Microseismic Signals Research Articles

Energy and Frequency Characteristics of Microseismic Particles in a Complex Geological Model

AbstractThis study aims to better understand the energy–frequency characteristics of microseismic wave propagation by monitoring particles in complex geology. Artificial seismic source experiments are conducted using a complex geological model, and variational mode decomposition method is applied to process the microseismic signals, obtain reconstructed microseismic signals and decomposition modes, and calculate the microseismic particle input energy and signal energy at each measurement point. Computations derived from the Sadoff empirical formula are used to evaluate the attenuation of the microseismic particle input energy with the propagation distance. The relationships between microseismic wave center frequency and energy and between the microseismic particle input energy and microseismic signal energy are elucidated. The results show that the microseismic particle input energy follows a power exponential decay trend with increasing propagation distance; the values and distributions of the microseismic particle input energy and signal energy conform to a Gaussian trend in the frequency domain; various geological structures and combinations thereof impose frequency modulation effects on the microseismic wave, where the frequency modulation interval depends on the microseismic waves across the different interfaces; and the linear slope describing the relationship between microseismic particle input energy and signal energy can reflect the microseismic wave propagation state.

Automatic arrival-time picking of P- and S-waves of microseismic events based on object detection and CNN

Automatic, quick and accurate picking of the arrival-times of both the P- and S-waves of microseismic (MS) events is essential to real-time data processing for MS monitoring. This study proposes a two-stage arrival-time picking method that involves extracting the region of interest of the waveform and P- and S-wave arrival-time prediction. The dataset collects 6107 MS signals with low and high signal-to-noise ratios to train and test the proposed method. The YOLO network is employed to extract the MS signal region to narrow the range of arrival-time picking and significantly improve the problem of serious interference caused by excessive negative samples in the MS signal. Then, the convolutional neural network is utilised further to precisely pick the arrival-times of both the P- and S-waves. The results showed that the proposed two-stage method is more accurate than the one-stage method in picking the arrival-time. The effects of different convolutional kernel sizes and fully connected layers on the two-stage model are also discussed. Furthermore, a new multi-index comprehensive evaluation method that considers both the accuracy and the size of the model is proposed to further evaluate the proposed model's practicality. This study presents a new method suitable for both P- and S-wave arrival-time picking to enable MS monitoring, which is valuable for various rock engineering applications.

First observation of paired microseismic signals during solution salt mining

To monitor the status of cavern roof instability induced by solution salt mining, we deployed a surface microseismic monitoring system for Dongxing salt mine, Dingyuan county, China. The microseismic monitoring system consists of 11 three-component geophones installed in shallow holes of 5 m deep. From 7-month continuous monitoring data, we have detected a novel type of signal that consists of two events. The first event has a long duration (>15 s) and generally has higher frequencies at the beginning. The 2nd event has a shorter duration of ∼3 s with lower frequencies. The two events are separated by ∼30–90 s in time. From May 2017 to November 2017, there are a total of 88 distinct paired signals with two events in sequence. We propose the first event corresponds to a series of fracturing processes on the cavern roof, and microseismic signals associated with different fractures are mixed and overlapped to produce a mixed signal with a longer duration. Due to a series of fracturing processes, some zones of cavern roof become instable and some rock debris may fall from the roof and collide on the cavern floor. This collision can produce a short duration signal, which corresponds to the 2nd event of the paired signal. The further analysis of polarities of the first arrivals for the 2nd event further proves its collision origin. The time interval between two events is related to the time for the falling rock debris through the brine, which is controlled by the cavern height and various physical properties of the rock debris and brine. Through the detailed analysis of paired signals, we can have a better understanding of the cavity development status for solution salt mining.

A high-resolution microseismic source location method based on contrast source algorithm

SUMMARY Passive source location estimation technology plays an extremely important role in microseismic monitoring. However, the energy of microseismic signals is generally weak and prone to be interrupted as noise, which will affect the detection and location accuracy of microseismic events. Besides, the existing wave-equation-based source location estimation methods are not suitable for strongly heterogeneous strata, such as faults, salt bodies, fractures filled with proppant and other fluids, etc. These complex geological structures often trigger scattered waves. Thus, based on scattering theory, a high-resolution microseismic source location estimation method with a new grouping imaging condition is proposed by implementing a contrast source numerical simulation method in the process of wavefield backpropagation. First, the contrast source algorithm is used to perform wavefield extrapolation by solving the Helmholtz equation in the frequency domain, which can provide more comprehensive, effective information to get high-precision locations. Moreover, an optimized grouping imaging condition is proposed, which can improve the imaging resolution and computational efficiency. The effectiveness of the method we propose is illustrated using numerical examples. Compared with traditional source location estimation methods, including with low signal-to-noise ratio or multisource location estimation, this method results in strong energy focusing achieved in an effecient manner.

Characteristics and Classification of Microseismic Signals in Heading Face of Coal Mine: Implication for Coal and Gas Outburst Warning

Characteristics and Classification of Microseismic Signals in Heading Face of Coal Mine: Implication for Coal and Gas Outburst Warning

Research on casing deformation mechanism and prevention measures based on micro-seismic signal distribution

Casing deformation is a severe problem in shale gas well development. The field data show that multiple deformations usually occur in the same casing well, and the distance is close, and the degree of deformation is different. In this paper, firstly, based on the distribution of micro-seismic event points in the Ning M well, the fracture group is inverted, and to establish the finite element numerical model to simulate the multi-stage fracturing, Then, the stress at each point of the casing is analyzed with the fracturing time, the circumferential stress at each point after fracturing and the axial stress of casing, and the displacement variation of casing in the directions of maximum, minimum and vertical principal stress are discussed. Finally, the effect of the intensive fracturing technology on controlling casing deformation is discussed. The results show that: a) Casing stress increases in different amplitude at a local point, circumferential direction, an axial direction, and regional stress difference is formed; b) Some local areas around the wellbore are subjected to twice or more re-fractures, and the corresponding rock properties (such as elastic modulus and Poisson's ratio) will change differently; c) Casing deformations with different directions and displacements appear in some reconstructed areas, showing “S” shape deformation, consistent with the actual deformation situation on site; d) The intensive fracturing technology can alleviate the non-uniformity of fracturing and reduce the non-uniform load of the casing, and effectively reduce the casing deformation rate after field application.

Detection and Characterization of Microseismic Events from Fiber-Optic DAS Data Using Deep Learning

Abstract Microseismic analysis is a valuable tool for fracture characterization in the Earth’s subsurface. Distributed acoustic sensing (DAS) fibers are deployed at depth inside wells, so they hold vast potential for high-resolution microseismic analysis. However, the accurate detection of microseismic signals in continuous DAS data is challenging and time consuming. We designed, trained, and deployed a deep learning model to detect microseismic events in DAS data automatically. We created a curated dataset of nearly 7000 manually selected events and an equal number of background noise examples. We optimized the deep learning model’s network architecture together with its training hyperparameters by Bayesian optimization. The trained model achieved an accuracy of 98.6% on our benchmark dataset and even detected low-amplitude events missed during manual labeling. Our methodology detected more than 100,000 events, allowing for a far more accurate and efficient reconstruction of spatiotemporal fracture development than would have been feasible by traditional methods.

An Effective Denoising Method Based on Cumulative Distribution Function Thresholding and its Application in the Microseismic Signal of a Metal Mine With High Sampling Rate (6 kHz)

Microseismic events can be used to analyze the risk of tunnel collapse, rock burst, and other mine hazards in space and time. In practice, the artificial activities and other signals at the mining site can seriously interfere with the microseismic waveforms, reducing the signal-to-noise ratio. In this study, we propose a denoising method based on the threshold of the cumulative distribution function (CDF) of the wavelet coefficients in the wavelet domain using synchrosqueezed continuous wavelet transform (SS-CWT). First, the ratio of microseismic signal variance between two adjacent time windows is used to determine the range of background noise. Then, the microseismic signal is transformed into a wavelet domain using SS-CWT, and the threshold of wavelet coefficients at each scale is estimated based on the cumulative distribution function (CDF) of background noise. At last, a post-processing step is applied by utilizing an amplitude smoothing function, to further suppress the noise. The proposed denoising method is tested by both synthetic and filed microseismic data recorded in a metal mine. The results show that the method is effective in denoising and can improve the SNR of mine microseismic data with a high sampling rate.

Surface microseismic data denoising based on sparse autoencoder and Kalman filter

Microseismic technology is widely used in unconventional oil and gas production. Microseismic noise reduction is of great significance for the identification of microseismic events, the location of seismic sources and the improvement of unconventional oil and gas production. In this paper, a denoising filter is proposed based on sparse autoencoder and Kalman filtering. Firstly, a sparse autoencoder is pre-trained to learn the feature of the microseismic data. Sparse autoencoding is a back-propagation neural network algorithm based on unsupervised learning, in which there are three layers: the input layer, the hidden layer and the output layer. The hidden layer is the spare, which makes the algorithm learn features better, represents samples in harsh environments and reduces dimensionality effectively. Besides, Kalman filter is used to deal with the uncertainty factors. Using a dataset of 600 surface microseismic synthesis traces and simulation noise. Sparse autoencoders and Kalman filtering are trained to suppress noise. The denoising filter based on sparse autoencoder and Kalman filter model obtains a higher signal noise ratio than the conventional model. The experiment results for the filtering of surface microseismic signals show the feasibility and effectiveness of the proposed method.

Mine-Microseismic-Signal Recognition Based on LMD–PNN Method

The effective recognition of microseismic signal is related to the accuracy of mine-dynamic-disaster precursor-information processing, which is a difficult method of microseismic-data processing. A mine-microseismic-signal-identification method based on LMD energy entropy and the probabilistic neural network (PNN) is proposed. First, the Local-Mean-Decomposition (LMD) method is used to decompose the mine microseismic signal. Considering the problem of vector redundancy, combined with the correlation-coefficient method, the energy entropy of the effective product-function component (PF) is extracted as the feature vector of mine-microseismic-signal classification. Furthermore, the probabilistic neural network (PNN) is used for learning and training, and the blasting-vibration signal and the coal–rock-mass-rupture signal are effectively identified. The test results show that the recognition accuracy of the PNN is up to 90%, the calculation time and classification effect of the PNN are better, and the recognition accuracy is increased by 15% and 7.5%, respectively, compared with the traditional PBNN and GRNN. This method can accurately and effectively identify the microseismic signals of mines and has good generalization performance.

A denoising method of mine microseismic signal based on NAEEMD and frequency-constrained SVD

A denoising method of mine microseismic signal based on NAEEMD and frequency-constrained SVD

Analysis of rock microseismic signal based on blind source wavelet decomposition algorithm

At present, microseismic technology is a widely used method for monitoring the rock burst phenomenon during the construction of deep-buried tunnels. The rock fracture in the tunnel will generate seismic waves. The seismic wave has strong randomness and low energy and is usually mixed with environmental noise, which is called the microseismic signal. The microseismic signal received by the geophone will contain other noises. So, there is an urgent need for an algorithm that can quickly decompose the mixed signal. To solve this problem, this paper proposes a blind source wavelet algorithm to extract the rock fracture signal. First, the signal is preprocessed, and the mixed signal matrix is established. Second, the blind source decomposition algorithm is used to process the signal, and the effective signal is reconstructed. Finally, the wavelet algorithm is used to further remove the noise to enhance the microseismic signal. The proposed method is compared with the wavelet decomposition method and the empirical mode decomposition (EMD) method through the laboratory simulation data and the actual signal of Baihetan Hydropower Station. It is concluded that the proposed algorithm can effectively decompose and remove the noise in the mixed signal, and the decomposition accuracy is better than the wavelet decomposition method and the EMD method. The proposed algorithm has certain practical significance for the identification of microseismic signals of rock fracture and for rock burst early warning.

Design and Optimization of a Triangular Shear Piezoelectric Acceleration Sensor for Microseismic Monitoring

Aiming at the characteristics of low sensitivity and narrow frequency range of existing microseismic monitoring sensors for mine water hazard prevention and control, a piezoelectric acceleration sensor for microseismic monitoring based on a kind of triangular shear structure is proposed. Firstly, the structure of the triangular shear piezoelectric acceleration sensor is designed, and its dynamic model is built. The structural and material parameters related to natural frequency and sensitivity are analyzed. Then, the selection of piezoelectric ceramic materials is discussed. The parametric design of the designed sensor is carried out, and its finite element structural model is built by ANSYS. The modal analysis, resonance response analysis, and piezoelectric analysis of the designed sensor are carried out. The simulation results indicate that the working frequency and sensitivity of the designed sensor meet the requirements of microseismic monitoring. Response surface optimization is adopted to analyze the influence of sensor element design variables on the sensitivity and resonant frequency of the designed sensor. The reoptimized design of the reference sensor improves the resonant frequency of the designed sensor by 9.46% and the charge sensitivity by 18.96%. Finally, the designed sensor is calibrated, and the microseismic signal detection experiment is carried out. The results indicate that the resonant frequency of the designed sensor is 6150 Hz, the working frequency is 0.1-2050 Hz, and the charge sensitivity is 1600 pC/g. The sensor can detect microseismic signals with a wide frequency range and high sensitivity.

Microseismic Time Delay Estimation Method Based on Continuous Wavelet.

The microseismic signal is easily affected by observation noise and the inaccurate estimation of traditional methods will seriously reduce the location accuracy of the microseismic event. Therefore, based on the continuous wavelet spectrum and the similarity coefficient, a fast and efficient microseismic time delay estimation method is proposed. Firstly, the original signals are denoised by continuous wavelet transform. Subsequently, the time-frequency transform of the original signal by continuous wavelet transform, time-frequency signal extraction is the process of band-pass filtering, which can further reduce the influence of noise interference on the time delay estimation. Finally, we calculated the similarity between the time-frequency signals via the time domain and frequency domain integration. The similarity function is based on correlation and proposed according to the time-frequency transformation provided by the phase spectrum to evaluate the similarity between two noisy signals. The time delay estimation is determined by searching for the similarity function peak. The experimental results show the precision and accuracy of the method over the cross-correlation method and generalized cross-correlation phase transformation method, especially when the signal-to-noise ratio is low. Therefore, a new time delay estimation method for non-stationary random signals is presented in this paper.

Research on classification and identification of mine microseismic signals based on deep learning method

Research on classification and identification of mine microseismic signals based on deep learning method

Intelligent Microseismic Events Recognition in Fiber-Optic Microseismic Monitoring System Compared With Electronic One

Fiber-optic seismic sensor has become an emerging effective tool for microseismic monitoring, which is of importance for oil and gas production improvement. When processing the fiber-optic data, the recognition of microseismic events is the first crucial step, which is rarely reported due to the lack of such data. In this paper, based on the features of central data distribution, skewness, kurtosis, energy entropy, etc., the classification accuracy of microseismic (MS) events obtained by fiber-optic MS monitoring system and electronic one is compared using machine learning algorithm (e. g. SVM and KNN) for the first time. The results show that fiber-optic data has higher classification accuracy than electronic data when using the feature of central data distribution owing to the high signal-to-noise ratio of fiber-optic data. However, when choosing the features of energy entropy, zero-crossing rate and the energy proportion of specific frequency band, electronic data has higher classification accuracy than fiber-optic data benefiting from the longer events duration and the lower frequency components of electronic data. When using skewness and kurtosis, the classification performance in fiber-optic data is almost consistent with the electronic one. Moreover, the results indicate that the characteristics of MS signal itself have a greater impact on the discrimination ability of MS events than the applied machine learning classification algorithm.

Attenuation characteristics of impact-induced seismic wave in deep tunnels: An in situ investigation based on pendulum impact test

The radiated seismic energy is an important index for the intensity assessment of microseismic (MS) events and the early warning of dynamic disasters. However, the energy of MS signals is significantly attenuated due to the heterogeneity and viscous damping of rock media. Therefore, the study on attenuation characteristics of MS signals in underground engineering has practical significance for the accurately estimation of radiated seismic energy. Based on a pendulum impact test facility and MS monitoring system, an in situ investigation was carried out to explore attenuation characteristics at a deep tunnel. The results show that the seismic energy and peak particle velocity (PPV) attenuation are exponentially related to the propagation distance. The attenuation coefficient of energy is larger than that of PPV. With the increase in the input impact-energy, the seismic energy attenuation coefficient decreases as a power function. An empirical relationship between energy attenuation coefficient and wave impedance of rock mass was established in this scenario. Moreover, the time-frequency characteristics and energy distribution laws of impact-induced signals were investigated by the continuous wavelet transform (CWT) and wavelet packet analyses, respectively. The dominant frequency of signals decreases gradually as the propagation distance increases. Based on the energy attenuation characteristics, a new method was proposed to calculate the released source energy of MS events in the field. This study can provide an insight into energy attenuation characteristics of seismic waves and references for attenuation correction in seismic energy calculation.

Automatic Events Recognition in Low SNR Microseismic Signals of Coal Mine Based on Wavelet Scattering Transform and SVM

The technology of microseismic monitoring, the first step of which is event recognition, provides an effective method for giving early warning of dynamic disasters in coal mines, especially mining water hazards, while signals with a low signal-to-noise ratio (SNR) usually cannot be recognized effectively by systematic methods. This paper proposes a wavelet scattering decomposition (WSD) transform and support vector machine (SVM) algorithm for discriminating events of microseismic signals with a low SNR. Firstly, a method of signal feature extraction based on WSD transform is presented by studying the matrix constructed by the scattering decomposition coefficients. Secondly, the microseismic events intelligent recognition model built by operating a WSD coefficients calculation for the acquired raw vibration signals, shaping a feature vector matrix of them, is outlined. Finally, a comparative analysis of the microseismic events and noise signals in the experiment verifies that the discriminative features of the two can accurately be expressed by using wavelet scattering coefficients. The artificial intelligence recognition model developed based on both SVM and WSD not only provides a fast method with a high classification accuracy rate, but it also fits the online feature extraction of microseismic monitoring signals. We establish that the proposed method improves the efficiency and the accuracy of microseismic signals processing for monitoring rock instability and seismicity.

Microseismic signal denoising by combining variational mode decomposition with permutation entropy

Microseismic signal denoising by combining variational mode decomposition with permutation entropy

Experimental research on the evolutionary characteristics of acoustic signals for concrete cracking under uniaxial compression

In this study, the evolutionary characteristics of the acoustic emission (AE), microseismic (MS), and sound signals released during concrete fracture in terms of their amplitude, b value, activity level, and frequency spectrum were investigated. The results indicated that the b values are all characterised by a continuous decline immediately before concrete failure, accompanied by the appearance of low-frequency signals with high amplitudes. The occurrence of these two phenomena can be used as precursor information for the subsequent concrete failure. Nevertheless, there are significant differences among the three signal types. Frequency ranges of these acoustic signals and hierarchical characteristics of the concrete fracturing result in different signal activeness, which can be evidenced by the following order of the initial stress of activeness: sound (85% σpeak), MS (70% σpeak), and AE (57% σpeak). Diversities in signal acquisitions modes lead to differences near the failure stage. A quiescent period of AE, characterized by only a few signals with low frequency and high amplitude, was observed immediately before concrete failure, whereas MS and sound were distributed in various frequency bands without quiescent periods. These differences indicated that the three types of signals have different sensitivities and applicability. A reliable concrete fracturing analysis and prediction can be achieved by the integrated usage of the three types of acoustic techniques.