To obtain rockburst characteristics at rock engineering, the rockburst event, in situ stress characteristics, and rockburst proneness are studied in a deep gold mine. The severity of rockburst increases with cover depth at the mine. The main locations of spalling and rockburst events include roof and sidewall. The in situ stress measurement results show this area that is controlled by horizontal constructional pressure, and the self-weight stress also has an important influence. The rockburst mechanism is analyzed from the principal stress and tangential stress. When the maximum tangential stress (MTS) does not exceed 0.4 times UCS of surrounding rock, no obvious disaster in tunnel is found. When the MTS of tunnel is between about 0.4 ~ 0.55 times UCS of rock, the tunnel may suffer spalling. When the MTS of tunnel exceeds 0.55 times UCS of rock, serious failure may occur in tunnel, such as rockburst and large roof collapse. When surrounding rock is relatively hard and complete (high impact energy index, elastic strain energy index, linear elastic energy index, and RQD index), the tangential stress plays a very important role in the rockburst at the mine.

With the growth of human social productivity, the demand of industry and agriculture for mineral resources is getting higher. Less shallow surface resources have been difficult to provide enough demand of human beings, and the deep mining of underground resources becomes more important. [14]. The mining depth for more than 100 metal mines exceeds 1000 m [58]. However, some special phenomena in deep hard rock mines have occurred, such as spalling, rockburst, and rock core disking [911]. In order to deal with these disasters, a lot of researches were done on the mechanism of rockburst, rock mass support, and rockburst prediction. The research on rockburst mechanism includes characteristics of in situ stress [1214], rock mass quality [1517], indoor rock experiment [1824], and rock failure criterion [2530]. The rock mass support research includes energy-absorbing anchor [3134], goaf filling [3538], and combined support [3942]. The rockburst prediction research includes microseismic monitoring and early warning [4346] and rockburst intelligent prediction [4754].

In the past, many useful achievements have been made in the research of rockburst. For example, Martin and Christiansson [55] analyzed the failure of AECL’s and SKB’s Äspö test tunnels in combination with the corresponding rock strength and stress. The surrounding rock spalling strength is 0.58, 0.56, and 0.65 times of UCS, respectively. The proportion of spalling strength to tunnel’s maximum tangential elastic stress was defined as the safety factor. As the average safety factor exceeds 1.25, the spalling decreases to a very low value probability. Li et al. [56] obtained the conditions for transition from shear failure to slabbing failure of rock through compression tests under different height-width proportion and found that slabbing strength of granite is close to 0.6 times UCS. Jiang et al. [57] conducted numerical analysis on the characteristics of rockburst of Jinping II Hydropower Station, found that the strain energy played a key part in inducing rockburst, and proposed an energy index to predict rockburst, which can better estimate the strength of rockburst and rockburst pit depth. Cai and Kaiser [58] analyzed the relationship between in situ rock spalling strength and UCS through theoretical methods and numerical simulation. They believed that the in situ rock strength was not measured but interpreted and could not be simply equal to the rock cracking stress under laboratory testing. Irregularity of the excavation boundary and size effect is the main reason that causes the (apparent) in situ rock spalling strength being about 0.3 ~ 0.5 times the laboratory UCS.

However, rockburst, rock spalling, large-area collapse, large deformation, and other disasters still occur often in deep rock engineering. The rockburst often leads to the failure of supporting structure. The rockburst disasters still pose a threat to personnel safety and production. In the aspect of rockburst disaster mechanism, some studies are carried out from the stress concentration. Some studies have simulated the spalling and ejection process of rock through true triaxial test, but the stress value of the test machine acting on the rock sample far exceeds the in situ stress of general deep engineering. In a deep project, the sidewall of the tunnel with 800 m coverage depth is easy spalling, and the in situ vertical principal stress is close to 22 MPa [29]. In a true triaxial test, when the hole inside the sample is spalling, the vertical stress applied by the testing machine to the sample is close to 150 MPa [59], which is 7 times of vertical stress in actual project. This reflects that the mechanism of rockburst in deep engineering needs more research.

Therefore, based on in situ investigation of rockburst and in situ stress characteristics in a gold mine, this paper analyzes the rockburst mechanism from two aspects: tangential stress concentration and rockburst tendency. The rockburst characteristics in the mine are summarized, and the rockburst events and proneness with different cover depths are analyzed and discussed. This is conducive to promoting people’s understanding for deep ground pressure disasters.

2.1. Project Overview

As shown in Figure 1, the Xincheng gold mine is situated in Shandong Peninsula, China. It is a large deep well gold mine. The average elevation of mining area is 30 m, covered by the Quaternary. The ore body strike is 360 m, and the control-inclined depth exceeds 1200 m. The ore body strikes northeast and tends to northwest, with dip angle of 29°. The mining depth has exceeded -1080 m, which belongs to deep mining category. The surrounding rock is sericite granodiorite. The average Brazilian tensile strength and uniaxial compressive strength (UCS) of surrounding rock are 9.4 MPa and 150.5 MPa. With the continuous development of engineering and mining operations to the deep, the ground pressure disasters such as rock collapse, spalling, and rockburst are significantly enhanced.

2.2. Characteristics of Ground Pressure Disasters

To obtain the rockburst characteristics at the gold mine, the project site was investigated. The project investigation includes in- situ investigation and consultation with technicians. Table 1 and Figures 25 show the disaster events. The spalling, rockburst, and large area collapse are the main disaster events at the mine. The cracking of sidewall starts from 745 m cover depth, the tunnel spalling starts from 830 m cover depth, and the rockburst and roof collapse events start from 980 m cover depth. With the increase in cover depth, the ground pressure disaster events become more serious. The main locations of rockburst and spalling in the mine include roof and sidewall, while some rockburst occurred in excavation face, as shown in Figure 6. The number of rockburst and spalling in the roof and sidewall is basically equal. Considering the large in situ stress in the deep, these disasters should be caused by the horizontal stresses and vertical stresses.

3.1. Characteristics of In Situ Stress Field

To analyze characteristics of in situ stress at the gold mine, the strain relief method is adopted to obtain the in situ stress. The layouts for measuring points (measuring holes) and drilling parameters are shown in Table 2. Six measuring points are selected in the deep area. Among them, two measuring points are selected in horizontal tunnel with cover depth of 860 m. One measuring point is selected on ramp with cover depth of 925 m. One measuring point is selected on ramp with cover depth of 970 m. Two measuring points are selected in horizontal tunnel with cover depth 1060 m. The measuring points basically avoid the stress concentration areas of tunnel and stope, such as crossing and inflection points, as well as rock fracture development zones. The measuring points are arranged as far away as possible from large goaf and chamber to ensure accuracy of the stress data.

After the in-site strain relief measurement, core tube elastic parameter determination, and three-dimensional in situ stress comprehensive calculation, the calculation results of stress components for 6 measurement points in deep location of the gold mine are listed in Table 3. The results for maximum, intermediate, and minimum principal stresses are listed in Table 4. The calculation results of horizontal and vertical principal stress are listed in Table 5. In order to facilitate the calculation of tunnel stress concentration, the concepts of horizontal and vertical principal stresses are adopted. The maximum horizontal and minimum horizontal principal stresses are in horizontal direction, which can be calculated through normal stress in x and ydirections and shear stress in xy direction. Vertical principal stress is normal stress in z direction. The azimuth angle is positive when rotating clockwise from the north direction.

From Tables 35, the stress distribution in the deep of the gold mine has the following characteristics:

  • (i)

    The principal stresses and normal stress components increase with increase in cover depth. In the same level, although magnitude and direction of in situ stress at different measuring points change to a certain extent, there is no sudden change, indicating that in situ stress distribution on the level in deep part is relatively uniform

  • (ii)

    The azimuth range for maximum principal stress is 260° ~290°, which is close to the east-west, slightly biased towards the NWW-SEE direction. The dip angles for maximum principal stress at six points are small, indicating this area is controlled by horizontal constructional pressure

  • (iii)

    The proportion of intermediate to maximum principal stress is between 0.65 and 0.76. The dip angle of the intermediate principal stress mostly exceeds 60°, which is in the near-vertical direction, indicating that self-weight stress also has an important influence in the deep mining area

  • (iv)
    The stress values of six points were analyzed by regression method. The regression curves (Figure 7) and regression equation (Equation (1)) of principal stresses with cover depths are obtained
    (1)σH=0.0461h7.11,σv=0.0266h,σh=0.0294h9.34,
    where σH, σv, and σh are maximum horizontal, vertical, and minimum horizontal principal stress, respectively; h is cover depth. The effective range for Equation (1) is about from 600 m to 1200 m of cover depth.

3.2. Rockburst Analysis Based on Stress

Figure 8 is the stress state of horizontal transportation tunnel at the mine, and σH is basically orthogonal to tunnel axis. The ground pressure disaster events in Table 1 are marked in Figure 9 according to the corresponding cover depth. The relationship between disasters, σH and σv, is clearly obtained. Table 1 shows that spalling and rockburst occurred in roof and sidewall. Roof failure is related to the horizontal stress, and sidewall failure is related to the vertical stress, which shows that the maximum horizontal and the vertical principal stresses should be paid attention to. The surrounding rock’s UCS is about 150.5 MPa. Figure 9 shows that when σv exceeds 0.15 times UCS or the σH exceeds 0.20 times UCS, the tunnel may suffer spalling. When σv exceeds 0.18 times UCS or σH exceeds 0.25 times UCS, the tunnel is prone to rockburst.

As shown in Figure 10, according to Kirsch formula, maximum tangential stress (MTS) around circular tunnel is 3q under uniform stress q [60, 61]. In this study, the MTS σθ of arch tunnel is determined by MTS of circular tunnel, which is an approximate value, which is shown in Equation (2).
(2)σθ=3σHσv=0.1117h21.33.

The ground pressure disaster events in Table 1 are marked in Figure 11 through the MTS. When the MTS at the gold mine does not exceed 0.4 times UCS of rock, no obvious disaster phenomenon is found. When the MTS of tunnel is between about 0.4 ~ 0.55 times UCS of rock, tunnel may suffer spalling. When the MTS of tunnel exceeds 0.55 times UCS of rock, rockburst and large-area roof collapse may occur in tunnel. As the mining depth at the mine has not exceeded 1100 m, the rockburst at the cover depth of 1100 m is unclear. According to the criterion of Wang and Park [62], strong rockburst is easy to occur when the tangential stress exceeds 0.7 UCS. In some studies, in situ spalling strength is about 0.4±0.1 times of UCS [58, 63], and the empirical conclusion is widely accepted. The research in this work also conforms to the conclusion.

4.1. Evaluation Indicators

The internal cause of rockburst is the sudden elastic energy release of hard rock, which results rock block ejection and rockburst formation. The external cause of rockburst is high stress, and excavation causes stress redistribution and stress concentration around the cavern. Therefore, in order to obtain rockburst proneness, internal and external causes of rockburst should be comprehensively considered. A total of six evaluation indicators are selected from three aspects of surrounding rock: mechanical properties, rock conditions, and stress state. These indicators include impact energy index, deformation brittleness index, elastic strain energy index, linear elastic energy index, rock mass quality designation, and tangential stress index.

4.1.1. Impact Energy Index

The impact energy index (WCF) refers to the proportion of elastic strain energy stored (E1) before peak load by rock test to energy consumed (E2) in process after peak load until complete failure. The impact energy index is calculated according to
(3)WCF=E1E2.

The impact energy index can be divided into 4 grades: (1) no rockburst as WCF<1; (2) weak rockburst as 1WCF<2; (3) strong rockburst as WCF2 [64].

4.1.2. Deformation Brittleness Index

The deformation brittleness index (Ku) is obtained by permanent deformation (u1) and total deformation (u) before the peak strength of sample in uniaxial loading and unloading test, and its expression is as follows:
(4)Ku=uu1.

The deformation brittleness index can be divided into 4 grades: (1) no rockburst as Ku<2; (2) weak rockburst as 2Ku<6; (3) moderate rockburst as 6Ku<9; (4) strong rockburst as Ku9 [65].

4.1.3. Elastic Strain Energy Index

The index (Wet) is proportion of stored elastic strain energy sp in sample to plastic deformation dissipation energy st before the peak strength of the rock, and its calculation expression is
(5)Wet=ϕspϕst.

The index is divided into 4 grades: (1) no rockburst as Wet<2; (2) weak rockburst as 2Wet<5; (4) strong rockburst as Wet5 [66, 67].

4.1.4. Linear Elastic Energy Index

The index (We) before sample’s peak load in uniaxial compression is
(6)We=σc22Es,
where σc is UCS (MPa) and Es is unloading tangent elastic modulus (GPa). The index can be divided into 4 grades: (1) no rockburst as We<50; (2) weak rockburst as 50We<100; (3) moderate rockburst as 100We<150; (4) strong rockburst as We150 [62].

4.1.5. Rock Mass Quality Designation

The rock quality designation (RQD) index refers to the proportion of sum length in core longer than 10 cm to the total length of rock core drilled in this cycle. The rock quality designation can be divided into 4 grades: (1) no rockburst as RQD<25%; (2) weak rockburst as 25%RQD<50%; (3) moderate rockburst as 50%RQD<70%; (4) strong rockburst as RQD70% [68].

4.1.6. Tangential Stress Index

The proportion of MTS (σθ) in tunnel to UCS (σc) of rock is defined as the tangential stress index (T):
(7)T=σθσc.

The tangential stress index can be divided into 4 grades: (1) no rockburst as T<0.3; (2) weak rockburst as 0.3T<0.5; (3) moderate rockburst as 0.5T<0.7; (4) strong rockburst as T0.7 [62, 69].

4.2. Evaluation of Rockburst Proneness

Indoor rock mechanical tests were carried out using borehole cores, including rock uniaxial compression, uniaxial loading, and unloading testing. Part of the test curves is shown in Figure 12. The boreholes are located near the new main shaft, and the drilling depth ranges from -930 m to -1570 m. Seven measuring points are arranged every 100 m from -950 m to -1550 m. Uniaxial compression and loading and unloading test at each depth have been conducted for 3 times, respectively, 42 times in total. The rock is mainly sericitized granite. According to mechanical experiment results and calculation method for each index of rockburst proneness, Table 6 shows the calculation results of each index of rockburst proneness.

According to the grading standard for rockburst proneness of each index, Table 7 shows the evaluation results of each index of rockburst proneness. At the coverage depth of 950 m to 1550 m, rockburst proneness obtained through impact energy and linear elastic performance indexes is strong. The rockburst proneness obtained through elastic strain energy index and RQD indexes is basically strong. The rockburst proneness obtained through tangential stress index increases from medium to strong with the increase of burial depth. The rockburst proneness obtained through deformation brittleness index has no obvious trend. It can be known from Section 2.2 and Section 4.2 that the rock mass buried below -950 m has strong rockburst proneness, and rockburst events have occurred. This shows that the five indexes of impact energy, elastic strain energy, linear elastic energy, tangential stress, and RQD indexes are more suitable for rockburst proneness analysis of the gold mine.

When the surrounding rock is relatively hard and complete (high impact energy and elastic strain energy, linear elastic energy, and RQD indexes), the tangential stress plays a very important role in the rockburst. This is the main reason why the rockburst mostly occurs in the deep, because the deep in situ stress is greater. Figure 11 shows that when the tangential stress index does not exceed 0.4, no obvious disaster phenomenon is found at the gold mine. When tangential stress index is between 0.4 and 0.55, there are some spalling phenomena, which can be regarded as weak rockbursts. When the tangential stress index exceeds 0.55, there are some rockbursts and large-area collapse. When the tangential stress index exceeds 0.7, the surrounding rock has strong rockburst proneness according to the criteria of Wang et al. [62, 69]. For the gold mine, the rockburst proneness analysis can be carried out according to the following modified tangential stress index criterion, as shown in Table 8.

The rockburst proneness is mainly determined by mechanical properties and integrity of surrounding rock and in situ stress. Considering that mechanical properties and integrity of rock are difficult to be changed artificially, a reasonable engineering arrangement considering Azimuth of maximum principal stress is an important method to reduce tangential stress concentration. In addition, the faster excavation speed results the faster strain energy release. The larger the single excavation volume, the higher the energy release, which are also important reasons of rockburst events. Energy-absorbing materials can be used to assimilate the released strain energy, which can obtain better support effect, and this has attracted the attention of some scholars [34, 70]. Therefore, controlling the excavation speed and single excavation volume, adopting energy-absorbing materials, and arranging pressure relief holes in advance on the working face can effectively reduce the rockburst.

With the increase of cover depth, ground pressure disaster events become more serious. The cracking of sidewall starts from 745 m cover depth, the tunnel spalling starts from 830 m cover depth, and the rockburst and roof collapse events start from 980 m cover depth. The main location of rockburst and spalling in the mine includes roof and sidewall. Roof failure is related to horizontal stress, and sidewall failure is related to vertical stress.

Principal stresses and stress components show trend of increasing with increase of cover depth. Dip angles of σ1 are all small, indicating that this area is controlled by horizontal constructional pressure. Dip angle of σ2 is in near-vertical direction, indicating that self-weight stress has an important influence in the deep mining area.

The rockburst mechanism is analyzed from the principal stress and tangential stress concentration. When the MTS of tunnel does not exceed 0.4 times UCS, no obvious disaster is found. When the MTS of tunnel is between about 0.4 ~ 0.55 times UCS, surrounding rock may suffer spalling. When the MTS of tunnel exceeds 0.55 times UCS, rockburst and large-area roof collapse may occur in tunnel.

According to the in situ investigation results and rockburst proneness analysis, it is shown that the gold mine has strong rockburst proneness below -950 m. The five indexes of impact energy, elastic strain energy, linear elastic energy, RQD, and tangential stress indexes are more suitable for rockburst proneness analysis of the gold mine. When the surrounding rock is relatively hard and complete (high impact energy, elastic strain energy, linear elastic energy, and RQD indexes), the tangential stress plays a very important role in the rockburst.

The data used to support the results of this study can be found in this manuscript text.

The authors declare no competing interests.

This work was supported by the Fundamental Research Funds for the Central Universities of Central South University (No. 2021zzts0278) and the State Key Research Development Program of China (No. 2018YFC0604606).

Exclusive Licensee GeoScienceWorld. Distributed under a Creative Commons Attribution License (CC BY 4.0).