Microseismic Data Research Articles

Risk Field of Rock Instability using Microseismic Monitoringdata in Deep Mining.

With the gradual depletion of surface resources, rock instability caused by deep high stressand mining disturbance seriously affects safe mining. To create effective risk management, a rockinstability risk field model using microseismic monitoring data is proposed in this study. Rockinstability risk was presented visually in 3D visualization. The in-situ microseismic monitoringdata was collected and analyzed to make calculation of peak ground velocity (PGV), peak groundacceleration (PGA), energy flux, energy and seismic moment. Indicator weights of PGV, PGA, energyflux are confirmed by using the analytic hierarchy process (AHP) to calculate risk severity. The Copulafunction is then used to solve the joint probability distribution function of energy and seismic moment.Then the spatial distribution characteristics of risk can be obtained by data fitting. Subsequently, thethree-dimensional (3D) risk field model was established. Meanwhile, the established risk field isverified by comparing monitoring data without disturbance and the blasting data with disturbance.It is suggested that the proposed risk field method could evaluate the regional risk of rock instabilityreasonably and accurately, which lays a theoretical foundation for the risk prediction and managementof rock instability in deep mining.

Microseismic positioning of an isolated working face under complex geological conditions and its engineering application

In view of the inability to accurately locate vibrations in isolated workings under complex geological conditions, an adaptive rotational categorization method, a downhill comparison method based on time-frequency analysis (TFA-DC method), a variable-step acceleration search method, and a dual-phase seismic source location method (TD-DL method) are proposed. A set of integrated software with features of “visualization”, “interactive” and “one-click” was developed for microseismic data processing. Results show that compared with the improved STA/LTA method, the recognition accuracy of the island working face microseismic signal by the adaptive wheel classification method is increased by 4.8%. Compared with the improved STA/LTA method, the TFA-DC method has the advantage that it can simultaneously pick up the exact p wave and the peak S wave, and the failure ratio is 0. Compared with simulated annealing algorithm and genetic algorithm, stepwise accelerated search method has better results. The standard deviation of objective function value, location error and wave velocity error are all 0. Method improves the positioning of the TD - DL detector coordinates, dual phase and coherence of known information, such as the positioning result positioning error is only the p wave and S wave ChanZhen phase positioning method of 9.5% and 14.5%, to a certain extent offset the p wave and S wave ChanZhen phase calculation of the positioning error, so as to improve the effect of source localization precision of the inversion.

Research on in situ stress inversion of deep‐buried tunnel based on pressure/tension axis mechanism and geological structure

AbstractThe investigation of the in situ stress distribution has always been a key condition for engineering design of deep tunnels and analysis of surrounding rock stability. In this paper, a comprehensive judgment method coupled with pressure/tension (P/T) axis mechanism and geological structure is proposed to invert the in situ stress in the Duoxiongla tunnel in Tibet. In the process of TBM tunnel excavation, 3887 groups of microseismic events were collected by means of microseismic monitoring technology. By studying the temporal and spatial distribution of 3887 groups of microseismic events, 42 groups of microseismic data were selected for in situ stress inversion. Then the focal mechanisms of 42 groups of microseisms were inverted. Combined with the analysis of the previous geological survey, the inversion results of the in situ stress were analyzed. According to the focal mechanism of the tunnel area, the linear in situ stress inversion method was used to invert the in situ stress in the source area. Finally, according to the PTGS (pressure/tension axis mechanism and geological structure) comprehensive judgment method proposed in this paper, the in situ stress of the tunnel microseismic region was determined. The results show that there are mainly three groups of fissures and joint surfaces in the tunnel area, and the in situ stress is dominated by the horizontal tectonic stress; the main driving force of the rupture surface in the excavation process of Duoxiongla tunnel is the horizontal tectonic stress; the distribution of the maximum and minimum principal stress obtained by the inversion is consistent with the distribution of the P/T axis; combined with the linear in situ stress inversion method and the comprehensive judgment of PTGS, the azimuth and dip angles of the three principal stresses are finally determined as N90.71°E, 4.06°, N5.35°W, 3.06°, and N8.10W, 85.32°, respectively. The study verifies the feasibility of microseismic inversion of in situ stress.

Anisortopic Modeling of Hydraulic Fractures Height Growth in the Anadarko Basin

Correct estimation of hydraulic fracture height growth is a critical step in the design of Hydraulic Fracturing (HF) treatment, as it maximizes the reservoir stimulation and returns on investment. The height of the fractures is governed by several in situ conditions, especially stress variation with depth. The common workflow to estimate stress is by building the mechanical earth model (MEM) and calibrating it using the Diagnostic Fracture Injection Test (DFIT). However, DFIT interpretation is a complex task, and depending on the method used, different results may be obtained that will consequently affect the predicted hydraulic fracture height. This work used the tangent and compliance methods for DFIT interpretation, along with isotropic and anisotropic stress profiles, to estimate the HF height growth using numerical modeling in a 3D planar HF simulator. Data from two wells in the Anadarko Basin were used in this study. The predicted height was compared with microseismic data. The results showed that even though the tangent method fits better to the isotropic stress profile, HF did not match with the microseismic data. On the contrary, the anisotropic stress profile showed a good match between the compliance DFIT model and the microseismic events. Based on the discussions presented in this study, the validity of the DFIT interpretation is debatable, and when the formations are anisotropic, the isotropic model fails to correctly estimate the minimum stress profile, which is the main input for the estimation of the fracture height. This is in addition to the fact that some researchers have questioned the use of the tangent method in low-permeability formations.

A novel long-short term memory network approach for stress model updating for excavations in high stress environments

ABSTRACT Digitalisation has increased access to large amounts of data for rock engineers. Machine learning presents an opportunity to aid data interpretation. The operators of Garson Mine use a microseismic database to calibrate a mine-scale finite difference model, used to assess seismic risk to inform mine operations. A Long-Short Term Memory (LSTM) network is proposed for stress model updating. The model is trained using microseismic data, geology, and geomechanical parameters from the FLAC3D model. Two LSTM networks are developed for Garson Mine: (1) predicting far field principal stresses in the FLAC3D model, and (2) predicting the far field six-component stress tensors in the model. Various LSTM network hyperparameters were analyzed to determine the architecture for the targets: input encoding and pre-processing, training solver, network layer architecture, and cost function. Architectures were chosen based on the corrected Akaike Information Criterion (AICc), coefficient of determination (R2), and percent capture (%C). When predicting principal stresses, AICc = −59.62, R2 = 0.996, and %C = 97%, and when predicting the six-component stress tensor AICc = −45.50, R2 = 0.997, and %C = 80%. This research represents progress towards continuous, automated updating of numerical models such that rapid, more accurate forecasts of changes in stress conditions will allow earlier reaction to challenging stress environments, increasing safety of excavations.

Micromechanical characterisation of overburden shales in the Horn River Basin through nanoindentation

The paper presents a micromechanical characterisation of Fort Simpson shale, which overlies unconventional gas-producing lithologies in the Horn River Basin, NW Canada. The Fort Simpson formation is clay-rich and microseismic data recorded during hydraulic fracturing events in the underlying reservoir has shown the formation acts as a barrier to fracture development, with a notably anisotropic seismic response. Samples were prepared from core fragments and the composition and texture of the shale was characterised using X-ray diffraction, mercury injection porosimetry and scanning electron microscopy (SEM). Nanoindentation testing was used to obtain the mechanical response of the shale microstructure, at grain-scale. The indentation was conducted on a grid pattern and samples were oriented both parallel and perpendicular to the bedding plane to assess the inherent mechanical anisotropy. Chemical analysis of the grids was also undertaken through SEM/EDS (energy dispersive X-ray spectroscopy) and the coupled chemo-mechanical data was used to characterise the material phases of the shale through a statistical clustering procedure. The results show that Fort Simpson shale broadly consists of a soft clay phase, with strongly anisotropic elastic stiffness, and stiffer but effectively isotropic grains of quartz and feldspar. A simple upscaling scheme was also applied to link the grain-scale elastic stiffness to the field-scale microseismic data.

A Reliable Online Dictionary Learning Denoising Strategy for Noisy Microseismic Data

Improving the quality of microseismic recordings is a critical step in microseismic data processing. We introduce a wavelet-weighted online dictionary learning (WWODL) denoising strategy for noisy microseismic recordings. We develop an adaptive parameters estimation approach for tunable Q-factor wavelet transform, which provides accurate periodic and non-stationary subband information from microseismic data for online dictionary learning (ODL). A sliding time window is employed to divide the obtained subbands into a series of patches of equal length, which are then assembled into a matrix and fed into the ODL. The subband kurtosis information is weighted to the constraint function of the ODL, further enhancing the sparse coding ability for each subband. Shortening the oscillation duration, an improved ODL is developed with a faster convergence speed in calculating sparse coefficients. Our results confirm that the WWODL can suppress high-frequency, low-frequency, and shared-bandwidth noises and has a minimal impact on the first arrival. The time consumption and signal-to-noise ratio of the WWODL are average 1/8.675 and 56.82% higher than ODL, respectively.

An investigation of data analysis method for hydraulic fracturing based on the water hammer effect

Hydraulic fracturing plays a crucial role in the development of unconventional resources, and assessing fracture geometry is a fundamental aspect of hydraulic fracturing analysis. Among the various technologies for evaluating fracture geometry, the water percussion signal-based method has gained popularity due to its cost-effectiveness, convenience, and real-time capabilities. In this study, we present a methodology for analyzing fracture data that utilizes the water percussion effect. By leveraging the impact of fractures on water percussion pressure attenuation, we propose an inverse calculation approach to determine fracture geometry. Firstly, we introduce the RCI (Reservoir-Completion-Interaction) circuit model, which effectively addresses the bottom hole fracture boundary and simulates the variation of water percussion pressure in the wellbore. Secondly, the simulation results are utilized in an iterative process to determine the RCI values, which are then utilized for the inversion calculation of fracture geometry. Finally, we apply this method to a field case and compare the simulation results with microseismic monitoring data. A larger resistance (R) value indicates a smaller fracture volume, while the (C) value can be used to monitor any poorly performing stages of the fracturing simulation process. The (I) value primarily affects the parameter calculation of the fracture size, especially the fracture width. The results of the field application show that the accuracy rate exceeding 80%, validating the reliability of the model and providing a valuable reference for field fracturing data analysis.

Unsupervised Clustering of Microseismic Signals Using a Contrastive Learning Model

Distinguishing useful microseismic signals is a critical step in microseismic monitoring. Here, we present the time series contrastive clustering (TSCC) method, an end-to-end unsupervised model for clustering microseismic signals that employs a contrastive learning network and a centroidal-based clustering model. The TSCC framework consists of two successive phases: pre-training and fine-tuning. In the pre-training phase, two random cropping augmentations are used to transform the time series microseismic data into two distinct but correlated views. Then, the multi-scale temporal and instance contrasting learning are used to discriminate between negative and positive views, thus motivating the encoder to capture microseismic signal contextual information from multiple perspectives and generate distinct representations from unlabeled data. During the fine-tuning phase, the encoder weights are iteratively fine-tuned by simultaneously performing contrast learning and clustering. The corresponding loss is a weighted combination of the contrastive and clustering loss functions, which induces the encoder to learn representations that improve clustering performance. The test results demonstrate that the proposed method can achieve better clustering accuracy than popular clustering methods, including k-means, deep embedding clustering, unsupervised clustering with deep convolutional autoencoders, and deep clustering with self-supervision. Moreover, the TSCC model can produce results comparable to supervised deep learning approaches while requiring no labeled data, manual feature extraction, or large training datasets. In practice, the TSCC model has a clustering accuracy of 98.07% and a normalized mutual information of 86.26%.

Case Study on Dynamic Identification of Overburden Fracture and Strong Mine Pressure Mechanism of Isolated Working Face Based on Microseismic Clustering

Strong mine pressure has a great impact upon the safety production of coal mines. Microseismic information provides a more advanced technical means for overburden fracture dynamic identification and mine pressure mechanism research, since it contains rich information on rock fracture sources. In this study, the isolated LW8102 working face in Tongxin Mine was investigated in order to propose a spatio-temporal microseismic event data analysis method based on the k-means clustering algorithm. This algorithm can handle dynamic identification of overburden fractures constrained by spatiotemporally discrete distributions of microseismic events. This provided the dynamic extension process and the fracture distribution pattern of the overburden: eight fracture extensions were formed in the overburden. In each extension, vertical fractures connected the low and high rock layers in the LW8102 and LW8103 goafs, and through fractures connected the LW8102 and LW8103 goafs in their high, middle, and low levels. Some extensions had fractures that were connected to form a closed loop structure. In the vertical fracture, there was a tendency for one or two layers of the stratum to fail first, and then extend to one or both sides. The process of through and vertical fracture propagation followed a certain temporal sequence, reflected primarily in two forms: firstly, as the vertical fracture extended to a certain layer, it provided the initial rupture space for through fracture spreading; secondly, the through fracture first broke, and then extended to the vertical fracture until it intersected with the vertical fracture or provided the initial rupture space for the expansion of the vertical fracture. By matching the overburden fracture to the mine pressure that responded to the support resistance, we analyzed the mechanism of mine pressure at the working face. Through fracture at the high level was found to be the primary cause of the occurrence of mining pressure. It was precisely placed that the formation of multiple adjacent high through fractures 110 m from the floor, triggering simultaneous instability motion of the lower multi-layer level rock; this was the main reason for the phenomenon of strong mine pressure at the working face. Meanwhile, high through fracture at 80 m from the floor was the main reason for the phenomenon of large mine pressure at the working face.

Analysis Of Liquifaction Potential Using The Method Microzonation in Coastal Zone at North Aceh District

Liquefaction is a natural phenomenon that occurs when previously sandy soil turns into mush as a result of an earthquake and has the potential to become a natural disaster if not anticipated early. the coastal zone of Dewantara District is included in the earthquake-prone area (referring to the results of previous studies, the value of the seismic vulnerability index is between 0.41 to 18.3 Kg, with the potential for damage caused to be medium and high risk) because most of the area is an area that has alluvial plains a wide beach and located on a tectonic route so that it is estimated that there is a potential follow-up threat after an earthquake occurs, in the form of liquefaction. Determination of the distribution of liquefaction potential due to earthquakes is determined based on data obtained from microtremor measurements. The specific objective of this study is to provide information regarding the potential and vulnerability of liquefaction to Policy Makers/Local Governments in evaluating plotting areas and/or preparing regional spatial plans based on disaster risk reduction/mitigation aspects. Based on the results of the processed data analysis, several important conclusions can be identified in this study. The character of soil vulnerability is related to the value of dominant frequency (f0), soil amplification (A0), and seismic vulnerability index (seismic vulnerability index). This relationship can be used to formulate whether an area has the potential to experience lateral ground motion (liquefaction) which potentially can affect the coastal area of ​​Dewantara District. Based on microseismic data analysis (standard deviation), several disaster-prone zone points were obtained in the north and east of the measurement point. In detail, this is shown by the map of the distribution of SVI values ​​in the Coastal Area which is in the range of 20-100, especially in the north it is at a value above 75, which allows for a higher chance of liquefaction on the condition that there are major earthquakes accompanied by with rock pores filled with fluid.

How can casing deformation be prevented during hydraulic fracturing of shale gas?—A case study of the Weiyuan area in Sichuan, China

In the process of hydraulic fracturing of shale gas horizontal wells in the Weiyuan area of Sichuan, China, casing deformation (CD) is a frequent problem that has seriously affected the benefits of shale gas development and remains to be solved for many years. In this paper, based on the analyses of Multi-finger Imaging Tool-24 (MIT-24) data, microseismic data, and the relationship between CD and engineering parameters, we confirm that the wellbore morphology in all CD sections in the Weiyuan area is consistent with shear deformation characteristics and clarify the cause of all CD: hydraulic fracturing leads to increased fluid pressure in well-intersecting faults and large fractures (F-Fs), inducing their shear slip, and eventually leads to asymmetric compression on the casing by strata on both sides of F-Fs, i.e., CD. On this basis, we propose a set of methods to control (or moderate) the shear slip of F-Fs by reducing the additional fluid pressure in target F-Fs during hydraulic fracturing to prevent CD, including multi-cluster perforations, temporary plugging F-Fs and fracturing fluid flowback. We also analyze the theoretical basis, feasibility and workflow of these methods. These methods have been tested in 44 horizontal wells in the Weiyuan area, of which the result is a radical decrease in the CD rate of the single well (from 54% to 9.1%), indicating that the slip of F-Fs and the resulting CD during hydraulic fracturing can be controlled and providing a reference for the prevention of hydraulic fracturing-induced earthquakes.

Microseismic Dynamic Response and Multi-Source Warning during Rockburst Monitoring Based on Weight Decision Analysis

To prevent rockburst disasters and improve the accuracy of warnings for rockburst, based on the microseismic data of the 1366 working face of Hengda Coal Mine collected by the microseismic monitoring system, Fourier transform, wavelet packet transform, and Hilbert-Huang transform analysis methods are used for time-frequency domain joint analysis. The time-frequency differences of the main frequency, amplitude, frequency band percentage, and instantaneous energy of the high-energy microseismic event and the events before high-energy microseismic event are obtained. The analysis shows that the high-energy event has obvious low frequency characteristics, and when the high-energy event occurs, the instantaneous energy shows an obvious "inverted V" trend. At the same time, it is found that the acoustoelectric indexes show a trend of "rising" or "inverted V" when the high-energy event occurs. On this basis, the unascertained measure comprehensive evaluation model of rock burst hazard is established by analytic hierarchy process (AHP). Based on the analysis of microseismic data and the acoustoelectric index of the 1366 working face in Hengda coal mine, it is of great significance to determine the warning indicators for rockburst, improve the accuracy of uncertainty quantitative analysis for rockburst, and improve the discrimination accuracy of rockburst risk.

Impact of Lithologic Heterogeneity on Brittleness of Cenozoic Unconventional Reservoirs (Fine-Grained) in Western Qaidam Basin

In order to understand the impact on reservoir brittleness of lithologic heterogeneity in the continental mixed fine-grained sedimentary rocks in the western Qaidam Basin, the mechanical properties of the rocks and their correlation with mineral composition and petrographic characteristics were studied. A total of 20 samples from two parallel groups (10 samples in each group) were analyzed by triaxial stress test mineralogy, and morphology. The results show that the reservoir rocks can be divided into five different types according to the mechanical properties of the reservoir (characterized by stress–strain curves), among which Types I and III belong to a similar elastoplastic failure model, Type II shows a special pulse failure mode for plastic material, Type IV shows a failure mode of mixed characteristics, and Type V exhibits a typical plastic failure model. The correlation between minerals and mechanical properties indicates that quartz and feldspar, which are often considered brittle minerals, do not contribute much to the brittleness of these continental fine-grained sedimentary rocks. The main minerals affecting the reservoir brittleness are dolomite and clay minerals, contributing positively and negatively to it, respectively. The petrographic analysis results prove that the abnormal correlation between rock mechanical properties and quartz and feldspar is caused by the different rock fabrics. When dolomite forms a rock skeleton, it typically exhibits greater strength, brittleness and physical properties than other minerals. Based on the results, a brittleness evaluation standard for continental unconventional reservoirs (fine-grained) is proposed, and the validity of the standard is verified by the spatial correlation between the lithology probability model and the micro-seismic monitoring data. This indicated that the spatial heterogeneity of the dolomite-rich rock is the main controlling factor for the effective development of the Cenozoic continental unconventional reservoirs (fine-grained) in the Western Qaidam Basin.

Deformation characteristics of sidewall and anchorage mechanisms of prestressed cables in layered rock strata dipping steeply into the inner space of underground powerhouse cavern

Local stability of underground powerhouse cavern subject to low to moderate geo-stresses, is frequently threatened by the unpredictable displacements of high sidewall in layered rock masses dipping into cavern space. Further understanding these displacements and the anchoring behaviours helps provide effective measures for ensuring the safe, rapid construction of major projects. This study investigates the deformation features of sidewall and anchoring mechanisms of long prestressed cables in layered strata steeply inclined toward the inner space of underground powerhouse, using field tests and numerical tools. First, the development characteristics of sidewall in layered strata steeply dipping into cavern space are analysed based on the deformation monitoring data. Time-dependent relationships between those displacements and depths from the excavation surface distinguished from that in other typical underground caverns in layered rock masses, are critically discussed. Second, damage characteristics of these strata caused by excavating activities are presented based on the microseismic (MS) monitoring data. Effects of damage zone on the evolution of deep deformations distinctively involved, are described in detail. Then, the calculation platform coupling MS data and continuum modelling, suitable for simulating progressive damage of layered rock masses, is introduced to reveal new insights into the mechanisms behind those distinctive deformations. Finally, the modified models properly capturing the mechanical responses of prestressed anchor cables, are integrated into such a calculation platform. Anchoring mechanisms of long prestressed cables associated with those deformations of layered strata dipping into the cavern are thus further investigated. This can explain why the prestressed anchor cables provide less effective control effects on deep deformations mainly induced by shear dislocations on bedding planes.

A Continuous Data Acquisition System for Three-Component Surface Microseismic Real-Time Monitoring

Surface microseismic monitoring is one of the most critical technologies for monitoring the dynamic subsurface reservoir processes. Due to the weak signals and poor signal-to-noise ratio (SNR), there are two major challenges to monitor the reservoir processes in an accurate and efficient way. One challenge is concerned with weak signals acquisition and how to evaluate the instrument’s abilities to record the microseismic events. The other one is how to realize real-time data monitoring during dynamic processes. Taking these challenges into consideration, a design approach to develop buried sensors consisting of three-component sensors and a low-cost, low-power consumption embedded device for microseismic monitoring data logger with the acquisition, synchronization, record, and transmission functions are presented. The circuits are divided into two printed circuit boards (PCBs). One is for a signal processing circuit, which consists of signal conditioning and analog-to-digital converter (ADC), and the other is for a digital circuit, which consists of the microcontroller and peripherals, including the global navigation satellite system (GNSS), Micro Secure Digital Memory Card (SD card), and Ethernet transceivers. The circuit for signal conditioning and ADC is a combination of a preamplifier, low-pass filter, programmable gain amplifier, and ADC. It has three modes: real-time monitoring, continuous recording, and combination. A software is designed to work in cooperation with the hardware to adjust the working parameters and modes. The self-noise of the data logger is tested in the laboratory, and the monitoring abilities of the data logger are evaluated. Finally, the surface microseismic monitoring system has been tested in the field to evaluate its performance for academic purposes.

Image-domain DAS 3D VSP elastic transmission tomography

SUMMARY Full-wavefield elastic imaging of active-source seismic data acquired by downhole receivers commonly offers higher-resolution subsurface images in the vicinity of a borehole compared to conventional surface seismic data sets, which can lack higher-frequency wavefield components due to longer travel paths and increased attenuation. An increasingly used approach for downhole acquisition is vertical seismic profiling (VSP), which has become more attractive when coupled with distributed acoustic sensing (DAS) using optical fibres installed in wells. The main difficulty for generating high-quality images with full-wavefield imaging tools for DAS VSP data, though, is the need for an accurate velocity model. To build plausible velocity models using active-source DAS VSP data, we adopt a 3-D image-domain elastic transmission tomography technique, originally developed for surface-recorded passive (microseismic) data, by exchanging the source and receiver positions (i.e. reciprocity) to mimic a passive-seismic surface monitoring scenario. The inversion approach exploits various images for each source constructed through time-reverse imaging (TRI) of downgoing P- and S-wave first-arrival waveforms. The TRI process uses the kinetic term of the (extended) PS energy imaging condition that exhibits sufficient sensitivity to velocity model errors. The method automatically updates the P- and S-wave velocity models to optimize image focusing via adjoint-state inversion. We illustrate the efficacy of the adopted elastic inversion technique using an active-source DAS 3-D VSP field data set acquired in the North Slope of Alaska. The numerical experiments demonstrate that the inverted elastic velocity models can be further used in full-wavefield acoustic/elastic imaging algorithms to obtain accurate subsurface images.

Traveltime-based microseismic event location using artificial neural network

Location of earthquakes is a primary task in seismology and microseismic monitoring, essential for almost any further analysis. Earthquake hypocenters can be determined by the inversion of arrival times of seismic waves observed at seismic stations, which is a non-linear inverse problem. Growing amounts of seismic data and real-time processing requirements imply the use of robust machine learning applications for characterization of seismicity. Convolutional neural networks have been proposed for hypocenter determination assuming training on previously processed seismic event catalogs. We propose an alternative machine learning approach, which does not require any pre-existing observations, except a velocity model. This is particularly important for microseismic monitoring when labeled seismic events are not available due to lack of seismicity before monitoring commenced (e.g., induced seismicity). The proposed algorithm is based on a feed-forward neural network trained on synthetic arrival times. Once trained, the neural network can be deployed for fast location of seismic events using observed P-wave (or S-wave) arrival times. We benchmark the neural network method against the conventional location technique and show that the new approach provides the same or better location accuracy. We study the sensitivity of the proposed method to the training dataset, noise in the arrival times of the detected events, and the size of the monitoring network. Finally, we apply the method to real microseismic monitoring data and show that it is able to deal with missing arrival times in efficient way with the help of fine tuning and early stopping. This is achieved by re-training the neural network for each individual set of picked arrivals. To reduce the training time we used previously determined weights and fine tune them. This allows us to obtain hypocenter locations in near real-time.

Adaptive filtering of microseismic data for monitoring a water-conducting fractured zone in a mine

Water-conducting fractured zones in a rock mass can cause problems in mining. Attempts have been made to monitor their development using microseismic signals. However, due to the lack of prior information, it is difficult to filter out mixed low-frequency interference with traditional denoising methods. In this work, the proposed adaptive filtering algorithm is applied after the wavelet packets are decomposed. It is based on a cross-correlation analysis. The algorithm takes a high-quality signal in the common source waveform as prior information and applies the corresponding correlation coefficients between subbands as a threshold. The algorithm was verified with simulations. The results show that low-frequency interference can be effectively suppressed by filtering. For single-frequency interference, the signal-to-noise ratio increased from − 10.18 to 13.97, and the root-mean-square error was 43.88. For multi-frequency interference, it increased from − 10.01 and − 2.63 to 13.50 and 7.99. The root-mean-square errors were 46.31 and 138.07. The narrower the main frequency band of the interference signal and the less the overlap of the main frequency band of the interference signal and the effective signal, the better the filtering effect. When the algorithm was applied to microseismic data collected in the field, the number of effective channels increased and the accuracy improved. The development of a water-conducting fractured zone in the field was consistent with the microseismic location obtained after interference was removed by the algorithm, which indicates that it is feasible to monitor a water-conducting fractured zone by analyzing microseismic waveforms with the adaptive filtering algorithm.

A novel microseismic classification model based on bimodal neurons in an artificial neural network

Microseismic monitoring systems deployed in deep underground engineering can capture massive waveform signals in real time. However, some noise signals are highly deceptive and similar to microfracture signals. This problem usually requires engineers to compare the signal characteristics from different domains and poses a challenge for the rapid detection of microseismic data. To address this problem, we establish a new waveform multiclassification model based on the multimodal feature extraction technology. The proposed model constructs a dual-branch convolutional neural network to learn the bimodal features from the time and frequency domains. Furthermore, the attention mechanism (channel and spatial attention) with skip connection is used to increase the focus on the prominent features of the waveform and reduce the interference caused by irrelevant information. We conduct experimental studies using microseismic data from deep tunnel projects under complex geological conditions and classify the recorded data into three target types: microseismic, blast, and noise waveforms. The results show that the proposed approach achieves good multiclassification performance for noise signals highly similar to microseismic signals as well as for noisy microseismic signals with different signal-to-noise ratios. In summary, this study lays the foundation for further exploration of the potential application of multimodal fusion in seismic engineering and various rock engineering fields.