J. Cent. South Univ. Technol. (2011) 18: 1709-1718
DOI: 10.1007/s11771-011-0892-2
Comparison of vibrations induced by excavation of deep-buried cavern and open pit with method of bench blasting
LU Wen-bo(卢文波)1, 2, LI Peng(李鹏)1, 2, CHEN Ming(陈明)1, 2,
ZHOU Chuang-bing(周创兵)1, 2, SHU Da-qiang(舒大强)1, 2
1. State Key Laboratory of Water Resources and Hydropower Engineering, Wuhan University, Wuhan 430072, China;
2. Key Laboratory of Rock Mechanics in Hydraulic Structural Engineering, Ministry of Education,Wuhan University, Wuhan 430072, China
? Central South University Press and Springer-Verlag Berlin Heidelberg 2011
Abstract: The measured data of vibrations induced by excavation of deep-buried cavern and open pit with method of bench blasting were analyzed by Fourier Transform and Wavelet Transform, and the characteristics of vibrations induced under these two circumstances were studied. It is concluded that with the similar rock condition and drilling-blasting parameters, vibration induced by bench blasting in deep-buried cavern has a higher main frequency and more scattered energy distribution than that in open pit. The vibration induced by bench blasting in open pit is mainly originated from the blast load, while the vibration induced by bench blasting in deep-buried cavern is the superposition of vibrations induced by blast load and transient release of in-situ stress. The vibration induced by transient release of in-situ stress increases with the stress level.
Key words: deep-buried cavern; open pit; blasting vibration; power spectra; time-energy density; in-situ stress
1 Introduction
Blasting vibration is an inevitable adverse effect during blasting excavation of rock. The vibrations induced by blasting excavation of open pit and underground cavern under low-level in-situ stress are generally the elastic vibrations caused by the effect of blast load in rock mass. However, during the blasting excavation of deep-buried cavern with high-level in-situ stress, rock stress on the excavation contour will release to zero transiently, which will inevitably cause intensive adjustments of stresses in the surrounding rock mass and induce transient unloading vibration. Thus, the vibration in the surrounding rock mass induced by blasting excavation of deep-buried cavern differs greatly from the vibration induced in open pit blasting. The former includes vibration induced unloading of in-situ stress besides common blasting vibration.
Many researchers have studied the characteristics of blasting vibrations under different explosive sources. ROY [1] analyzed blasting vibrations induced by surface blasting and underground blasting and revealed the characteristic differences between the two cases. SINGH et al [2-3] compared single-hole blast signatures with multi-hole ones, and proposed guidelines on determining the effective charge weight in the blasts of iron ore and limestone mines in India. KAHRIMAN et al [4] investigated the environmental impacts of bench blasting at Hisarcik Boron open pit mine. LING and LI [5-6] applied wavelet packet analysis to study the energy distributions of blasting vibrations with different maximum charge amounts of shot per delay, delay time and distances from explosion. LI et al [7] analyzed the energy distribution characteristics of blast vibrations under different blasting sources and concluded that with similar blasthole depth and diameter, main frequency band of open pit blasting vibration is more inclined to low frequency and the energy proportion at 40 Hz for blast vibration of underground engineering is very small.
The works of researchers worldwide revealed that the effects of transient release of in-situ stress should be studied to further understand the vibration characteristics of deep-buried cavern blasting. ABUOV et al [8] indicated that remained rock mass near excavation work face could be damaged because of the quick unloading of in-situ stress during blasting. By theoretical analysis, CARTER and BOOKER [9] implied that vibration in the surrounding rock could be induced by the transient unloading of in-situ stress, and the intensity of vibration should increase with the unloading rate of in-situ stress. YI et al [10] also indicated that under medium-to-high in-situ stress, the excavation-induced unloading wave could be an important factor for the loosening of excavated rock mass, and under higher in-situ stress levels, the amplitudes of excavation load would be larger and induce more intensive vibration. LU et al [11] showed that under a certain level of in-situ stress, the vibration induced by transient release of in-situ stress would exceed the vibration induced by blast load and become the main part of total vibration in the surrounding rock.
Traditionally, Fourier Transform had been widely used to analyze blasting vibration. However, since Wavelet Transform was proposed in signal analysis field, it has been popularized and applied in many fields due to its advantages in analyzing non-stationary random signals [12]. It was also introduced into blasting vibration analysis [13]. The aim of this work is to combine the two methods to analyze the measured vibration data in open pit and deep-buried cavern excavated with method of bench blasting and to compare the vibration characteristics of the two cases.
2 Blasting condition and measured vibration data
2.1 Open pit excavation case
The case of open pit excavation was taken from ash storage base excavation of The Third Stage Project of Huaneng Fuzhou Electric Power Station. Huaneng Fuzhou Electric Power Station locates at the Choudong village of Changle in Fujian province on the south bank of Minjiang. The rock types of the foundation are moderately weathered to slightly weathered granites.
The drilling-blasting parameters adopted were as follows: blasthole diameter was 90 mm; bench height was 7-9 m; spacings between holes and between rows were 3.0 and 2.0 m, respectively; No.2 rock emulsion explosive was used. Blasting was divided into 10 delays with Nonel relay detonators. One blasthole was detonated each delay and MS7 Nonel detonators were used for delays, as shown in Fig.1(a).
2.2 Deep-buried cavern excavation case
The case of deep-buried cavern excavation was from the 4th layer excavation of Pubugou underground powerhouse. Pubugou Hydropower Station locates at the middle of Daduhe River in Changjiang river valley and its installed capacity is 3 300 MW. The underground chamber of this project is a complex underground space structure composed of crisscross chambers such as main power house, main transformer house, tailwater lock chambers, bus tunnels, travel tunnels, tailwater tunnels, and pressure diversion tunnels. It locates in the left bank of the river and thickness of rock mass above the chamber is 220-360 m. Tectonic stress and weight stress mainly constitute the in-situ stress field of the region. Its in-situ stress level is medium-to-high and the value is between 9.61 and 29.16 MPa. The rock mass of the region is mainly granite.
The drilling-blasting parameters were as follows: blasthole diameter was 90 mm; bench height was 7.9 m; spacings between holes and between rows were 2.1 and 2.0 m, respectively; No.2 rock emulsion explosive was used. Applying delay-in-hole detonation network, the blasting was divided into 8 delays. Four or five blastholes were detonated on each delay and odd series of Nonel detonators from MS1 to MS15 were used for the detonators inside blastholes, as shown in Fig.1(b). In blasting areas, in-situ stress parallel to longitudinal axis of main power house, σ1, was about 24 MPa. For the intensive unloading of in-situ stress in this direction caused by former excavation, residual in-situ stress was approximately 0-3 MPa. In-situ stress perpendicular to the longitudinal axis of main power house, σ2, was about 12 MPa.
Fig.1 Monitoring point arrangements and blasting designs: (a) Open pit case; (b) Deep-buried cavern case
2.3 Measured vibration data
Rock conditions and drilling-blasting parameters such as blasthole diameters and bench heights of the two cases were close. Monitoring point arrangements and initiation networks are shown in Fig.1. The typical measured data of open pit excavation case were taken from 1# and 2# points whose distances from blasting center were 31 m and 51 m, respectively. The two points were at the same side of the blasting area. Measured data of deep-buried cavern excavation case were from 3# and 9# points. The 3# point was at the backward of the blasting area and the 9# point was arranged at the side wall of main transformer house. Distances from blasting center of the two points were both nearly 41 m. Velocity-time history curves of measured vibrations are shown in Fig.2. Considering more influence of in-situ stress on the horizontal vibration component, only horizontal component data were presented.
3 Vibration data analysis
3.1 Vibration analysis methods
The Fourier Transform and Wavelet Transform were adopted in the vibration analysis. The Fourier Transform is a traditional signal analyzing method, and could attain complete frequency information of signals; The Wavelet Transform is appropriate for analyzing non-stationary random signals, and can properly detect the mutational points of signals.
3.1.1 Fourier Transform
The Fourier Transform is defined as
(1)
Its inverse transform is given by
(2)
where f(t) is a random signal and ω is the circular frequency.
Fig.2 Velocity-time curves of measured vibration signals: (a) 1# point of open pit case; (b) 2# point of open pit case; (c) 3# point of deep-buried carven case; (d) 9# point of deep-buried carven case
Amplitude spectrum and power spectrum are mostly applied in the spectral analysis of vibration signals. Generally, function |F(ω)| is the amplitude spectral function of f(t), and spectral density |F(ω)|2 is the power spectral function which describes power (or energy) of every unit interval in frequency scale. The power spectral analysis was adopted in this work.
After the vibration signal data were collected, utilizing the FFT toolbox functions of Matlab, power spectral analysis of the signals could be carried on by simple programming. In the process of power spectral analysis in part of time domain, the frequency spectral characteristics of vibration during specific time period could be investigated.
3.1.2 Wavelet Transform
Assuming ψ(t)=L2(R) (L2(R) represents a square integratable space, say, a finite energy signal-space), its Fourier Transform is 
satisfies the admissible condition:
(3)
ψ(t) is defined as basic wavelet or mother wavelet. After stretch and translation of the mother wavelet ψ(t), a wavelet serial could be given by
(4)
where a is the stretch factor (or the scale factor) and b is the translation factor.
For any signal f(t)
L2(R), its Wavelet Transform is [14]
(5)
where 
represents the inner product of f(t) and ψa,b, and 
is the conjugate function of 
From above, it is known that the Wavelet Transform wf (a,b) represents the amount of wavelet component ψa,b that the signal contains. The Wavelet Transform under specific scale a is equivalent to bandpass filtering of a signal: when a is small, the time window is narrow and the transform corresponds to a detailed observation of the signal with high-frequency wavelet; when a is large, the time window is broad and the transform corresponds to approximate observation of the signal with low- frequency wavelet. Thus, for high-frequency signals (corresponding to small-scale), the Wavelet Transform has a high time resolution and lower frequency resolution and for low-frequency signals (corresponding to high-scale), it has a high frequency resolution and lower time resolution.
Modulus maximum method and time-energy density analysis method based on the Wavelet Transform are mainly applied in the mutation detection of signals. The time-energy density analysis method was adopted in the analysis. Its basic theories are as follows.
According to Moyal inner product theorem and Eq.(5), the following equation could be given:
(6)
It shows that energy of a signal is proportional to modulus square of the Wavelet Transform coefficients. According to the conception of energy density, Eq.(6) yields [15]
(7)
where
(8)
In a sense, scale a corresponds to frequency ω. So, the energy distribution of all frequency bands varying with time b is given by Eq.(8), which is called time- energy density function. In the actual application, through changing integral upper and lower limits in Eq.(8) to make the integrating interval lying in specific frequency band of the signal, time-energy density distribution in this frequency band could be obtained.
LING and LI [15] determined the actual delay time of millisecond blasting with this method. Regarding one blast as a system, blast of every detonator is a process of energy input and will inevitably arouse the mutation of energy density in the system. So, the actual delay time of every detonator can be identified according to peak positions on time-energy density curves. In the same way, if single delay blasting vibration is originated from different excitation sources, arriving time of different induced vibration waves could be identified by the peak positions on the time-energy density curves.
3.2 Power spectral analysis of measured vibration signals
The power spectral analyzing results of waveforms in Fig.2 are shown in Fig.3, and the analyzing results of typical single delay vibration waveforms are given in Fig.4.
Fig.3 Power spectra of measured vibration signals: (a) 1# point of open pit case; (b) 2# point of open pit case; (c) 3# point of deep-buried carven case; (d) 9# point of deep-buried carven case
3.3 Time-energy density analysis of measured vibration signals
The selection of wavelet basis is an important issue in the application of wavelet analysis, since adopting different wavelet bases to analyze the same problem will lead to different outcomes. Daubechies (abbr. Db) wavelet series are compactly supported, smooth and approximately symmetrical [16]. Thus, they have been successfully applied to analyze the non-stationary signals including blasting vibration signals [13, 15]. Db8 is the mostly used wavelet basis in blasting vibration analysis and also adopted here.
Besides the selection of wavelet basis, the determination of scale a is also an important problem in time-energy density analysis. It mainly depends on the minimum working frequency of blasting vibration recorders and the frequency ranges of measured signals. The minimum working frequency of blasting vibration recorders in this case is 10 Hz, and the sampling frequency is 2 000 Hz. According to the Nyquist sampling theorem, the integral upper limit of scale a in Eq.(8) should be larger than 100.
Taking the scale lower limit as 1 and the upper limit as 125, the time-energy density distributions of single delay vibration signals from Fig.2 are computed through Eq.(8). The vibration waveforms of single delays and corresponding time-energy curves are shown in Fig.5 and Fig.6, respectively.
4 Comparisons of vibrations from two cases
4.1 Comparison of frequency spectra
Figure 3 shows that the energy of open pit excavation case mainly distributes under 50 Hz and its frequency band is narrow. However, as for deep-buried cavern excavation case, the energy distributes in the range from 30 Hz to 120 Hz and the frequency band is wider. Figure 4 illustrates that the power spectrum of every single delay vibration signal in open pit excavation case only has one dominant frequency, but for deep-buried cavern excavation case, the power spectrum has at least more than two dominant frequencies. This shows that the main frequency of vibration induced by the deep-buried cavern excavation is higher than that induced by the open pit excavation, and the energy distribution of the former is more scattered than the latter in the frequency domain.
Fig.4 Power spectra of single delay: (a) 1# point of open pit case (3rd delay); (b) 2# point of open pit case (3rd delay); (c) 1# point of open pit case (4th delay); (d) 2# point of open pit case (4th delay); (e) 3# point of deep-buried carven case (MS9 delay); (f) 9# point of deep-buried carven case (MS9 delay); (g) 3# point of deep-buried carven case (MS11 delay); (h) 9# point of deep-buried carven case (MS11 delay)
Fig.5 Measured vibration waveforms of single delay: (a) 1# point of open pit case (3rd delay); (b) 2# point of open pit case (3rd delay); (c) 1# point of open pit case (4th delay); (d) 2# point of open pit case (4th delay); (e) 3# point of deep-buried carven case (MS9 delay); (f) 9# point of deep-buried carven case (MS9 delay); (g) 3# point of deep-buried carven case (MS11 delay); (h) 9# point of deep-buried carven case (MS11 delay)
Fig.6 Corresponding time-energy curves: (a) 1# point of open pit case (3rd delay); (b) 2# point of open pit case (3rd delay); (c) 1# point of open pit case (4th delay); (d) 2# point of open pit case (4th delay); (e) 3# point of deep-buried carven case (MS9 delay); (f) 9# point of deep-buried carven case (MS9 delay); (g) 3# point of deep-buried carven case (MS11 delay); (h) 9# point of deep-buried carven case (MS11 delay)
4.2 Comparison of time-energy density curves
From Figs.6(a)-6(d), it is known that the time-energy density curves of single delay vibrations from open pit excavation case are mainly composed of three to four larger peaks, which are centralized and constitute a peak group. For the open pit excavation case, one hole blasts each delay, and the in-situ stress of rock mass is approximately equal to zero. So, the induced vibration comes from the effect of blast load. This means that the peak group corresponds to an excitation source. From Figs.6(e)-6(h), it can be seen that the time-energy curves of single delay vibrations chosen from deep- buried cavern excavation case are clearly divided into two peak groups, corresponding to two excitation sources. According to the detonation networks and the in-situ stress condition, the two excitation sources might be caused by the delay errors of detonators or the blast load and transient release of in-situ stress.
In order to judge whether the two excitation sources are caused by the delay time errors of detonators, multi-delay blasting vibration waveforms on different delays (time intervals between delays are the same) are simulated by the linear superposition of wavelet taken from the single-hole blasting vibration waveform in Fig.5(a). And then, power spectral analyses of the simulated waveforms are carried on. The results are shown in Fig.7.
By comparing Fig.7 and Fig.4(a), it is known that the power spectra of simulated waveforms and the single hole blasting waveform are similar and each has one dominant frequency. Moreover, the dominant frequency values of four simulated models are close. However, for the deep-buried cavern excavation case, the power spectrum of every single delay vibration signal has at least more than two dominant frequencies. This suggests that the two excitation sources are not caused by the delay errors of detonators but the blast load and transient release of in-situ stress.
4.3 Comparison of vibrations induced by transient release of in-situ stress
From the in-situ stress distribution of the deep-buried cavern excavation case, it is known that in-situ stress in the direction from 9# point to blasting area is higher than that of 3# point on MS9 delay. By comparing Fig.4(e) and Fig.4(f), the power spectrum of signal measured at 3# point on MS9 delay has two dominant frequencies but the peaks corresponding to the two dominant frequencies are still overlapped. However, for 9# point on MS9 delay, the two dominant frequencies are clearly separated and the power spectrum is more distinct from that of the single delay blasting vibration signals in Figs.4(a)-4(d) (not affected by the in-situ stress). This shows that the vibration induced by transient release of in-situ stress at 9# point on MS9 delay is more significant than that at 3# point.
Fig.7 Power spectra of simulated multi-delay blasting vibrations on different delays: (a) Two delays (50 ms delay); (b) Two delays (100 ms delay); (c) Three delays (50 ms delay); (d) Three delays (100 ms delay)
According to the blasting initiation network and monitoring point location in Fig.1, as MS9 blasts ahead of MS11, the in-situ stress along the direction of the row that MS9 blastholes locate has been released. Thus, the in-situ stress level of this direction on MS11 delay is much lower than that on MS9 delay. Figure 4(h) indicates that the peaks corresponding to the two dominant frequencies of blasting vibration measured at 9# point on MS11 delay are obviously overlapped. As analyzed above, this means that the transient release of in-situ stress has less effect on MS11 delay. So, it can be concluded that the vibration induced by the transient release of in-situ stress increases with the stress level.
5 Conclusions
1) With the similar rock condition and drilling- blasting parameters, the vibration energy induced by deep-buried cavern excavation with method of bench blasting distributes mainly in high frequency band, while that induced by open pit excavation distributes mostly in lower frequency band. And energy distribution of the former is more scattered than the latter in the frequency domain.
2) The vibration induced by open pit excavation with method of bench blasting mainly comes from blast load. However, the vibration induced by deep-buried cavern excavation includes the vibration induced by the transient release of in-situ stress besides the vibration induced by blast load.
3) In the case of deep-buried cavern excavation, the vibration induced by transient release of in-situ stress increases with the in-situ stress level.
References
[1] ROY P P. Characteristics of ground vibrations and structural response to surface and underground blasting [J]. Geotechnical and Geological Engineering, 1998, 16(2): 151-166.
[2] SINGH P K, SIRVEIYA A K, ROY M P, PRASAD A, MOHAPARRA T. A new approach in blast vibration analysis and prediction at iron ore mines [J]. Mining Technology: IMM Transactions section A, 2005, 114(4): A209-218.
[3] SINGH P K, SIRVEIYA A K, BABU K N, ROY M P, SINGH C V. Evolution of effective charge weight per delay for prediction of ground vibrations generated from blasting in a limestone mine [J]. International Journal of Mining, Reclamation and Environment, 2006, 20(1): 4–19.
[4] KAHRIMAN A, OZER U, AKSOV M, KARADOGAN A, TUNCER G. Environmental impacts of bench blasting at Hisarcik Boron open pit mine in Turkey [J]. Environmental Geology, 2006, 50(7): 1015-1023.
[5] LING Tong-hua, LI Xi-bing. Features of energy distribution for blast vibration signals based on wavelet packet decomposition [J]. Journal of Central South University of Technology, 2005, 12(S1): 135-140
[6] LING Tong-hua, LI Xi-bing. Influence of maximum decking charge on intensity of blasting vibration [J]. Journal of Central South University of Technology, 2006, 13(3): 286-289.
[7] LI Hong-tao, YANG Xing-guo，SHU Da-qiang, LU Wen-bo, GAO Xing-ji. Study on energy distribution characteristics of seismic waves induced by different forms of blasting resource [J]. Journal of Sichuan University, 2010, 42(1): 30-34. (in Chinese)
[8] ABUOV M G, AITALIEV S M, ERMEKOV T M, ZHANBYRBAEV N B, KAYUPOV M A. Studies of the effect of dynamic processes during explosive break-out upon the roof of mining excavations [J]. Soviet Mining Science, 1989, 24(6): 581-590.
[9] CARTER J P, BOOKER J R. Sudden excavation of a long circular tunnel in elastic ground [J]. International Journal of Rock Mechanics and Mining Sciences and Geomechanics Abstracts, 1990, 27(2): 129-132.
[10] YI Chang-ping LU Wen-bo, XU Hong-tao, ZHOU Chuang-bing. The dynamic unloading effect study of initial stress field in excavation process of large-scale rock mass structure [J]. Journal of Rock Mechanics and Engineering, 2005, 24(s1): 4750-4754. (in Chinese)
[11] LU Wen-bo, CHEN Ming, YAN Peng, ZHOU Chuang-bing. Study on the characteristics of vibration induced by tunnel excavation under high in-situ stress [J]. Journal of Rock Mechanics and Engineering, 2007, 26(s1): 3329-3334. (in Chinese)
[12] NEWLAND D E. Wavelet analysis of vibration: Part 2. Wavelet maps [J]. Journal of Vibration and Acoustics, 1994, 116(4): 417-425.
[13] HE Jun, YU Ya-lun, LIANG Wen-ji. Wavelet analysis for blasting seismic signals [J]. Chinese Journal of Geotechnical Engineering, 1998, 20(1): 47-50. (in Chinese)
[14] CHUI C K. An introduction to wavelets [M]. New York: Academic Press Incorporated, 1992: 49-176.
[15] LING Tong-hua, LI Xi-bing. Time-energy analysis based on wavelet transform for identifying real delay time in millisecond blasting [J]. Journal of Rock Mechanics and Engineering, 2004, 23(13): 2260- 2270. (in Chinese)
[16] DAUBECHIES I. Wavelet transform, time-frequency localization and signal analysis [J]. IEEE Transactions on Information Theory, 1990, 36(5): 961-1005.
(Edited by YANG Bing)
Foundation item: Project(2010CB732003) supported by the National Basic Research Program of China; Projects(50725931, 50779050 and 50909077) supported by the National Natural Science Foundation of China
Received date: 2010-09-03; Accepted date: 2011-02-11
Corresponding author: LU Wen-bo, Professor, PhD; Tel: +86-27-68772221; E-mail: wblu@whu.edu.cn
