Research on optimized excavation approach applied in long-span underground cavern

Abstract

Long-span underground cavern excavation has been receiving wide application. This paper aims to give a theoretical study on the long-span underground cavern excavation as well as the damage zone of rocks subjected to various modes of blasting. Analysis is performed based on cylindrical wave theory with a start from blasting vibration and rock damage study, and then an optimized excavation method is proposed. Moreover, ultrasonic tests demonstrate validation of damage zone. It shows the proposed excavation approach is applicable.

Keywords

Long-span underground cavern blasting vibration rock damage zone excavation method ultrasonic test

1 Introduction

The long-span underground caverns are frequently seen in current infrastructure with the development of technology and become a forefront for geotechnical engineering. The applications include large-scale powerhouses in hydraulic engineering, road tunnels, railway tunnels and underground aircraft hangar etc.

Construction safety and schedule in a larger scale of excavation can be significantly affected by the factors including footage driving cycle, quality of cavern wall, blasting fragmentation and holes utilization ratio. Damages are inevitable on the remaining rocks outside the excavation contour, no matter what sort of blasting modes are adopted during underground cavern excavation. It is important to determine the damage zone so as to provide strong criterion for stability analysis and reinforcement on rock structure. In engineering practice, a post-measuring was previously carried out to find the damage scale. Damage mechanics, which was introduced in the past 20 years, has led to a wide use of peak particle velocity (PPV). Lu and Hustrulid [1] studied the designed method for blasting excavation associated with the contour surface near rock slope. Gao [2] conducted the field test and studied damages on slopes subjected to the presplitting during the excavation of a strip copper mine. Zeng [3] evaluated the result using current model for blast-induced rock damage calculation, and compared with the damage zone measured in a diversion tunnel excavation located in Laxiwa hydropower station. However, a theoretical study on long span underground cavern is still a challenge to catch up with the engineering application of explosion-induced damage zone. Moreover, an optimized excavation approach is needed based on a better understanding of damage.

In this paper, a theoretical study is given on long-span underground cavern excavation as well as the damage zone of rocks subjected to various modes of blasting. Analysis is first performed based on cylindrical wave theory with a start from blasting vibration and rock damage study, and then an optimized excavation method is proposed.

2 Ppv used as the experience criterion to estimate blast-induced damages

A design method for blasting excavation, based on PPV induced by shock wave, was proposed by Holmberg and Person [4] and then received wide attention. The key points include: (1) find the PPV distribution in remained rock around boreholes in which each particle’s PPV is gotten by overlaying all the individual blasting charges’ effects; (2) conduct the necessary field tests to capture the attenuation velocity field and thus find the different degrees of damage effects over the positions corresponding to PPV; (3) use the obtained PPV as an experience criterion for damage calculation and one can thus predict how far the damages can spread based on comparison to the actual PPV; (4) on the other hand, if a maximum damage zone is determined, it becomes possible to find the relevant PPV distribution. Then the blasting mode and the maximum explosive quantity in a sound, which is a significant parameter of design for drilling pattern, will be discussed to optimize excavation procedure.

It’s important to specify PPV distribution around the rock with explosive materials as that is the premise for Holmberg-Persson design method. Therefore, the PPV experience criterion determined properly is the prerequisite of numerous studies that have been performed. Generally, the commonly used PPV experience criterion was proposed by Bauer-Calder [5], Mojitabai-Beattie [6] and Savely [7]. For the hard rock, Holmberg-Persson set an upper-limit of 70–100 cm/s for PPV.

With a comprehensive consideration of PPV experience criterion for explosion-induced damages proposed by Bauer-Calder, Mojitabai-Beattie and Savely and that for hard rock introduced by Holmber and Persson, some upper-limits for PPV can be estimated. In generation, 45–100 cm/s are used for slightly-weathered or complete, middle-coarse-grained, blocky-structure granite. For example, 45 cm/s is used for slightly-weathered rock and 100 cm/s is adopted for fresh rocks [8 –10].

3 Damage zones induced by various blasting patterns

Attenuation function for PPV is expressed as Sadov’s formula given in Equation (1) [11]: $V = K {(\frac{Q^{n}}{R})}^{α}$ (1) where Q is the explosive quantity in a sound (kg), R the distance form explosive center (m), V the peak particle velocity (cm/s), and K and α the coefficient prior to site condition and charge parameter. n = 1/3 is used in China and Russia from dimensional analysis according to the concentrated charge condition; n = 1/2, defined for column charge, is used in the U.S. and other western countries. If K and α are known after applied regression, one can determine V once Q and R are given.

When R reaches the limit, n values exerts little influence on V. Equation (1) cannot reflect the effect of n on V when n = 1/2 or n = 1/3 that depends on the factors such as explosive type, charge form, drilling type or lithological condition of surrounding rock. Equation (1) shows a satisfactory outcome in fitting the ground shaking attenuation in a middle-far distance from explosive source, while errors become significant when predicting PPV near the explosive. It is thus crucial to find an attenuation formula that can reflect the physical mechanism of explosion, specifically for those PPV near the explosive resource [12].

Lu and Hustrulid proposed the attenuation formula for PPV at excavation contour of near-explosion region in rock slope excavation, with Heelan solution analysis applied in the cylindrical wave theory, sub-wave theory in long-cylindrical or short-cylindrical charge as in Equation (2). Field test and numerical study indicate the effectiveness of equation. $V = {kV}_{0} {(\frac{b}{R})}^{β}$ (2) where k is the coefficient associated with holes number adopted in a single explosion. k ≈ l in near-explosive region, whereas k equals to the hole number when considering remote-explosive region; b is the diameter of hole, β the attenuation exponent, and [PPV] the safety criterion of PPV in remained rock.

The blast-induced damage zone of remained rock is thus given in Equation (3): $[R] = b {(\frac{{kV}_{0}}{[PPV]})}^{\frac{1}{β}}$ (3)

The PPV on hole wall, V₀, is expressed as [13]: $V_{0} = \frac{P_{o}}{ρ C_{p}}$ (4) where ρ is the rock density, C_P the longitudinal wave velocity and P₀ the initial detonation air pressure in holes, that is equal to P_e(mean explosive pressure) when coupling charged. P_e can be obtained using detonation wave theory [14]: $P_{e} = \frac{ρ_{e} D^{2}}{2 (γ + 1)}$ (5) where ρ_e is the charge density, D the detonation velocity, and γ the isentropic exponent of charge.

Detonation gas expands in the holes when coupling charged. Assume that the gas is polytropic, the equation of state is [15]: $P = A ρ^{ν_{0}}$ (6) where ρ is the density of detonation gas, A is a constant, and ν₀ is the isentropic exponent.

Expansion of detonation gas would follow 2 phases including P ≥ P_k and P < P_k. If the decoupling coefficient $\frac{a}{b}$ (where b is drilling diameter, a is the charge diameter) is large, the initial mean pressure in hole P₀ can be easily given as [16]: $P_{0} = {(\frac{ρ_{e} D^{2}}{2 (γ + 1)})}^{\frac{ν_{0}}{γ}} P_{k}^{\frac{γ - ν_{0}}{γ}} {(\frac{a}{b})}^{2 ν_{0}}$ (7)

If the decoupling coefficient is small, the expansion could only experience P > P_k. Then P₀ is expressed as [16]: $P_{0} = \frac{ρ_{e} D^{2}}{2 (γ + 1)} {(\frac{a}{b})}^{2 γ}$ (8) where ν₀ = γ = 3.0 when P ≥ P_k; ν₀ = γ = 1.3 when P < P_k. P_k is the critical pressure of charge.

Therefore, V₀ on hole wall can be found with defined P₀ using Equation (4), then the damage zone could be determined by Equation (3). The damage zone equation in Equation (3) reflects the factors influencing PPV, e.g., explosive types, drilling size, charge structure and lithologic parameter.

Based on what was actually used in the underground excavation of power station including drilling size and explosive property, the parameters applied in the blasting design in this study are given as follows: (1) drilling diameter: 76 mm in benching blasting part and presplitting, 45 mm for protective layer excavation and smooth blasting; charge diameter: 60 mm for bench blasting, 32 mm for presplitting, 32 mm for protective layer and 25 mm for smooth blasting; (2) emulsion explosive was adopted, with ρ_e = 1100kg/m³, blasting speed D = 3800 m/s, P_k= 100MPa, γ= 3.0 when P≥100MPa; (3) slightly weathered rock: ρ= 2700 kg/m³, C_p= 4500 m/s, β= 1.3; almost fresh rock, ρ= 2700 kg/m³, C_p= 5000 m/s, β= 1.1.

Damage zones with respect to different types of blasting method are determined in compliance with its geological condition and drilling diameter adopted. The calculation should follow the following steps: (1) define the initial average press of detonation gas in the holes, and make sure whether it is coupling charged, if so, process with Equation (5) and use otherwise Equations (7 or 8) depending on its decoupling coefficient; (2) find peak particle velocity V₀ on holes wall; (3) PPV safety criterion for rocks weathered on different degree is thus determined with mentioned steps done. Combined with Equation (3) one can define the damage zones of this power station during the blasting excavation as seen in Table 1.

Table 1

The defined damage zone and its corresponding blasting pattern

Patterns of Blasting	Damage Zone of Rock(m)
	Slightly weathered	Fresh
Bench blasting	2.52	2.06
Presplitting	0.35	0.20
Small bench for shallow holes	0.92	0.69
Smooth blasting for small holes	0.36	0.23

The results show different damage zones for different blasting patterns adopted. The bench blasting applied in middle-deep holes cause zone to exceed 2 m, while in shallow holes the value is less than 1 m. Difference is marginal for smooth blasting and presplitting as both are of smaller value because the free face conditions were not considered.

4 Optimization for excavation sequence

Stress wave and fragments induced near the explosion region, which differs from the mechanism of middle-far earthquake, can impose dynamic damage effect on remained rocks. Explosion damage, as is justified, should be given precedence when considering response of explosion region. Optimization for excavation process should be thus based on controlling explosion damage.

Dynamic damages on rock surfaces are inevitable as rocks begin to fracture subjected to blasting load, such as new cracks appearing, extension of existing cracks and rise in hydraulic conductivity. To mitigate the problem, one can adopt the controlled blasting approach at base surfaces or slope contour. Underground caverns explosion should be performed adjacent to rock contour with an applied smooth blasting or presplitting.

Presplitting and smooth blasting are similar in cracking mechanism, hence each of them can create a smooth contour. Smooth blasting indicates that the advantages lie in less damage on surrounding rocks while drawback lies in inevitable damages on bearing rocks within the excavation region, whereas presplitting holds the opposite properties. Therefore, an optimized approach is needed, according to rock and construction conditions, to improve long-span underground cavern blasting.

Rocks are usually presplitted in a short distance near the side wall when conducting the long-span underground excavation. The formed cracks can isolate shocks and mitigate damages on side wall when the bench blasting is adopted. Clamping effect in underground construction is however not as obvious as in the open-air because the existence of ground stress weakens the isolation and crack resistance, making it hard to form a consistent presplitting crack. An appropriate excavation approach should hence not only avoid the drawbacks, but also take advantages of the two methods mentioned above. Presplitting combined with a lateral free face is quiet applicable in feeding those requirements.

Therefore, it is recommended to broach the grooves first, come to excavation of middle rocks and blast the protective layers. Broached grooves and excavation should be conducted within the same network and set another for protective layer blasting. The blasting excavation in middle rock follows the sequence: broached grooves ⟶ presplitting holes at protective layer boundary ⟶ blasting ladders for middle rock excavation as shown in Fig. 1. In this procedure, the formed broached grooves provide lateral free faces for presplitting where blasting energy can thus focus on. This approach overcomes the difficulty in crack formation due to high ground stress to serve the functions, involving shock isolation and cracks resistance.

Fig.1

The optimized sequence for underground excavation in lower layers.

In terms of the presplitting approach with lateral free faces, thickness of its reserved protective layers is a significant parameter. Thickness selection requires both bench blasting and presplitting applied in the middle rock cause no damages to the remained part on the side wall. Considering the safety allowance and requirement for construction surface under excavation, the value can be chosen as 3 m (see Table 1).

5 Comparative study for sound waves

Sound wave is a practical method to reflect blasting effect on rocks and an accessible way to verify the theoretical damage zone, even though it is an after-the-fact criterion and does not allow for the coupling effect of blast-induced damages as well as unloading-induced damages during the excavation. The test, which aimed to apprehend the damage zone and its peculiarity, was applied at the remained rock behind the sidewall of cross-sections at the third floor of a power station’s main house. The holes, set 10 m in depth and 90 mm in diameter, went through the protective layer and reached the rocks of sidewall. The classical curve of depth versus velocity is shown in Fig. 2.

Fig.2

Results for sound wave test at the third floor of main house before and after explosion.

Figure 2 shows comparison of sound wave velocity distribution along holes depth before or after explosion in section 0+17.5. It indicates a less-than-average velocity appeared at orifice of holes after explosion because excavation of protective layers can partly reduce the effects on the mechanical property of the remained rock surfaces caused by stress relax and blasting loading. It is difficult to determine the exact depth of affected rocks merely judging from the characteristic of curves, while wave velocity less than average value shows an affected depth of 0.2 m∼0.8 m among most sections in surrounding rocks. The fact indicates a reserving 3-meter-thick protective layer is useful in avoiding damages on sidewall induced by bench blasting. Moreover, this study provides an optimized excavation sequence for which the proposed blast-induced damage zone is proved to be applicable.

6 Conclusion

This study provides the damage zones subjected to various blasting modes during the long-span underground cavern excavation. The study is conducted using the attenuation equation for PPV (particle peak velocity) based on the wave theory, and PPV safety criterion applied in estimating the blast-induced damages on rocks. An optimized excavation approach is then proposed with the given damage zones. Ultrasonic tests demonstrate validation that feed the requirement of damage zone in near-explosion region, which can be a technical support for long-span underground excavation to process with efficiency and safety.

References

Wenbo

L.U.

and Hustrulid

, An improvement to the equation for the attenuation of the peak particle velocity, Engineering Blasting8(3) (2002), 1–4.

Gao

, Study on pre-splitting blasting on slope damage of strip mine, Copper Engineering112(6) (2011), 15–17.

Billi

, Salvini

and Storti

, The damage zone-fault core transition in carbonate rocks: Implications for fault growth, structure and permeability, Journal of Structural Geology25(11) (2003), 1779–1794.

Holmberg

and Persson

P.A.

, The Swedish approach to contour blasting, In: Proceedings of the 4th Conference on Explosive and Blasting Technique. [s. l.]:ISEE, 1978, pp. 113–127.

Bauer

and Calder

P.N.

, Open pit and blast seminar, Mining Engineering Department, Queens University, Course No. 63221. Kingston, Ontario, Canada, 1978.

Mojitabai

and Beatti

S.G.

, Empirical approach to prediction of damage in bench blasting, Trans Inst Min and Metall Sect A105 (1996), pp. A75–A80.

Savely

J.P.

, Designing a final blast to improve stability, Preprint No. 86-50, Presented at the SME Annual Meeting, New Orleans, Louisiana, USA, 1986.

and Hustrulid

, Optimization of the contour blasting during rock slope excavation at Three Gorges project,In: Elsworth

, Tinucci

, Heasley

ed., Proc 38th US Rock Mech Sypm, DC Rocks: Rock Mechanics in the National Interest[C] [s. l.]:[s. n.], (1), 2001, pp. 387–394.

Wang Xianyi

H.X.S.J.

, Discussion on safety evaluation of vibration effect by rock excavation, Coal Science And Technology7(31) (2003).

10.

Gang

, Analysis of the safety criteria for blasting vibration, Blasting2(21) (2004).

11.

Dianzhong

, Practical handbook for engineering blasting, Metallurgical Industry Press, 1999.

12.

Hustrulid

W.L.W.

, Some general design concepts regarding the control of blast-induced damage during rock slope excavation, In: Wang

ed., Proc 7th Rock Fragmentation by Blasting, 2002.

13.

Erxin

and Renshu

, Blasting Engineering, Xuzhou: China University of Mining and Technology Press, 1999.

14.

J.G.

, Qin

W.S.

and Li

Y.R.

, Experimental study of regularity of effect of bottom air column length on cut blasting, In: Theory and Practice of Energetic Materials, Beijing: Science Press, 2003, pp. 1059–1062.

15.

Jinshi

and Yang

, Study on rock blasting crack development direction and mechanics under simu-static pressure effects, Explosion and Shock Waves10(1) (1990), 76–83.

16.

Zong

, Lu

and Luo

, Theoretical study on axial decoupling coefficients of smooth blasting with air cushion charging construction, Chinese Journal of Rock Mechanics and Engineering24(6) (2005), 1047–1051.