Analysis of support requirements for underground water-sealed oil storage cavern in China

Tunnelling and Underground Space Technology 71 (2018) 36–46

Contents lists available at ScienceDirect

Tunnelling and Underground Space Technology journal homepage: www.elsevier.com/locate/tust

Analysis of support requirements for underground water-sealed oil storage cavern in China

MARK

⁎

Jie Liua, Xing-Dong Zhaoa, , Shu-Jing Zhanga, Lian-Ku Xiea,b a b

School of Resources & Civil Engineering, Northeastern University, Shenyang, Liaoning 110819, China Beijing General Research Institute of Mining & Metallurgy, Beijing 100160, China

A R T I C L E I N F O

A B S T R A C T

Keywords: Underground oil storage cavern Rock mass classification Numerical method Support design

The aim of this study is to provide a sound support design for a large-scale underground water-sealed oil storage facility in China. The lithology of this study area consists of reddish-gray, medium-coarse grained granites interpenetrated by diabase, amphibole dioritic porphyrite and aplite. In order to obtain the geotechnical properties of intact rocks and rock masses, detailed field and laboratory studies were carried out. Rock masses at three sites were characterized in terms of Rock Mass Rating (RMR), Rock Mass Quality (Q-system) and Geological Strength Index (GSI), and then rock mass properties of underground caverns were estimated, accordingly. The support systems obtained from empirical methods were analyzed using FLAC3D commercial software considering the effect of the water curtain system. The maximum thickness of plastic zones and the maximum total displacement occurred around underground caverns after installing the support systems suggested by the empirical methods were compared to the unsupported case. The results show that more reliable support design could be obtained by using both empirical rock mass classifications and numerical analysis method.

1. Introduction China is short of oil resources and the contradiction between supply and demand has been increasingly prominent in recent years. In 2016, China consumed a total of 0.562 billon tons of oil, among which 67.3% was imported from other countries (Tian, 2017). As the energy issues becoming increasingly prominent, safe and stable natural petroleum supplies are of critical importance to China’s economic and social development. Underground water-sealed storage has many advantages of less land occupation, higher security, lower cost and more environmental benefits than aboveground storage, which is widely considered as a technically sound and economically feasible storage method. Underground water-sealed storages for crude oil, liquefied petroleum gas (LPG) and liquefied natural gas (LNG) have been developed worldwide since the early twentieth century, for example the LPG storage in the Seto inner-sea area of Japan (Tezuka and Seoka, 2003), the crude-oil storage in Korea (Lee and Song, 2003) and the hydrocarbon storage in the Perama area of Greece (Benardos and Kaliampakos, 2005). In 2003, China began to construct the national strategic oil storage bases. At the Stage Ⅱ Plan of China since 2008, the seven to eight underground water-sealed depots have been currently in the implementation stage. The four underground water-sealed storage caverns in Huangdao in Shandong Province, Jinzhou in Liaoning Province, Zhanjiang and

⁎

Huizhou in Guangdong Province have started construction and come into use gradually. The basic principle of underground oil storage in unlined rock caverns is that groundwater pressure around caverns should be higher than the storage cavern pressure and thus groundwater flows into caverns to prevent oil leakage. The water-sealing effect and rock mass stability are two key problems for the construction of underground storage caverns (Zhuang et al., 2017). Suitable water pressure can be obtained by locating the caverns at a sufficient depth or by installing a water curtain system. When the hydrodynamic containment of storage caverns cannot be maintained by natural groundwater, an artificial water curtain system consisting of galleries and arrays of the drilled holes above the crown of caverns would be needed. Moreover, groundwater flow into a cavern may induce a potential hazard for cavern stability (Sun and Zhao, 2010). Wang et al. (2015) investigated the design and testing of water curtain systems in detail. The seepage field analysis of underground water-sealed storage have also been conducted using numerical simulations by many researchers (Sun and Zhao, 2010; Sun et al., 2011; Yu et al., 2013; Lin et al., 2016; Li et al., 2017). Among them, Lin et al. (2016) developed a transient unified pipe network method to evaluate the water-sealing effectiveness and Li et al. (2017) presented a three-dimensional numerical method for seepage analysis of the water-sealed oil storage caverns considering the spatial effect of

Corresponding author. E-mail address: [email protected] (X.-D. Zhao).

http://dx.doi.org/10.1016/j.tust.2017.08.013 Received 25 December 2016; Received in revised form 11 August 2017; Accepted 11 August 2017 0886-7798/ © 2017 Elsevier Ltd. All rights reserved.

Tunnelling and Underground Space Technology 71 (2018) 36–46

J. Liu et al.

form an independent oil storage unit. The oil inside every storage group is sealed by the horizontal and vertical water curtain holes drilled from the water curtain tunnels. Fig. 2 shows the cross-section view of the project. Two horseshoeshaped access tunnels are 8 m wide and 8 m high. Every storage cavern is 19 m wide, 24 m high and 934 m long, and the cross-section shape is three-center arched roof and straight wall. The distance between the adjacent caverns is 38 m. The floor level of the storage caverns is at an elevation of −80 m. The water curtain tunnels, which are also arched, are 6.5 m wide, 6 m high and 974 m long. The floor elevation of water curtain tunnels is −32 m and the horizontal water curtain system is installed 25 m above the storage caverns. The vertical water curtain holes extend 10 m beyond the cavern floor. The diameter of the water curtain boles is 100 mm and the spaces of the horizontal and vertical holes are 10 m and 20 m, respectively. The tunnels and caverns are excavated using the drill-and-blast method. For the large-cross-section caverns, these caverns are excavated in three equal height sections (top heading, bench-1 and bench-2), each of which is 8 m high.

water curtain boreholes and the influence of oil vapor. Although underground water-sealed storage caverns are generally constructed within hard rock masses, these cavern are characterized by large-crosssection, multiply excavation faces, uncertain discontinuities and long term effect by a dynamic groundwater seepage, which determines that underground oil storage caverns have a lot of differences with other underground constructions. The stability of storage caverns is one of the key requirements during construction and operation stages. Many studies have been conducted on the stability of underground storage caverns using numerical methods and in-situ testing and monitoring. Lee et al. (1997) studied the behavior of oil-storage caverns during excavation based on the instrumentation measurements and numerical analysis. Chen et al. (2013) solved the hydro-mechanical problem for underground oil storage caverns using discontinuous deformation analysis. Aiming at the Huangdao project in China, Wang et al. (2013) studied the influences of construction sequence on the hydro-mechanical behavior using Finite Element Method (FEM); Li et al. (2014) analyzed the hydro-mechanical behavior using Discrete Element Method (DEM); Qiao et al. (2016) systematically presented the geotechnical monitoring. For the Jinzhou project in China, Ma et al. (2016) evaluated the stability and damage mechanism of underground caverns by integrating Continuous-Discontinuous Element Method (CDEM) and microseismic monitoring. On this basis, Zhuang et al. (2017) investigated the temporal-spatial evolution of the micro-cracks and energy-release patterns induced by excavation unloading. The above studies provide valuable engineering references for constructing an underground oil-storage facility. However, studies involving the detail support design of the storage caverns have seldom reported. To determinate reasonable support systems is crucial to ensure projects to be carried out economically, safely and successfully. There are many influence factors on stability of storage caverns, such as geology, hydrogeological characteristics and mechanical properties of rock masses, and some of these factors are difficult to monitor during construction. Therefore, it would be improper to simply use the experience in support design. The Geomechanics Classification or the Rock Mass Rating (RMR) (Bieniawski, 1989), Rock Mass Quality (Q) (Barton, 2002) and Geological Strength Index (GSI) (Marinos and Hoek, 2000) classifications are the most widely used empirical methods, which are preferred by rock engineers and have gained a universal acceptance. However, these empirical methods cannot provide the stress and displacement information. Therefore, in this study, the numerical analysis of support systems obtained from empirical rock mass classifications are used to evaluate the support system based on an underground water-sealed oil storage facility in China. The groundwater flow is also considered in the numerical analysis.

2.2. Geology and hydrogeology There are no active and regional faults near the project area. Some inactive faults exist in near-field region and these faults extend mainly in three directions: SN, NE and NW. Based on core logging data, the strata at the project area are divided into the residual soil layer, the completely weathered layer, the strong weathered layer, the moderately weathered layer, the slightly weathered layer and the unweathered layer according to the weathering and integrity levels of rock masses. The bedrock at the area consists mainly of metamorphic of Archeozoic period and intrusive rock. The predominant rock types are reddish-gray, medium-coarse grained granites interpenetrated by diabase, amphibole dioritic porphyrite and aplite. According to the evaluation of Q-system for exploration boreholes, the Q-values are generally greater than 10 (see Table 1) and the rock mass quality is very good at the proposed cavern location (El. −20 m to −80 m). Thus, the excellent rock conditions are very suitable for the cavern construction. Underground water consists of Quaternary pore water and bedrock fissure water. Generally, the underground water flow direction is southwest and water table is at about 13.33–38.31 m below ground surface, which has an annual variation of 1–3 m. The type of hydrochemistry is sodium bicarbonate calcium chloride. The PH value is between 6.68 and 9.15. Water salinity is 138.14–232.37 mg/L. The water quality is good. The results of comprehensive hydraulic pressure tests show that the permeability coefficient is mainly less than 1.00 × 10−10 m/s, and ranging from 1.55 × 10−9 to 3.50 × 10−7 m/s (the mean value is 5.83 × 10−8 m/s) at fractured zones. The in-situ stress measurements indicate that in the buried depth of caverns the maximum principal stress is 6.19–11.5 MPa with a direction N71.7°-78.5 °E; the middle principal stress is 3.63–9.02 MPa in the horizontal plane; and the minimum principal stress is about 1.81–3.61 MPa in the vertical direction.

2. Project overview 2.1. Design and construction The Jinzhou national oil storage project is located in Liaoning Province, the northeast area of China. The geomorphic unit belongs to a hilly region and the elevation of the ground level ranges from 12.70 m to 42.83 m ACD (where ACD is the abbreviation of Admiralty Chart Datum). The design storage capacity of the project is 3 × 106 m3 for crude oil. Fig. 1 shows the layout of the underground facility, which is located more than 100 m below the ground surface. The underground facility is composed of four groups of storage caverns (each group consists of two connected caverns), four inlet oil shafts with a diameter of 3 m, four outlet oil shafts with a diameter of 6 m, two access tunnels and an artificial water curtain system. The eight storage caverns, namely 1N-4N and 1S-4S, are parallel to one another, aligned in the East-West direction. Two storage caverns for each group are connected by horizontal tunnels in order to guarantee the same liquid level. During the operation stage, the connection tunnels between the groups will be separated by concrete sealing plugs and thus each group can

3. Field and laboratory studies The field and laboratory studies include field observation, boreholes and discontinuity survey and laboratory testing. The study region for the field investigation is the cavern 4N and three sites (the mileage locations: 4+05-4+30 m for site #1, 3+20-3+30 m for site #2 and 0+70-1+50 m for site #3) were selected according to the variations of discontinuities (see Fig. 3). ShapeMetrix3D photogrammetric measurement system (Liu et al., 2017) was adopted for the discontinuity survey of three sites. The discontinuity information was obtained from the wall surface exposures. Fig. 4 shows the identification effect for field discontinuities of site #1 with the size 9.5 m × 3.8 m, showing three discontinuity sets. The statistical data of discontinuity parameters obtained from all sites are given in Table 2. 37

Tunnelling and Underground Space Technology 71 (2018) 36–46

J. Liu et al.

Fig. 1. Layout of the underground oil storage facility.

Water curtain tunnel

Shaft

Ground

Horizontal water curtain hole -32

Access tunnel 26.25

57

57

77.50

77.50

57

-80

Storage cavern

-97

57

24

Vertical water curtain hole

-56

10 24

-31

57

-90

26.25

Fig. 2. Cross-section view of underground oil storage facility (Lin et al., 2016).

parameters: uniaxial compressive strength (UCS) of intact rock material, rock quality designation (RQD), spacing of discontinuities, condition of discontinuities, groundwater conditions and orientation of discontinuities. Each parameter is given a rating of importance for a particular situation. The six input parameters are summed up to yield RMR rating results. On the basis of RMR rating, the rock mass is sorted into five classes: very good (100–81), good (80–61), fair (60–41), poor (4–21), and very poor (< 20). The rock masses of three sites in cavern 4N were classified according to the 1989 version of RMR and the results were given in Table 4. The Q-system was developed in the Norwegian Geotechnical Institute (NGI) by Barton et al. (1974) based on about 200 case histories of tunnels and caverns. The Q-system was updated on several occasions and it is now based on 1260 case records (Grimstad and Barton, 1993). Barton (2002) compiled the system again and made some changes on support recommendations to reflect the increasing use of steel fibre reinforced shotcrete (SFR) in underground excavation support. The quality of rock masses is described using six parameters which are RQD, joint set number (Jn), joint roughness (Jr), joint alteration (Ja), joint water reduction factor (Jw) and stress reduction factor (SRF) and Qrating is derived from the following expression by combining these six parameters:

Table 1 Percentages of rock masses of different classes. Class number

I

II

III

IV

V

Proportion (%)

40.51

28.76

14.00

11.31

5.41

The lithology of these sites is unitary, which is mainly mediumgrained granite. The physical and mechanical properties of the rocks were determined from laboratory testing on intact rock samples following ISRM (1981) recommended methods and the test results are given in Table 3. 4. Empirical analyses Rock mass classification systems are usually employed to assess the quality of rock masses for preliminary support design and also to determine the rock mass properties. For a preliminary cavern design, at least two classification systems should be applied. In this study, the most widely used rock mass classification systems including RMR, Q and GSI were utilized for three sites in cavern 4N. 4.1. Rock mass classification systems

Q= The RMR classification system was initially developed by Bieniawski (1973) on the basis of shallow tunnels in sedimentary rocks. It was modified several times and at present the latest 1989 version of RMR ratings (Bieniawski, 1989) is widely used. It employs six

RQD Jr Jw Jn Ja SRF

(1)

The three quotients in Eq. (1) represent the block size, the interblock frictional shear strength and the active stress, respectively. The Qvalue varies from 0.001 for exceptionally poor quality squeezing38

Tunnelling and Underground Space Technology 71 (2018) 36–46

J. Liu et al.

Elevation m W 46 1

6 -14

2

r4 r3 r2

4

r4 3 r3 r2 r1

r4

r4

r4 r3 r2

r2

E

r3 r2

r3

r1

r1

5 6

-34 -54

ZK10 20.21

ZK24 20.76

26

ZK40 41.64 r4 r3r2 r1

ZK36 30.55

ZK8 25.58

r1 r1

-74 Site #1

-94

Site #2 200m

-114

Legend Residual soil r4 r3 r2 r1

Site #3

Granite

Aplite granite

Diabase

Completely weathered strong weathered moderately weathered slightly weathered

2

Formation number

100m

0m

Amphibole diorite ZK11 Borehole 16.79 porphyrite Storage cavern

Fig. 3. Geological section of cavern 4N.

suggested by RMR can be applied in this project to some extent. Palmstrom and Stille (2007) pointed out the recommendations for rock support by RMR are somewhat out of date for modern tunneling. It should be noted that RMR support suggestions have not had a major revision since 1973. In many engineering applications, SFR may be considered in place of wire mesh and shotcrete. In this oil storage cavern project, the thickness of SRF was not defined independently based on the RMR system suggestions. The Q-values are combined with the size of the cavern in a support chart (see Fig. 5). The Q-index is related to the support requirements of underground excavations with the equivalent dimension De of the excavation. This dimension, which is a function of the size and purpose of the excavation, is obtained by:

ground to 1000 for exceptionally good quality rock and the quality is divided into nine categories. A stress-free form QN was defined by Goel et al. (1995), which is given in Eq. (2):

QN =

RQD Jr Jw Jn Ja

(2)

Barton (2002) defined a new parameter Qc to improve correlation among the engineering parameters:

Qc = Q

σci 100

(3)

where σci is the strength of intact rock in MPa. The GSI was developed by Hoek et al. (1995) as an input parameter for the Hoek-Brown criterion to estimate the reduction in rock mass strength for different geological conditions. This classification is simple and fast and it is based on the appearance of rock mass and its structure. The GSI values can be obtained from the quantitative GSI chart proposed by Marinos and Hoek (2000). The evaluations of three sites by using Q and GSI systems are summarized in Table 5. It can be found from Tables 4 and 5 that the results of Q-system is more conservative than those of RMR system though their results are very similar.

De =

Excavation span, diameter or height (m) Excavation sup portratio (ESR)

(4)

ESR is called the excavation support ratio. It is related to the degree of safety influencing on the support system and to the intended use of the excavation. It ranges between 0.8 and 5 and the suggested values have been given in a table by Barton et al. (1974). For the storage caverns, ESR is defined as 1. The De, plotted against the Q-index, is used to define a number of support categories in a chart (Fig. 5) published by Grimstad and Barton (1993). The bolt length Lb can be estimated in terms of excavation width B or height H for roofs and walls by the following equation proposed by Barton et al. (1974):

4.2. Support design by using RMR and Q classification systems

Lb = 2 + (0.15B or H /ESR)

GSI was developed only for the purpose of estimating rock mass strength in conjunction with the Hoek-Brown failure criterion. Bieniawski (1989) provided support guidelines for different rock mass classes in the RMR89 system. The suggested support schemes based on RMR for each site is shown in Table 6. Note that these RMR guidelines are for a 10 m wide horseshoe shaped, drilling and blasting tunnel under 25 MPa of vertical stress equivalent to a depth less than 900 m below ground surface. The design of rock bolt length is related to cavern size. It is not appropriate to determinate the bolt length based on RMR suggestions. Lowson and Bieniawski (2013) pointed out that bolt spacing is taken as a function of RMR only. Thus, the bolt spacing

(5)

Roof and wall support can be found by applying the span (19 m) and wall height (24 m) of the storage caverns, respectively. The bolt lengths for roofs and walls were calculated by Eq. (5) as 4.85 m and 5.6 m, respectively; therefore, the proposed bolt length design were 5 m for roofs and 6 m for walls. The estimated support categories based on Qsystem for each site have been shown in Fig. 5 and Table 6. According to the above analysis, the primary support systems proposed by Q-system were applied and RMR system was secondly considered and the further analysis was shown in Section 6. 39

Tunnelling and Underground Space Technology 71 (2018) 36–46

J. Liu et al.

Fig. 4. Identification effect for field discontinuities in site #1: (a) stereoscopic restructuring model, and (b) trace distribution.

(a)

Set 2

Set 1

Set 3

(b)

numerical modelling.

Table 2 Statistical results of discontinuity parameters. Site

Major orientations Dip direction/dip(°)

Set number

Spacing (m)

#1

269.49/72.20 245.13/29.87 87.01/68.21 268.11/32.85 125.11/66.17 249.74/46.08

3

1.5 1.3 2.0 2.7 1.0 2.3

#2 #3

2 1

5.1. Hoek-Brown parameters Hoek-Brown failure criterion for rock masses uses mb, s and a constants. Hoek et al. (2002) suggested the following equations for calculating these constants:

5. Estimation of rock mass properties The rock mass properties such as Hoek-Brown constants, UCS (σcmass) and uniaxial tensile strength of rock mass (σtmass), deformation modulus (Emass) and shear strength parameters were calculated by means of different empirical equations based on rock mass classification systems. The averages of these parameters are used as input data in the

GSI−100 ⎞ mb = m iexp ⎛ ⎝ 28−14D ⎠

(6)

GSI−100 ⎞ s = exp ⎛ ⎝ 9−3D ⎠

(7)

a=

1 1 GSI ⎞ 20 + ⎡exp ⎛− −exp ⎛− ⎞ ⎤ 2 6⎢ 15 ⎝ 3 ⎠⎥ ⎝ ⎠ ⎦ ⎣

(8)

where mi is the intact rock parameter. This constant mi can be obtained by triaxial testing of rock. Additionally, the approximate values can also be estimated by a table presented by Marinos and Hoek (2000). mi is 40

Tunnelling and Underground Space Technology 71 (2018) 36–46

J. Liu et al.

was assumed that blasting quality was excellent and so the value of D is consider zero. The calculated Hoek-Brown constants are listed in Table 7.

Table 3 Physical and mechanical properties of intact rocks. Parameters

Medium-grained granite

Unit weight γ (kN/m3) Uniaxial compressive strength σci (MPa) Young’s modulus E (GPa) Poisson ratio υ Porosity n (%) Tensile strength σt (MPa) Cohesion c (MPa) Internal friction angle φ (°)

26.20 122.74 47.50 0.22 0.40 7.70 18.63 62.60

5.2. Deformation modulus of rock masses The deformation modulus Emass of rock masses is an important parameter in numerical analysis. In-situ determination of Emass is very difficult due to being affected by the time and cost. So the empirical methods based on rock mass classification systems are used generally to determinate Emass. The values of deformation modulus of intact rocks from laboratory tests are given in Table 2. The several equations and the calculated rock mass deformation modulus values are presented in Tables 8 and 9, respectively.

Table 4 Evaluation results using RMR system. Parameters/ rating

Site #1

UCS(MPa) Rating RQD (%) Rating Spacing (m) Rating Joint conditions

122.74 12 33 8 1.3 15 Wall is moderately weathered; surfaces are smooth and apertures filled by soft materials (3–5 mm) 10 Dripping

57 13 1.0 15 Wall is covered by silt soil and minority surfaces are smooth 18 Wet

78 17 2.3 20 Wall is covered by silt soil and surfaces are rough

4 −2

7 −2

10 −2

47 III Fair

63 II Good

82 I Very good

Rating Groundwater conditions Rating Orientation rating Total rating Grade Rock mass quality

Site #2

Site #3

5.3. Strength of rock masses Different researchers have proposed different empirical equations to calculate the strength of rock mass (σcmass) based on rock mass classification systems. The most widely used equations are tabulated in Table 10. The calculated σcmass values are given in Table 11. In addition, the tensile strength σtmass of rock masses (Hoek et al., 2002) is: (30)

σtmass = −sσci/ mb

The calculated σtmass for three sites using Eq. (30) are 0.045, 0.174 and 0.730 MPa, respectively.

25 Damp

5.4. Equivalent Mohr-Coulomb strength parameters In rock engineering, many numerical model software packages use Mohr-Coulomb failure criterion. Therefore, it is necessary to determine equivalent friction angle and cohesive strength of rock masses. The equations for the friction angle φ′ and cohesive strength c′ were given by Hoek et al. (2002):

6αmb (s + mb σ3n )α − 1 ⎤ φ′ = sin−1 ⎡ ⎢ 2(1 + α )(2 + α ) + 6αmb (s + mb σ3n )α − 1 ⎦ ⎥ ⎣ Table 5 Evaluation results using Q and GSI systems. Site

Rating

Q

c′ = QN

Qc

Rock mass quality

GSI

(31)

σci [(1 + 2α ) s + (1−α ) mb σ3n](s + mb σ3n )α − 1 (1 + α )(2 + α ) 1 + (6αmb (s + mb σ3n )α − 1)/((1 + α )(2 + α )) (32)

#1 #2 #3

RQD

Jn

Jr

Ja

Jw

SRF

33 57 78

12 6 3

1 3 3

2 2 1

0.4 1 1

1 1 1

1.4 14.3 78

1.4 14.3 78

1.3 13.3 72.6

Poor Good Very good

where σ3n = σmax / σci . σ3max is the upper limit of confining stress over which the relationship between the Hoek-Brown and Mohr-Coulomb criteria is considered. For the tunnel engineering, the relationship between σ3max and σcmass is:

41 59 78

−0.94

σ3max σ = 0.47 ⎜⎛ cmass ⎟⎞ σcmass ⎝ γH ⎠

defined as 32 for medium-grained granite. D is a factor that depends upon the degree of disturbance to which rock masses have been subjected by blast damage and stress relaxation. It varies from 0 for undisturbed in-situ rock masses to 1 for very disturbed rock masses. For the underground oil storage caverns, the high blasting quality is required, which reduces surrounding rock damage as much as possible and thus prevents oil leak during the operation stage of the caverns. It

(33)

where σcmass is the rock mass strength; γ is the unit weight of rock masses and H is the depth of the tunnel below ground surface. When the horizontal stress is higher than the vertical stress, the horizontal stress value should be instead of γH. The calculated equivalent Mohr-Coulomb parameter values are given in Table 12.

Table 6 Support categories for selected sites. Classification system

Site #1

Site #2

Site #3

RMR

Value Support requirements Value Support requirements

63 Locally, bolts in crown 3 m long, spaced 2.5 m, with occasional wire mesh. Shotcrete 50 mm in crown where required. 14.3 Systematic bolts 5 m long in crown, 6 m long for wall, spaced 2.3–2.5 m. Shotcrete 40–50 mm

82 Generally, no support required except for occasional spot bolting

Q

47 Systematic bolts 4 m long, spaced 1.5–2 m in crown and walls with wire mesh in crown. Shotcrete 50–100 mm in crown and 30 mm in sides 1.4 Systematic bolts 5 m long in crown, 6 m long for wall, spaced 1.7–2.1 m. Fibre reinforced shotcrete 90–120 mm

41

78 Spot bolting spaced 3–4 m

Tunnelling and Underground Space Technology 71 (2018) 36–46

J. Liu et al.

Fig. 5. Estimated support categories based on the updated Q-support chart (Grimstad and Barton, 1993).

Table 7 Hoek-Brown parameter values.

Table 8 Empirical equations for calculation of Emass.

Site

mb

s

a

Authors

Equations (units GPa)

Number

#1 #2 #3

3.891 7.400 14.585

0.0014 0.0105 0.8677

0.5106 0.5030 0.5007

Bieniawski (1989) Serafim and Pereira (1983) Barton et al. (1980)

Emass = 2RMR−100, RMR > 50

(9) (10)

Nicholson and Bieniawski (1990) Mitri et al. (1994) Verman et al. (1997)

6. Numerical analysis In order to verify the results of the empirical analyses, FLAC3D (Fast Lagrange Analysis of Continua in 3D) (Zhu et al., 2010) was used to calculate the deformation and the thickness of the plastic zones around caverns due to the staged excavation and rock support considering the effect of the water curtain system. Since the ratio of length to crosssection dimension for underground caverns is large, a plane strain analysis was used. Two adjacent caverns with a distance of 38 m were selected to conduct the numerical analysis and the generated numerical model is shown in Fig. 6. The model area is 220 m wide and 180 m high. The rock mass is assumed “ideally” elastic-plastic material and the rock mass properties used in this analysis were obtained from the estimated values given in Section 5. The generalized Mohr-Coulomb strength criterion was used to identify elements undergoing yielding and plastic behavior in rock masses. The rock mass is simplified to an equivalent isotropic medium. The permeability of the rock mass is 5.83 × 10−8 m/s based on the water pressure tests and meanwhile the groundwater flow obeys the Darcy law. Vertical in-situ stress was generated to increase linearly with the

Emass = 10(RMR − 10)/40, RMR < 50

(11)

Emass = 25log10 Q, Q > 1 Emass =

Ei 100

(

0.0028RMR2 + 0.9exp

( )) RMR 22.8

(12)

Emass = 0.3(H )α10(RMR − 20)/38

(13) (14)

Read et al. (1999)

Emass = 0.1(RMR/10)3

(15)

Barton (2002)

Emass = 10Qc1/3

(16)

Hoek and Marinos (2000)

Emass = (1−D/2) σci/100 10(GSI − 10)/40, σci ⩽ 100

(17)

Emass = (1−D/2)10(GSI − 10)/40, σci > 100

(18)

Emass = Ei (0.5(1−cos(π RMR/100)))

Sonmez et al. (2004)

Emass = Ei (s a)0.4

Hoek and Diederichs (2006)

Emass = Ei ⎛0.02 + ⎝

1−D/2

⎞

(19)

1 + e ((60 + 15D − GSI)/11) ⎠

Notes: σci is uniaxial compressive strength of intact rock in MPa. Ei is deformation modulus of intact rock in GPa. α = 0.16−0.35, 0.16 for hard rocks and 0.35 for weak rocks. H is overburden in meters.

depth due to the overburden weight. Rock stress measurements indicate the maximum horizontal stress makes an angle of approximately 15° with the cavern axis. In this simulation, the maximum horizontal stress was assumed to be parallel to the cavern axis. The ratios of the maximum horizontal and the minimum horizontal stress to vertical stress were assumed 3.1 and 1.8, respectively. Thus, two principal in situ 42

Tunnelling and Underground Space Technology 71 (2018) 36–46

J. Liu et al.

Table 9 Estimated deformation modulus Emass of rock mass. Site

Equation number

#1 #2 #3

Average

(9)

(10)

(11)

(12)

(13)

(14)

(15)

(16)

(17)

(18)

(19)

8.4 – –

– 26.0 64.0

3.7 28.9 47.3

6.3 12.1 24.5

21.5 33.2 43.8

3.2 8.5 26.8

10.4 25.0 55.1

11.2 24.3 42.7

6.0 16.8 50.1

12.5 19.0 29.1

8.1 23.6 40.7

9.1 21.7 42.4

Table 10 Empirical equations for calculation of σcmass . Authors

Equations (units MPa)

Number

Hoek and Brown (1980)

σcmass/ σci =

(20)

Yudbir et al. (1983) Kalamaras and Bieniawski (1993) Goel (1994)

σcmass/ σci = exp(7.65((RMR−100)/100))

σcmass = 5.5γQ N1/3/ B 0.1

(23)

Bhasin and Grimstaad (1996) Singh (1997)

σcmass = (σci/100)7γQ1/3, Q > 10

(24)

Sheory (1997)

σcmass/ σci =

Aydan and Dalgiç (1998) Hoek et al. (2002) Barton (2002)

σcmass/ σci = RMR/(RMR + 6(100−RMR)) σcmass = σci s α

Ramamurthy (2004)

σcmass = σciexp((RMR−100)/25)

σcmass/ σci =

exp(RMR−100)/9 exp(RMR−100)/24

Water curtain tunnel

Water curtain borehole

(21) (22)

σcmass = 7γQ1/3, Q < 10 (25)

exp(RMR−100)/20

σcmass = 5γQ N1/3

Storage cavern

(26) (27) (28) (29)

Notes: γ is unit weight of rock mass in t/m3. B is the tunnel or cavern span in meters. Fig. 6. Numerical model of a group of caverns. Table 11 Estimated rock mass strength values. Site

#1 #2 #3

Equation number

the each stage of the excavations. This analysis includes both unsupported and supported cases. The proposed support systems are mainly based on Q-system and refer to the support guidelines based RMR system. The principal means of support elements consist of rock bolts (25 mm diameter, fully bonded) and fibre reinforced shotcrete. The properties of support elements, such as bolt length and spacing, and thickness of shotcrete are similar to those given in Table 6. 5 m length blot for roofs and 6 m length bolt for walls suggested by Q-system were used for all sites. The other support parameters consisting of bolt spacing and shotcrete thickness were designed as follows. For site #2, the bolt spacing and the thickness of shotcrete suggested by RMR (bolt spacing 2.5 m, shotcrete 50 mm) and Q-system (bolt spacing 2.3–2.5 m, shotcrete 40–50 mm) are almost identical. Therefore, the thickness of shotcrete 50 mm and bolting spacing 2.3 m can be selected easily. However, for site #1, the support parameters suggested by RMR and Q classification systems can vary within a relatively large range and thus to select an optimal support schemes is a little puzzled. If the excavation is required to be stable for a short amount of time, then it is advisable to try the less expensive support, namely bolts with 2 m spacing and 50 mm shotcrete in site #1. However, in this study, for this complex and important engineering whose service life exceeds 50 years, the caverns need to maintain longterm stability so that more substantial support, namely a conservative support pattern (bolts with 1.5 m spacing and 120 mm shotcrete) in site #1, should be installed. In the initial stages of design and construction, it is advisable to utilize the suggested conservative support. If the construction is progressing well with no stability problems and the support is performing very well, then it should be possible to gradually reduce the support requirements. The above analysis indicates that engineering judgement is also required in the application of rock mass classification to support design. Palmstrom and Broch (2006) pointed out that the suggested support systems by Q-system work best between 0.1 and 40 and outside this range, supplementary methods and calculations should be applied. For site #3, Q-value is greater than 40 and therefore the support suggestions by Q-system is not very reliable. In

Average

(20)

(21)

(22)

(23)

(24)

(25)

(26)

(27)

(28)

(29)

6.5 15.7 45.2

2.1 7.2 31.0

13.5 26.3 58.0

12.0 26.1 45.9

20.5 54.6 96.2

8.7 19.3 49.8

15.8 27.1 52.9

5.3 16.9 42.6

14.7 31.8 56.0

14.7 27.9 59.7

11.4 25.3 53.7

Table 12 Equivalent Mohr-Coulomb parameter values. Site

c′ (MPa)

φ′ (°)

#1 #2 #3

1.8 2.7 4.6

52.1 56.3 59.8

stresses are in the plane of the excavation and the third principal stress is out of the plane. The in-situ stress field was generated by applying stress on the right and back boundary faces of the model with the opposite faces fixed. The upper boundary is free. The stress-displacement conditions are changed into restraining all lateral boundary surfaces and the bottom face during the construction phase. The upper boundary of numerical model corresponds to the ground surface and the groundwater table is 21 m below the top boundary. The pore pressures varies with the depth. During the cavern excavation phase, atmosphere pressure was applied to the surface of the water curtain tunnels and caverns and the injection pressure of water curtain borehole was set as 0.41 MPa. The models include several stages following the proposed support and excavation steps. In the first stage, in-situ stress distribution was examined. In the second stage, water curtain tunnels were excavated and water curtain system was installed. In the following stages, two storage caverns were excavated simultaneously with top-heading and two benches (each 8 m high) and support systems were installed after 43

Tunnelling and Underground Space Technology 71 (2018) 36–46

J. Liu et al.

Unsupported case

Supported case

reduced to 1.50 m. For site #2, the displacement is slightly reduced to 3.83 mm and the thickness of plastic zones decreased to 1.50 m. The results indicate that the proposed support systems for site #1 and site #2 can guarantee the cavern stability. For site #3, the stability is good for unsupported cases and the values of maximum total displacement and maximum thickness of plastic zones are almost same and also very small before and after support installation. There is no any the plastic zones in the crown for both cases. For the walls, the support systems only cause slight change in the thickness of plastic zones. The results of the numerical modelling show that no significant failure is expected.

Top heading excavation

7. Discussion The wedge failures are the most common types in hard rock masses at relatively shallow depths. During the construction stage of storage caverns, the wedge failure is a main problem. The support schemes provided by RMR and Q-system combining the numerical modelling mainly aim at the general stability of the caverns. However, the wedge problem is locally concentrated and thus a supplementary reinforcement design to the unstable wedge is very necessary. The three dimensional wedges of pyramidal shape are formed by intersection of three or more joint sets on excavation boundary of tunnels (Panda et al., 2014). There are two discontinuity sets in site #2 and one set in site #3. Therefore, the wedge block analysis was conducted by Unwedge software (Rocscience, 2004) only for site #1 using the joint data in unsupported case. Fig. 9 shows the wedge blocks on the excavation surface. The bottom wedge is fully stable. The unstable blocks were not be found in the cavern crown. The factors of safety of two big blocks in the caver walls are greater than 10, which indicates that these blocks have no potential wedge failures. During the construction stage, wedge stability analysis needs to be used to identify the potential wedge failures. In addition, the method based on both rock mass classifications and numerical analysis in this paper aims at providing a preliminary support design in engineering practices. Typical rock support systems can be obtained based on empirical rock mass classifications. By further using the numerical method, the general stability and expected deformations can be calculated and thus the preliminary support design can be more reliable. Although this method can address geological and geotechnical uncertainties to a certain extent and is easy to use for engineers, it may be invalid in areas where there is damage due to blasting or a sudden change in the geology, or hidden weak zones and so on. Therefore, the measurements should be carried out during construction to verify rock mass behavior and check the validity of the proposed support system or to adapt the design of the support system. Generally, the conventional monitoring items include the displacement monitoring of the surrounding rock by the multi-point displacement meters and the rockbolt stress monitoring by rockbolt stress meters. However, it is difficult to deeply understand the stability mechanism of underground caverns based on these conventional monitoring data. Ma et al. (2016) presented an integrated CDEM and microseismic monitoring method to evaluate the cavern stability successfully. The main goals of the Ma et al. method are to understand the damage mechanisms of micro-crack evolution well and to identify and predict potential danger areas of surrounding rock masses for underground oil storage caverns. The microseismic monitoring technique can accurately and effectively monitor the micro-crack evolution in rock masses compared to the conventional monitoring methods. Meanwhile, CDEM can well simulate the progressive damage and failure process of surrounding rock masses, which is validated using microseismic monitoring data. As a complement of our study, the Ma et al. method can predict local instability and potential unstable zones during cavern construction, which can help to adjust the original support system or take other control measures when needed.

Bench-1 excavation

Bench-2 excavation Fig. 7. Numerical modelling stages for site #1.

addition, the use of the rock mass classification schemes should be updated and used in conjunction with site specific analysis. Particularly for underground water-sealed caverns, shotcrete’s primary function is not only to limit the deformation of rock masses but also to reduce groundwater seepage so as to lower drainage cost during the operation period. Sprayed concrete is introduced for all rock mass classes of all caverns. According to engineering practices, the layer of shotcrete is easily shirking and cracking when its thickness is less than 50 mm. Therefore, at least 50 mm thickness shotcrete are required for all caverns. Finally, the recommended support system is systematic bolts with 1.5 m spacing and 120 mm thick shotcrete for site #1; systematic bolts with 2.3 m spacing and 50 mm thick shotcrete were used for site #2; spot bolting and 50 mm thick shotcrete were used for site #3. Whether local spot bolting is required mainly depends on the wedge analysis in the discussion section. Only shotcrete was used for site #3 in this numerical modelling. The enlarged view of the left cavern excavation sequence for site #1 in unsupported and supported cases are given in Fig. 7. For unsupported and supported cases, the thickness of plastic zones and maximum total displacement at walls, roof and floor of the cavern for each site are shown in Fig. 8. The displacement at the sidewall is larger than that at roof. The maximum thickness of plastic zones and maximum total displacement are presented in Table 13. The maximum total displacements for unsupported cases in site #1 to #3 are 12.04, 4.99 and 2.24 mm, respectively, and occur at the left wall for left cavern. The maximum thicknesses of plastic zones for three sites are 2.75, 2.20 and 0.45 m, respectively, and develop at the wall and floor. After support installation, it can be seen from Fig. 8 and Table 13 that size of plastic zones and maximum total displacement decrease significantly compared with unsupported cases for site #1. The displacement is reduced to 8.38 mm and the maximum thickness of plastic zones is 44

Tunnelling and Underground Space Technology 71 (2018) 36–46

J. Liu et al.

Site

Unsupported case

Supported case

Fig. 8. Total displacement and extent of plastic zones for unsupported and supported cases.

4.45

#1 8.35

5.06 3.48

2.70

2.04

#2 4.99

3.07

3.83

2.36

2.06

1.60

1.38

#3

2.55

1.21

1.58

2.24

1.06

1.38 0.92

obtained from empirical methods, bolts and shotcrete as support elements were recommended. As verified in the numerical method, the recommended support systems were feasible. After support installation, area of plastic zones and maximum displacements significantly decreased. Therefore, it is suggested that a combination of both empirical and numerical method is used to design an optimum support system. Considering the wedge failures in hard rock masses, it is very necessary that the block analysis is conducted to identify the unstable wedges in the daily excavation management. However, the field measurements should also be carried out during the cavern construction.

Table 13 Maximum thickness of plastic zones and maximum total displacement obtained from FLAC3D. Site

#1 #2 #3

Maximum thickness of plastic zones (m)

Maximum total displacement (mm)

Unsupported

Supported

Unsupported

Supported

2.75 2.20 0.45

1.50 1.12 0.45

12.04 4.99 2.55

8.38 3.83 2.24

Acknowledgments This work was financially supported by the National Key Research and Development (Grant number: 2016YFC0600803), Natural Science Foundation of China (Grant numbers: 51474052), the Fundamental Research Funds for the Central Universities (Grant number: N150102001).

Fig. 9. Unsupported wedge blocks for site #1.

References

8. Conclusion

Aydan, Ö., Dalgiç, S., 1998. Prediction of deformation behaviour of 3-lanes bolu tunnels through squeezing rocks of North Anatolian Fault Zone (NAFZ). In: Regional Symposium on Sedimentary Rock Engineering, Taipei, pp. 228–233. Barton, N., Lien, R., Lunde, J., 1974. Engineering classification of rock masses for the design of tunnel support. Rock Mech. 6 (4), 189–239. Barton, N., 2002. Some new Q-value correlations to assist in site characterization and tunnel design. Int. J. Rock Mech. Min. Sci. 39, 185–216. Barton, N., Loset, F., Lien, R., Lunde, J., 1980. Application of the Q-system in design decisions concerning dimensions and appropriate support for underground installations. In: Proceedings of the International Conference on Subsurface Space. Pergamon Press, pp. 553–561.

An engineering geological assessment of rock masses has been carried out at three sites of the underground water-sealed oil storage project in Jinzhou, China. Based on the information collected in the field and laboratory, the RMR, Q and GSI classification systems were used for estimating the rock mass quality and determining the support systems for the cavern. The rock mass properties used as input parameters for numerical modelling were estimated from empirical equations related to the classification systems. According to the results 45

Tunnelling and Underground Space Technology 71 (2018) 36–46

J. Liu et al.

Nicholson, G.A., Bieniawski, Z.T., 1990. A nonlinear deformation modulus based on rock mass classification. Int. J. Min. Geol. Eng. 8, 181–202. Palmstrom, A., Stille, H., 2007. Ground behaviour and rock engineering tools for underground excavations. Tunn. Undergr. Space Technol. 22 (7), 363–376. Palmstrom, A., Broch, E., 2006. Use and misuse of rock mass classification systems with particular reference to the Q-system. Tunn. Undergr. Space Technol. 21, 575–593. Panda, M.K., Mohanty, S., Pingua, B.M.P., Mishra, A.K., 2014. Engineering geological and geotechnical investigations along the head race tunnel in Teesta Stage-III hydroelectric project. India. Eng. Geol. 181, 297–308. Qiao, L.P., Li, S.C., Wang, Z.C., Tian, H., Bi, L.P., 2016. Geotechnical monitoring on the stability of a pilot underground crude-oil storage facility during the construction phase in China. Measurement 82, 421–431. Ramamurthy, T., 2004. A geo-engineering classification for rocks and rock masses. Int. J. Rock Mech. Min. Sci. 41, 89–101. Rocscience, 2004. Unwedge User's Guide. Rocscience Inc., Toronto, Ontario, Canada. Read, S.A.L., Richards, L.R., Perrin, N.D., 1999. Applicability of the Hoek-Brown failure criterion to New Zealand greywacke rocks. In: Proceeding 9th International Society for Rock Mechanics Congress, Paris, vol. 2. pp. 655–660. Serafim, J.L., Pereira, J.P., 1983. Considerations of the geomechanics classification of Bieniawski. In: Proc. of Int. Symp. Eng. Geol. Underground Construction, LNEC. Lisbon. 1, II. 33–II. 42. Sheory, P.R., 1997. Empirical Rock Failure Criterion. Oxford IBH Publishing Co. and A.A. Balkema, Netherlands, pp. 176. Singh, S., 1997. Time-dependent Deformation Modulus of Rocks in Tunnels. (M.E. Thesis) Civil Engineering Department, University of Roorkee, India. Sonmez, H., Gokceoglu, C., Ulusay, R., 2004. Indirect determination of the modulus of deformation of rock masses based on the GSI system. Int. J. Rock Mech. Min. Sci. 41 (5), 849–857. Sun, J.P., Zhao, Z.Y., 2010. Effects of anisotropic permeability of fractured rock masses on underground oil storage caverns. Tunn. Undergr. Space Technol. 25 (5), 629–637. Sun, J.P., Zhao, Z.Y., Zhang, Y., 2011. Determination of three dimensional hydraulic conductivities using a combined analytical/neural network model. Tunn. Undergr. Space Technol. 26 (2), 310–319. Tezuka, M., Seoka, T., 2003. Latest technology of underground rock cavern excavation in Japan. Tunn. Undergr. Space Technol. 18, 127–144. Tian, Z.R., 2017. The analysis on the export and import of oil and gas products of China in 2016. Int. Petro. Econ. 25 (3), 15–25 (in Chinese). Verman, M., Singh, B., Viladkar, M.N., Jethwa, J.L., 1997. Effect of tunnel depth on modulus of deformation of rock mass. Rock Mech. Rock Eng. 30 (3), 121–127. Wang, Z.C., Li, S.C., Qiao, L.P., 2015. Design and test aspects of a water curtain system for underground oil storage caverns in China. Tunn. Undergr. Space Technol. 48, 20–34. Wang, Z.C., Li, S.C., Xue, Y.G., Qiu, D., Tian, H., 2013. Mechanical properties of surrounding rocks of large water-sealed underground oil storage caverns during construction process. Rock Soil Mech. 34 (1), 275–282 (in Chinese). Yu, C., Deng, S.C., Li, H.B., Li, J.C., Xia, X., 2013. The anisotropic seepage analysis of water-sealed underground oil storage caverns. Tunn. Undergr. Space Technol. 38, 26–37. Yudbir, F., Lemanza, W., Prinzel, F., 1983. An empirical failure criterion for rock masses. In: Proc. 5th Int. Cong. ISRM, Melbourne, vol. Balkema, Rotterdam, pp. 131–138. Zhu, W.S., Li, X.J., Zhang, Q.B., 2010. A study on sidewall displacement prediction and stability evaluations for large underground power station caverns. Int. J. Rock Mech. Min. Sci. 47 (7), 1055–1062. Zhuang, D.Y., Tang, C.A., Liang, Z.Z., Ma, K., Wang, S.Y., Liang, J.Z., 2017. Effects of excavation unloading on the energy-release patterns and stability of underground water-sealed oil storage caverns. Tunn. Undergr. Space Technol. 61, 122–133.

Benardos, A.G., Kaliampakos, D.C., 2005. Hydrocarbon storage in unlined rock caverns in Greek limestone. Tunn. Undergr. Space Technol. 20, 175–182. Bhasin, R., Grimstaad, E., 1996. The use of stress–strength relationship in the assessment of tunnel stability. Tunn. Undergr. Space Technol. 1 (1), 93–98. Bieniawski, Z.T., 1989. Engineering Rock Mass Classifications. Wiley, New York. Bieniawski, Z.T., 1973. Engineering classification of jointed rock masses. Trans. S. Afr. Inst. Civ. Engrs. 15, 335–344. Chen, H.M., Zhao, Z.Y., Sun, J.P., 2013. Coupled hydro-mechanical model for fractured rock masses using the discontinuous deformation analysis. Tunn. Undergr. Space Technol. 38, 506–516. Goel, R.K., Jethwa, J.L., Paithankar, A.G., 1995. Indian experiences with Q and RMR systems. Tunn. Undergr. Space Technol. 10 (1), 97–109. Goel, R.K., 1994. Correlations for Predicting Support Pressures and Closures in Tunnels. (PhD Dissertation) Nagpur University, Nagpur, India. Grimstad, E., Barton, N., 1993. Updating the Q-system for NMT. In: Int. Symp. Sprayed Concrete, Fagernes, Norway, pp. 46–66. Hoek, E., Kaiser, P.K., Bawden, W.F., 1995. Support of Underground Excavations in Hard Rock. Balkema, Rotterdam. Hoek, E., Carranza-Torres, C., Corkum, B., 2002. Hoek-Brown failure criterion-2002 Edition. In: Proceedings of the North American Rock Mechanics Society, Toronto, July 2002. Hoek, E., Marinos, P., 2000. Predicting tunnel squeezing. Tunnel sand Tunnelling International. Part 1 -November, Part 2-December. Hoek, E., Diederichs, M.S., 2006. Empirical estimation of rock mass modulus. Int. J. Rock Mech. Min. Sci. 43, 203–215. Hoek, E., Brown, E.T., 1980. Underground Excavation in Rock. Instn Min Metall, London. ISRM, 1981. Rock characterization, testing and monitoring. In: ISRM Suggested Method, International Society for Rock Mechanics. Pergamon Press, Oxford. Kalamaras, G.S., Bieniawski, Z.T., 1993. A rock mass strength concept for coal seams. In: 12th Conf. Ground Control in Mining, Morgan town, pp. 274–283. Lee, C.I., Song, J.J., 2003. Rock engineering in underground energy storage in Korea. Tunn. Undergr. Space Technol. 18, 467–483. Lee, Y.N., Suh, Y.H., Kim, D.Y., Jue, K.S., 1997. Stress and deformation behavior of oil storage caverns during excavation. Int. J. Rock Mech. Min. Sci. 34 (3/4), 305–319. Li, S., Wang, Z., Ping, Y., Zhou, Y., Zhang, L., 2014. Discrete element analysis of hydromechanical behavior of a pilot underground crude oil storage facility in granite in China. Tunn. Undergr. Space Technol. 40, 75–84. Li, Y., Chen, Y.F., Zhang, G.J., Liu, Y., Zhou, C.B., 2017. A numerical procedure for modeling the seepage field of water-sealed underground oil and gas storage caverns. Tunn. Undergr. Space Technol. 66, 56–63. Lin, F., Ren, F., Luan, H., Ma, G., Chen, S., 2016. Effectiveness analysis of water-sealing for underground LPG storage. Tunn. Undergr. Space Technol. 51, 270–290. Liu, J., Zhao, X.D., Xu, Z.H., 2017. Identification of rock discontinuity sets based on a modified affinity propagation algorithm. Int. J. Rock Mech. Min. Sci. 94, 32–42. Lowson, A.R., Bieniawski, Z.T., 2013. Critical Assessment of RMR based tunnel design practices: a practical engineer’s approach. In: Rapid Excavation and Tunneling Conference, Washington D.C., June 2013. Ma, K., Tang, C.A., Wang, L.X., Tang, D.H., Zhuang, D.Y., Zhang, Q.B., Zhao, J., 2016. Stability analysis of underground oil storage caverns by an integrated numerical and microseismic monitoring approach. Tunn. Undergr. Space Technol. 54, 81–91. Marinos, P., Hoek, E., 2000. GSI-A geologically friendly tool for rock mass strength estimation. In: Proc. Geo Eng Conference, Melbourne, pp. 1422–1442. Mitri, H.S., Edrissi, R., Henning, J., 1994. Finite element modelling of cable-bolted stopes in hard rock ground mines. In: SME Annual Meeting. Albuquerque. New Mexico, pp. 94–116.

46