ Measurement Results Evaluation

In this section, the numerical modeling results and in situ measurement results of vertical displacement values associated with the advancement of the coal face have been compared on the same graph for SAS and RBR support systems (Fig. 16). Vertical displacement values from the numerical models created for SAS and RBR support systems were compared with convergence measurement results obtained in the field at the CO-2 station located in the middle of the SAS zone and the CO-6 station located in the middle of the RBR zone, relative to the monitored U3 observation point. These results are presented on the same graph in Fig. 16.

Fig. 16
figure 16

Comparison of numerical modeling and in situ measurement results for vertical displacement values due to 18-m coal excavation for SAS and RBR

Full size image

As seen in Fig. 16, according to the numerical model results, a vertical displacement of 250 mm occurred in the roof over the 18-m advancement of the coal face for the SAS, while the in situ measurements indicate this value to be 363.76 mm. For the RBR, the numerical model results predict a vertical displacement of 51 mm over the 18-m advancement of the coal face, whereas the in situ measurements show this value to be 76.36 mm. Consequently, Fig. 16 reveals that both the numerical model results and the in situ measurement results indicate that the RBR reinforcement system performs its support function more effectively during the advancement of the coal face compared to the SAS reinforcement system.

In Fig. 16, differences between the vertical displacement results obtained from numerical analysis and in situ measurements are observed in both RBR and SAS systems. These differences are due to the local complex geological conditions of the measurement area and the unsystematic nature of mining activities. In other words, these differences are associated with problems hindering the systematic progress of mining activities in the A6 panel. While activities such as reinforcement design are systematically conducted in numerical analyses, real-world mining schedules are not always executed as planned due to issues like belt malfunctions, ventilation problems, and equipment breakdowns. Predicting and integrating such unforeseen circumstances into numerical models, especially in underground mining activities, is quite challenging. A similar situation holds for the Ömerler coal mine site discussed in this study. Therefore, the differences observed between the numerical analysis outputs and the actual measurement results in Fig. 16 can be attributed to irregular delays caused by unwanted issues in mining activities. In conclusion, the comparison presented in Fig. 16 demonstrates that the RBR system provides more effective support during the advancement of the coal face compared to the SAS system. This finding is supported by both the numerical model results and the in situ measurements.

4 Conclusions

In this study, the feasibility of rock bolting support in an underground coal mine gallery with a thick coal seam has been investigated through numerical modeling. In this context, the Ömerler underground coal mine area, belonging to the TKI-GLI which is located in the Tavşanlı district of Kütahya province, was selected as the study area. Rock mass and rock material properties were determined through in situ and laboratory studies conducted in the mine, and experimental, empirical, and numerical analyses were performed based on the obtained data. Numerical modeling was conducted using FLAC3D v6.0 finite difference method modeling software.

According to the design results obtained, the RBR and SAS were implemented in the pilot application area. In situ monitoring studies were carried out to evaluate the performance of the RBR and SAS systems in this pilot application area.

The results of the numerical modeling indicated that there were less displacement and less secondary stress change in the RBR-supported area compared to the SAS-supported area. Additionally, in situ measurements showed that the RBR more successfully supported the roof during coal production activities. It was determined that the RBR supported the roof 52% more effectively compared to the SAS.

These findings demonstrate that when evaluating the applicability of rock bolting support systems in underground galleries with thick coal seams in the Ömerler underground coal mine, the RBR system is a more effective solution. This study emphasizes the importance of more sustainable and secure support systems to enhance operational efficiency in the coal mining industry.