A hydraulic fracture monitoring system and method based on downhole microseismic sensor networking

Abstract

The present application discloses a hydraulic fracture monitoring system and method based on a network of underground microseismic sensors, which belongs to the field of coal mining, and includes: multiple microseismic sensors for collecting microseismic signals generated by rock fractures; multiple data acquisition substations for receiving and processing microseismic signals; a data transmission substation for converting the microseismic signals processed by the data acquisition substations into optical signals; a microseismic server for performing microseismic event analysis based on the optical signals; a mine network switch for connecting the data transmission substations and the microseismic server for network data exchange; an in-mine ring network cabinet for connecting the mine network switch and the microseismic server to form a communication network of the system; in view of the fact that it is difficult to accurately distinguish between microseismic signals generated by hydraulic fracturing and coal mining activities in a complex underground environment, resulting in low hydraulic fracturing detection accuracy, the present application improves the real-time monitoring accuracy of hydraulic fracture expansion by dividing the fracturing area and the mining area, distinguishing between microseismic signals generated by hydraulic fracturing and coal mining activities.

Description

Technical Field

[0001] The present application relates to the field of coal mining, and more specifically, to a hydraulic fracture monitoring system and method based on a network of underground microseismic sensors. Background Art

[0002] The geological conditions of my country's coal mines vary widely, with some deep strata containing thick, hard roofs that can easily form complex, hazard-prone structures. Traditional roof control techniques using blasting have demonstrated numerous drawbacks, including poor safety, high engineering workload and explosives requirements, and underground air pollution. Blasting is particularly effective in gassy mines or coal seams, posing a serious safety risk of gas or coal dust explosions.

[0003] With the continuous development of technology, hydraulic fracturing is increasingly being used to control roof hazards. For coal seams with thick, hard roof structures at high and low levels, hydraulic fracturing plays two primary roles: first, it transfers stress from the overlying rock strata; second, it alters the overlying rock structure, increasing the roof's resilience to collapse after mining, allowing the thick, hard roof to contact the waste rock as quickly as possible. This solution can effectively control goaf areas in coal mines, reducing the occurrence of geological hazards such as roof subsidence and surface collapse, offering advantages such as safety, environmental friendliness, and low construction costs.

[0004] Because hydraulic fracturing occurs underground in long, narrow spaces and complex geological structures, it's impossible to directly measure information such as the length and shape of the fractures, nor is it possible to determine the extent to which the fractures affect rock stress. This makes it difficult to accurately monitor and effectively evaluate the fracture propagation behavior during the fracturing process. Traditional fracturing monitoring methods, such as fluid pressure analysis and fracture geometry simulation, suffer from low resolution and blind spots, making them inadequate for coal mine safety management and production needs.

[0005] To better evaluate and observe the results of directional hydraulic fracturing, indirect methods are needed to accurately track crack propagation behavior and provide real-time disaster warnings. Microseismic monitoring technology, a well-developed, multi-optimized, and well-established engineering monitoring method, is widely used in the field of hydraulic fracturing crack propagation monitoring due to its reliability and timeliness, surpassing other crack observation methods. When a material fractures under external forces, it releases microseismic / acoustic emission signals in the form of elastic waves. These signals are closely related to the initiation of internal cracks and the development of macroscopic cracks. Each signal contains a wealth of information about the internal fractures of the rock mass.

[0006] For example, relevant technical solutions include: "Hydraulic Fracturing Monitoring Method and Device" (Application No.: CN201910680539.9), "A Microseismic Monitoring Sensor Arrangement Method for Coal Roof Hydraulic Fracturing" (Application No.: CN202411693891.3), and "A Coal Face End Suspended Top Hydraulic Fracturing Method and System Equipment" (Application No.: CN202411341738.4). However, these solutions suffer from one or more deficiencies: Irrational sensor arrangement results in low positioning accuracy, making it impossible to accurately determine the spatial location of microseismic events; a single-layer sensor arrangement results in a limited monitoring range, making it impossible to achieve full coverage monitoring of microseismic events around the working face; and most critically, in complex underground environments, it is difficult to accurately distinguish microseismic signals generated by hydraulic fracturing and coal mining activities, resulting in low hydraulic fracturing monitoring accuracy; and a lack of quantitative evaluation methods for hydraulic fracturing effectiveness makes it difficult to objectively assess crack propagation and fracturing effectiveness.

[0007] Therefore, there is an urgent need for a technical solution that can accurately monitor the expansion of hydraulic fractures. Summary of the Invention

[0008] In order to address the low accuracy of hydraulic fracturing detection caused by the difficulty in accurately distinguishing microseismic signals generated by hydraulic fracturing and coal mining activities in complex underground environments, the present application provides a hydraulic fracture monitoring system and method based on a network of underground microseismic sensors. By dividing the fracturing area and the mining area, the microseismic signals generated by hydraulic fracturing and coal mining activities can be effectively distinguished, and a microseismic event density cloud map is generated based on the kernel density estimation algorithm. Combined with the energy proportion and spatial dispersion, the real-time monitoring accuracy of hydraulic fracture expansion is improved.

[0009] One aspect of the present application provides a hydraulic fracture monitoring system based on a network of underground microseismic sensors, comprising: a plurality of microseismic sensors for collecting microseismic signals generated by rock fractures; a plurality of data acquisition substations for receiving and processing the microseismic signals; a data transmission substation for converting the microseismic signals processed by the data acquisition substations into optical signals; a microseismic server for performing microseismic event analysis based on the optical signals; a mine network switch for connecting the data transmission substations and the microseismic server for network data exchange; an in-mine ring main unit for connecting the mine network switch and the microseismic server to form a communication network of the system; the microseismic sensors, the data acquisition substations, and the data transmission substations are connected via mine sensor extension cables and mine cables; and the data transmission substations and the microseismic server are connected via mine optical fibers.

[0010] Another aspect of the present application provides a hydraulic fracture monitoring method based on a downhole microseismic sensor network, comprising: establishing a microseismic monitoring coordinate system, analyzing and adjusting the spatial positions of multiple microseismic sensors using Kijko's D-value criterion to obtain a reference coordinate system; arranging the microseismic sensors into an upper and lower layer structure according to the reference coordinate system to form a spatial network to collect microseismic signals generated by rock fracture; converting the microseismic signals into optical signals, and obtaining a basic event data set containing event type, energy, and source position in the reference coordinate system through waveform recognition, phase extraction, and source location calculation based on the optical signals; dividing the monitoring area into a fracturing area and a production area according to the basic event data set in the reference coordinate system; and extracting microseismic event characteristics of the fracturing area and the production area respectively; calculating the microseismic potential, apparent stress, energy index, and apparent volume parameter of each area according to the microseismic event characteristics as a microseismic activity parameter set characterizing the rock fracture characteristics; generating a microseismic event density cloud map in the reference coordinate system according to the microseismic activity parameter set and the regional division results, and performing hydraulic fracturing monitoring based on the density cloud map.

[0011] Furthermore, the upper layer depth is greater than the preset threshold , the depth of the lower layer is greater than the preset threshold ; yes N times, where N is a positive integer greater than 10; Preferably 70 to 90m, The microseismic event characteristics include the number of microseismic events based on timestamp statistics, energy values ​​calculated using the spectrum integration method, spatial distribution characteristics based on three-dimensional coordinates, and event type distribution based on moment tensor inversion.

[0012] Furthermore, a microseismic monitoring coordinate system is established. The spatial positions of multiple microseismic sensors are analyzed and adjusted using Kijko's D-value criterion to obtain a reference coordinate system. This includes: deploying an initial sensor array around the working face of a small coal pillar in a coal mine and establishing a rectangular coordinate system within a preset range of the working face, with the center of the coal pillar as the origin, the working face direction as the X-axis, the direction perpendicular to the working face as the Y-axis, and the direction perpendicular to the coal seam as the Z-axis; calculating the geometric configuration parameter D of the initial sensor array using Kijko's D-value criterion, where D is the determinant of the sensor geometric layout matrix. When D is less than a preset threshold, it indicates that the sensor spatial layout is unreasonable. The spatial positions of each sensor are adjusted until D exceeds the preset threshold; performing a calibration test based on the adjusted sensor positions to verify positioning accuracy. When the positioning accuracy meets the preset accuracy requirements, the established rectangular coordinate system is used as the reference coordinate system. The calibration test includes: generating an artificially excited microseismic signal at a known location; receiving the signal using a sensor network and calculating the source location; and comparing the error between the calculated location and the actual location. If the error is within the preset accuracy range, the sensor layout is reasonable.

[0013] In particular, microseismic location is essentially a spatial inversion problem. The basic principle is to calculate the spatial position of the earthquake source using the arrival time differences of microseismic waves received by multiple sensors. Positioning accuracy is directly related to the geometric arrangement of the sensors. Kijko's D-value criterion quantifies the rationality of sensor placement by calculating the determinant of the geometric configuration matrix. The D-value essentially measures the "volume" expansion of the sensor array in space. When sensors are arranged in a straight line or on a plane, the D-value approaches zero, and the uniqueness and accuracy of the solution to the positioning equation cannot be guaranteed.

[0014] Therefore, this application establishes a coordinate system with the center of the coal pillar as the origin. The center of the coal pillar is usually a relatively stable area in the working face. Using it as the origin can reduce systematic errors caused by geological deformation. The working face strike, the direction perpendicular to the working face, and the direction perpendicular to the coal seam are used as the X, Y, and Z axes, respectively. This aligns the coordinate system with the main direction of mining activities, facilitating subsequent analysis of the relationship between microseismic events and mining activities.

[0015] For the initial sensor placement, the geometric configuration parameter D is calculated using Kijko's D-value criterion. This calculation involves constructing a geometric matrix based on the sensor positions; calculating the determinant of this matrix; and comparing it with a preset threshold to determine the rationality of the sensor placement. If the D-value is less than the preset threshold, the spatial arrangement of the sensors is insufficient to provide good positioning accuracy, and the sensor positions need to be adjusted. Based on the D-value calculation results, an iterative optimization method is used to adjust the sensor positions, for example, by incorporating the roadway width and support structure as constraints.

[0016] Furthermore, a basic event data set containing event type, energy, and source location in a reference coordinate system is obtained, including: identifying effective event waveforms based on optical signals through short-time energy ratios; wherein, the short-time energy ratio is based on the principle of energy mutation detection. When a microseismic event occurs, the waveform energy will suddenly increase, at which point the system determines that an effective event waveform has been detected. This method can effectively suppress background noise interference in complex downhole environments and improve the signal-to-noise ratio of event identification. Based on the identified effective event waveforms, the AIC information criterion is used to identify and extract the P-wave and S-wave phase arrival times; wherein, as different types of elastic waves, P-waves and S-waves will produce different AIC minimum value features at the time of arrival. By identifying these feature points, the phase arrival information can be accurately extracted.

[0017] According to the time difference between the extracted P-wave and S-wave phase arrival times, the earthquake source position is calculated in the reference coordinate system by a nonlinear inversion algorithm; specifically, the nonlinear inversion algorithm is preferably a Levenberg-Marquardt iterative algorithm.

[0018] Based on the hypocenter location, the microseismic event type is determined through moment tensor inversion. This includes classifications of tension, shear, and combined types. Specifically, the moment tensor M is a 3×3 symmetric matrix that can be decomposed into: M = Miso + Mdev, where Miso is the isotropic component and Mdev is the deviatoric tensor component. Eigenvalue analysis reveals that tension-type events have a dominant positive eigenvalue, with the isotropic component dominating; shear-type events have two positive and one negative eigenvalues, with the deviatoric tensor component dominating; and combined events have mixed characteristics, reflecting a complex fracture mechanism. This classification directly reflects the physical mechanism of rock mass failure and provides a mechanical basis for distinguishing hydraulic fracturing from coal mining activities.

[0019] Based on the identified effective event waveforms, the spectral integration method is used to calculate the energy of microseismic events. This method avoids the interference of noise in time domain calculations and improves the accuracy of energy estimation. The earthquake source location, event type, and energy are used as the basic event data set.

[0020] Furthermore, the microseismic event characteristics of the fracturing area and the mining area are extracted respectively, including: according to the source position coordinates in the basic event data set, the fracturing area is delineated with the pre-designed directional long borehole hydraulic fracturing construction range in the reference coordinate system, and the mining area is delineated with the working face and the actual advancement progress; the microseismic events within the preset distance range from the fracturing borehole centerline are classified as fracturing area events; the number, energy and spatial coordinates of microseismic events in different time periods are extracted according to the timestamps to obtain the microseismic event characteristics of the fracturing area; the microseismic events in the mined area and the mining area of ​​the working face are classified as mining area events; the area is divided into the roof area and the non-roof area according to the Z coordinate of the source position; for the roof area, the number, energy and spatial coordinates of microseismic events in different time periods are extracted according to the timestamps to obtain the microseismic event characteristics of the mining area.

[0021] Furthermore, feature extraction is performed on the fracturing zone to obtain microseismic event characteristics for the fracturing zone. This includes: counting the number of microseismic events in the fracturing zone at different time periods based on the timestamps in the basic event dataset; by counting the number of events in different time periods, a time series model describing the dynamic development of the fracture can be constructed. The energy values ​​of the fracturing zone events in the basic event dataset are extracted and the energy distribution is calculated; the spatial coordinates of the fracturing zone events in the basic event dataset are obtained; the spatial coordinates reflect the geometric characteristics of the fracture and can be obtained through principal component analysis and kernel density estimation algorithms. The number, energy distribution, and spatial coordinates are combined to form the microseismic event characteristics of the fracturing zone; and a temporal correspondence is established between the pressure and flow parameters of the fracturing operation and the extracted microseismic event characteristics of the fracturing zone.

[0022] Furthermore, feature extraction is performed on the mining area to obtain the microseismic event characteristics of the mining area, including: dividing the microseismic events in the mining area into roof areas and non-roof areas in the reference coordinate system according to the height of the hypocenter position in the basic event data set; extracting the occurrence time series of microseismic events in the roof area from the basic event data set as the frequency feature; wherein, the time series reflects the periodicity, clustering and time correlation of the events, which can be used to identify the pattern of mining activities and microseismic responses, especially the abnormal time series when the roof is close to collapse.

[0023] The energy of microseismic events in the roof area is extracted from the basic event data set as a magnitude feature; wherein, the roof area in this application refers to a spatial area within a specific height range above the coal seam determined by the height (Z coordinate) of the microseismic event source position in the reference coordinate system. This area usually includes the direct roof and the basic roof, and is the main target area for hydraulic fracturing operations. In addition, from the perspective of rock mechanics, the roof area is a cantilever beam structure formed after coal seam mining, which is subjected to the action of its own weight and the pressure of the overlying rock formation, resulting in bending deformation and stress concentration. During the hydraulic fracturing process, the mechanical properties of the roof rock mass are artificially changed to control its fracture process and achieve orderly collapse of the roof. The frequency characteristics and magnitude characteristics are used as the microseismic event characteristics of the mining area.

[0024] Furthermore, based on the characteristics of microseismic events in each region, the microseismic potential, apparent stress, energy index, and apparent volume parameters were calculated for each region as a set of microseismic activity parameters characterizing rock mass failure. This includes: Based on the energy distribution characteristics of the fracturing region, the energy release rate is calculated by counting microseismic events of different energy levels to obtain the microseismic potential of the fracturing region, reflecting the energy release state of the rock mass under fracturing. The energy release rate directly quantifies the dynamic process of fracture propagation, overcoming the limitation of traditional methods that can only qualitatively describe fracture activity. For hydraulic fracturing, changes in the energy release pattern directly reflect the interaction mechanism between the fracturing fluid and the rock mass.

[0025] Based on the event types and energy values ​​in the basic event dataset, the focal mechanism solution algorithm calculates the apparent stress in the fractured area, reflecting the stress distribution state of the rock mass in the fractured area. The apparent stress parameter transforms the microscopic fracture signal into a macroscopic stress field distribution, filling the technical gap in traditional monitoring methods in obtaining deep stress states. During hydraulic fracturing, the redistribution of the stress field directly determines the dominant fracture propagation direction and is a key indicator for predicting fracture morphology. The focal mechanism solution algorithm, based on the calculation method of moment tensor inversion theory, solves the stress tensor at the earthquake source by analyzing the initial motion direction and amplitude ratio of the P and S waves of the microseismic events. Specifically, the 3×3 symmetric matrix is ​​decomposed into isotropic and deviatoric tensor components. The fracture type (tension, shear, or combined) is determined through eigenvalue analysis. The apparent stress parameter is then calculated to reflect the stress state of the rock mass at the moment of fracture.

[0026] Based on the timestamps and energy values ​​of the fractured area, a b-value analysis is constructed to calculate the energy index of the fractured area, reflecting the degree of rock failure caused by the fracture. The b-value analysis introduces fractal theory from statistical physics, which can distinguish between the development stages of network cracks and main cracks, solving the problem that traditional methods have difficulty identifying the structural characteristics of the fracture network. For roof control, the ability to predict whether a large main crack or a dense network of small cracks will form is directly related to the mode of roof collapse. The energy index is based on the b-value analysis and follows the Gutenberg-Richter law: , where b is an energy index parameter that reflects the energy-frequency distribution relationship of microseismic events. A high b value indicates that small-scale rupture events dominate, whereas a low b value indicates that large-scale rupture events are more numerous. a is a constant; N(E) represents the cumulative number of microseismic events with energy greater than or equal to E; and E is the energy of the microseismic event.

[0027] Based on the spatial coordinates of the events in the fracturing area, the kernel density estimation algorithm is used in the reference coordinate system to calculate the apparent volume parameters of the fracturing area, reflecting the expansion range of the hydraulic fracture. The kernel density estimation algorithm breaks through the limitations of the traditional geometric simplification model and can describe the complex and irregular fracture expansion morphology, solving the problem of inaccurate fracture spatial representation. Accurately grasping the fracture expansion range is the basis for evaluating whether the fracturing has achieved the expected design. The kernel density estimation algorithm, a non-parametric statistical method, estimates the spatial distribution probability density of events by calculating the weighted density of the microseismic events around each point in space. The algorithm uses a Gaussian kernel function to perform spatial interpolation on the microseismic events to generate a continuous density distribution, which is used to calculate the apparent volume parameters of the fracturing area and can describe the complex and irregular fracture expansion morphology.

[0028] In particular, the physical nature of hydraulic fracturing is a continuous process: energy injection → stress change → rock fracture → spatial expansion. For example, the energy accumulation (microseismic potential) caused by fracturing fluid injection inevitably leads to a redistribution of the stress field (apparent stress), which in turn determines the fracture mode (energy index) and spatial expansion morphology (apparent volume parameter).

[0029] Based on the magnitude characteristics of the roof region, the microseismic potential of the mining area is calculated using an energy accumulation analysis method, reflecting the energy release level of roof activity. Based on the types of roof region events in the basic event dataset, the apparent stress of the mining area is calculated using principal stress direction analysis, reflecting the changes in roof stress caused by mining activity. Based on the frequency and magnitude characteristics of the roof region, the energy index of the mining area is calculated, reflecting the intensity and scale of the roof rupture. Based on the spatial distribution of roof region events, the apparent volume parameters of the mining area are calculated using a volume expansion algorithm, reflecting the impact range of the roof activity. The microseismic potential, apparent stress, energy index, and apparent volume parameters of each region are combined to form a microseismic activity parameter set. The principal stress direction analysis determines the normal vector and slip direction of the rupture surface by analyzing the focal mechanism solutions of the microseismic events, and then infers the direction and magnitude of the regional principal stresses. This method, based on the Anderson fault theory, inverts the stress field distribution state of the mining area by statistically analyzing the rupture parameters of multiple microseismic events and calculates the apparent stress field parameters of the mining area. The volume expansion algorithm, based on the geometric statistics of the spatial distribution of microseismic events, estimates the impact range of rock mass failure by calculating the spatial envelope volume of microseismic events. Using a convex hull algorithm or ellipsoid fitting method, the algorithm calculates the equivalent fracture volume based on the three-dimensional coordinate distribution of microseismic events in the roof region. This reflects the impact range of mining activities on the roof and derives the apparent volume parameters of the mining area.

[0030] Furthermore, a density cloud map of microseismic events in a reference coordinate system is generated, and hydraulic fracturing monitoring is performed based on the density cloud map. This involves generating a density cloud map of microseismic events in a reference coordinate system using a kernel density estimation algorithm based on a set of microseismic activity parameters and the demarcation of fracturing and production areas. This map is then used to analyze the extent of roof rupture and the development of the fracture network. In particular, traditional microseismic monitoring focuses primarily on scattered individual events. This approach, through density cloud maps, transforms discrete microseismic event signals into a continuous three-dimensional density field representation, achieving a shift from "point-based monitoring" to "volume-based assessment." This shift enables the visualization of the holistic characteristics of the fracturing process, resolving the problem of traditional methods' inability to intuitively grasp the overall state of complex fracture networks.

[0031] Vertical sections of the density cloud map are extracted to analyze the dynamic changes in the microseismic event area during the advance of the coal cutter. Horizontal sections perpendicular to the coal seam are extracted to analyze the differences in microseismic event density between different lanes. Specifically, the vertical sections in this application reflect the spatiotemporal evolution of microseismic events under the influence of mining, while the horizontal sections reveal the differences in microseismic distribution between different lanes, achieving comprehensive monitoring that combines both dynamic and static aspects. Using the coal cutter advance direction as the main axis, the two processes of mechanical mining and hydraulic fracturing are organically linked.

[0032] Calculate the spatial distance from each microseismic event in the basic event data set to the small coal pillar working face of the coal mine; calculate the critical values ​​of the roof pressure distance and the pressure relief distance based on the spatial distance and the frequency characteristics and magnitude characteristics of the mining area; further divide the mining area into the pressure area and the pressure relief area based on the critical value; in particular, this application identifies two key areas in the mining stress field - the pressure area where stress is concentrated and the pressure relief area where stress is released, realizing the monitoring transition from homogeneous space to heterogeneous space. By analyzing the differences in fracturing effects under different mechanical environments, the expansion law of hydraulic fractures under different stress fields is revealed, providing a basis for the regional adjustment of fracturing parameters.

[0033] Based on the microseismic potential parameters, microseismic events with energy exceeding the preset threshold are regarded as large-energy events and marked in the density cloud map; based on the amount of energy in the microseismic activity parameter set, the energy share coefficient of the large-energy event is calculated; among them, the energy share coefficient is the ratio of the energy of the large-energy event to the total energy, reflecting the degree of large crack formation during the fracturing process.

[0034] Based on the apparent volume parameters and the marked density cloud map, the spatial distribution of large-energy events in the pressure-incoming and pressure-releasing areas is analyzed, and the spatial dispersion of large-energy events is calculated. The spatial dispersion is used to measure the uniformity of the spatial distribution of large-energy events, reflecting the coverage and uniformity of the fracture network.

[0035] The energy proportion coefficient calculated above is compared with the first preset threshold, and the spatial dispersion is compared with the second preset threshold. When the energy proportion coefficient is greater than the first preset threshold EPC and the spatial dispersion is greater than the second preset threshold, combined with the spatial distribution characteristics of the microseismic events displayed in the vertical and horizontal sections of the density cloud map, it is determined that the hydraulic fracturing has caused the roof to collapse in an orderly manner, and the real-time monitoring and evaluation of the expansion of the hydraulic fracture is realized. Among them, in this application, the first preset threshold ,in, is the total energy of the large-level event, The total energy of all microseismic events, the threshold EPC value range is: hard roof: 0.65 to 0.75; medium hard roof: 0.55 to 0.65; soft roof: 0.45 to 0.55. The second preset threshold ,in, is the Moran index, The SDI threshold ranges are: 0.65 to 0.80 for large-area uniform fracturing, 0.45 to 0.65 for directional fracture control, and 0.35 to 0.45 for targeted enhanced fracturing.

[0036] Specifically, the energy fraction reflects the effective utilization of fracturing energy, while the spatial dispersion characterizes the uniformity and coverage of the fracture network. The combination of these two metrics enables a comprehensive assessment of fracturing effectiveness. Setting a preset threshold and performing a dual-metric joint judgment transforms empirical judgments into quantitative decision-making criteria, significantly improving the objectivity and reliability of fracturing effectiveness assessments. Furthermore, the abstract concept of "orderly roof collapse" is transformed into specific energy fraction and spatial dispersion metrics, achieving a technical shift from qualitative description to quantitative characterization.

[0037] Compared with the existing technology, the advantages of this application are:

[0038] In traditional coal mine hydraulic fracturing monitoring, a single-layer sensor layout and empirical signal recognition are generally used. However, the unreasonable spatial arrangement of sensors leads to large positioning deviations, and it is impossible to effectively distinguish between microseismic signals caused by hydraulic fracturing and mining activities, resulting in low hydraulic fracturing monitoring accuracy.

[0039] This application first uses Kijko's D-value criterion to optimize the spatial arrangement of sensors and establishes a three-dimensional reference coordinate system with the center of the coal pillar as the origin; this coordinate system becomes the unified spatial reference framework for all subsequent microseismic event analyses, ensuring the spatial consistency of the entire monitoring system.

[0040] Secondly, based on the established reference coordinate system, a two-layer sensor space networking mode is designed, in which the upper layer has a depth of and the depth of the lower layer A specific proportional relationship of N times (N>10) is formed, forming a three-dimensional monitoring network with an ideal geometric coverage range, realizing full coverage monitoring of microseismic events in the vertical and horizontal directions around the working face, and effectively improving the capture rate of microseismic events.

[0041] Then, microseismic signals generated by rock fractures are collected through the upper and lower layers of sensors, and the collected signals are converted into optical signals for transmission. Based on the optical signals, short-time energy ratio recognition, AIC information criterion phase extraction and nonlinear inversion algorithm positioning are used. A monitoring area division method based on the propulsion position of the coal cutter head is proposed. Combined with source location and microseismic characteristic analysis, the microseismic signals generated by hydraulic fracturing and coal mining activities are effectively distinguished, reducing the signal misjudgment rate.

[0042] Finally, based on the distinguished microseismic event characteristics, the microseismic activity parameter set such as microseismic potential, apparent stress, energy index and apparent volume parameters is calculated. The microseismic event density cloud map generated by the kernel density estimation algorithm is combined with the two quantitative indicators of energy proportion coefficient and spatial dispersion. By comparing with the preset threshold, it can be determined whether hydraulic fracturing has caused orderly collapse of the roof, achieving a breakthrough from qualitative to quantitative monitoring of hydraulic fractures. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 This is a flow chart of a downhole microseismic sensing spatial network for monitoring hydraulic fractures in a specific embodiment of the present application;

[0044] Figure 2 This is a schematic diagram of a microseismic sensing spatial networking monitoring system according to a specific embodiment of the present application;

[0045] Figure 3 This is a diagram of the arrangement of microseismic monitoring sensors for directional long-hole hydraulic fracturing according to a specific embodiment of the present application.

[0046] Figure numerals: 1 - microseismic sensor, 2 - mining sensor extension cable, 3 - data acquisition substation, 4 - data acquisition substation power supply, 5 - mining optical fiber, 6 - mining cable, 7 - data transmission substation, 8 - data transmission substation power supply, 9 - mining network switch, 10 - mine ring network cabinet, 11 - microseismic server. DETAILED DESCRIPTION

[0047] The present application is described in detail below with reference to the accompanying drawings and specific embodiments.

[0048] To maximize coal resource recovery, a certain mine implemented measures to optimize the size of coal pillars in one of its working faces, reducing the size of the coal pillar between the belt conveyor chute and the auxiliary conveyor chute from the original design of 19m to 8.5m. However, in small-pillar working faces where two lanes are excavated simultaneously, the secondary reused auxiliary transport lanes commonly experience large tunnel deformation and floor heave. Areas with severe floor heave generally require floor excavation. To ensure mining safety at the working face and the stability of the coal pillars and surrounding rock of the lanes, while optimizing the coal pillar size, pressure relief measures are required to minimize the pressure caused by changes in coal pillar size. Therefore, directional long-hole hydraulic fracturing is performed in the forward direction of the mining area of ​​this working face to pre-crack the roof, improving mining efficiency while ensuring mining safety.

[0049] like Figure 1As shown, the hydraulic fracture monitoring method of the present application includes: establishing a microseismic monitoring coordinate system, using Kijko's D value criterion to analyze and adjust the spatial positions of multiple microseismic sensors to obtain a reference coordinate system; according to the reference coordinate system, the microseismic sensors are arranged into an upper and lower two-layer structure to form a spatial network to collect microseismic signals generated by rock fracture; the microseismic signals are converted into optical signals, and based on the optical signals, a basic event data set containing event type, energy and source position in the reference coordinate system is obtained through waveform recognition, phase extraction and source location calculation; according to the basic event data set, the monitoring area is divided into a fracturing area and a mining area in the reference coordinate system; and the microseismic event characteristics of the fracturing area and the mining area are respectively extracted; according to the microseismic event characteristics of each area, the microseismic potential, apparent stress, energy index and apparent volume parameter of each area are respectively calculated as a microseismic activity parameter set characterizing the rock fracture characteristics; according to the microseismic activity parameter set and the regional division result, a microseismic event density cloud map in the reference coordinate system is generated, and hydraulic fracturing monitoring is performed based on the density cloud map.

[0050] Specifically, such as Figure 2 As shown in Figure 1, the monitoring system is constructed. Its hardware includes: a microseismic server 11, a microseismic sensor 1, a data acquisition substation 3, a data transmission substation 7 (containing a photoelectric converter), a mining sensor extension cable 2, a mining cable 6, and a mining optical fiber 5. The post-processing software primarily includes: high-precision data acquisition instrument embedded software (DST), high-precision microseismic waveform real-time display software (WDD), high-precision microseismic signal automatic analysis software (GMS), and high-precision microseismic information 3D simulation software (VisGMD).

[0051] Based on the principles of seismology, microseismic signals generated by rock fractures can be used to invert underground rock deformation and crack expansion. Microseismic signals, generated by high-pressure water injection to fracture the rock, have propagation characteristics similar to seismic waves. Analysis can reveal information such as the earthquake source location, type, and energy.

[0052] Specifically, the drilling location was determined and, based on this location, microseismic sensors 1 were installed. Simultaneously, transmission lines such as the mining sensor extension cable 2, mining cable 6, and mining optical fiber 5 were connected, and the data acquisition substation 3 and data transmission substation 7 were debugged to ensure the proper operation of the entire hardware system. The drilling location was determined and, based on this location, microseismic sensors 1 and other equipment were installed. Considering the characteristics of hydraulic fracturing operations, high-sensitivity velocity-type microseismic sensors 1 with a sensitivity of 220 V / m / s were selected. Two data acquisition substations 3 were located in the connecting lanes, and data transmission substation 7 was located in the connecting lanes, in the same location as one of the data acquisition substations 3.

[0053] The microseismic sensor 1, data acquisition substation 3, and data transmission substation 7 are connected via a mine sensor extension cable 2 and a mine cable 6. The data transmission substation and microseismic server 11 are connected via a mine optical fiber 5. A data acquisition substation power supply 4 supplies power to the data acquisition substation 3; a data transmission substation power supply 8 supplies power to the data transmission substation 7. A mine ring main unit 10 connects the mine network switch 9 and the microseismic server 11, forming the system's communication network.

[0054] During installation, ensure that the sensor's installation angle and depth meet design requirements to ensure accurate microseismic signal acquisition. Debug data acquisition substation 3 and data transmission substation 7 to ensure the proper functioning of the entire hardware system. A complete microseismic sensing spatial network monitoring system consists of 11 components.

[0055] The rationality of the spatial arrangement of sensors is analyzed by applying Kijko's D-value criterion, where the D-value is the determinant of the sensor geometric arrangement matrix, which characterizes the "volume" expansion of the sensor array in space. In this embodiment, 16 microseismic sensors 1 are divided into 4 groups, with an interval of 80m between two groups. Each group is equipped with 4 sensors with different parameters. The sensor arrangement area extends from the belt conveyor chute to the auxiliary conveyor chute. The sensors are arranged in groups of 4, with a group spacing of 80m. The sensor installation depth is about 80m for the upper layer and about 6m for the lower layer. Inclined holes are used for installation on the adjacent roadway side to ensure coverage of the fracturing area and improve monitoring accuracy. Figure 3 As shown in the figure, the green sphere represents the designed installation position of the sensor. Preferably, the sensor is a high-sensitivity velocity-type microseismic sensor 1 (with a measurable range of 10Hz to 2000Hz and a sensitivity of 220V / m / s).

[0056] To establish coordinates and analyze errors, and to establish a reference coordinate system that better meets the needs of mechanical analysis, we use the center of the small coal pillar as the origin, the direction of the auxiliary transport chute as the positive direction of the X-axis, the direction perpendicular to the auxiliary transport chute pointing to the belt transport chute as the positive direction of the Y-axis, and the direction perpendicular to the horizontal plane of the roadway as the positive direction of the Z-axis. Specifically, in this application, a small coal pillar working face refers to a coal mining working face that uses an optimized coal pillar size design to improve resource recovery rate by reducing the coal pillar size.

[0057] During the hydraulic fracturing process, Figure 2As shown, microseismic sensor 1 collects microseismic signals generated by rock fractures in real time and transmits them to data acquisition substation 3. After preliminary processing of the signals by data acquisition substation 3, the signals are converted into optical signals by data transmission substation 7 and transmitted via optical fiber to microseismic server 11 on the surface. During the signal acquisition phase, the system uses a short-time energy ratio algorithm to identify valid event waveforms, setting the short-time window to 10ms and the long-time window to 100ms. When the energy ratio of the short-time window to the long-time window exceeds a preset threshold of 3.5, the system identifies the event as a valid microseismic event. This method effectively suppresses background noise interference in the downhole environment. For the identified valid event waveforms, the system uses the Akaike Information Criterion (AIC) method to accurately extract the arrival times of the P-wave and S-wave phases. This method can produce significant AIC minima at the arrival times of different wave phases, improving the accuracy of phase identification.

[0058] The software system in the microseismic server 11 further analyzes and processes the collected data. The high-precision data acquisition instrument embedded software (DST) monitors and sets the parameters of the data acquisition instrument and timing server. Data acquisition instrument and timing server parameters can be browsed, set, and diagnosed remotely using a web browser. The high-precision microseismic waveform real-time display software (WDD) displays the waveform signals picked up by the sensors in real time and is primarily used for debugging the sensor's waveform noise floor and for engineering construction and maintenance. The high-precision microseismic signal automatic analysis software (GMS) performs in-depth data analysis, calculating the earthquake source location using the Levenberg-Marquardt nonlinear inversion algorithm and determining the microseismic event type (tension, shear, and combined) through moment tensor inversion.

[0059] During moment tensor inversion, the system decomposes a 3×3 symmetric matrix into isotropic and deviatoric tensor components. Eigenvalue analysis is used to distinguish different fracture types: tension (primary positive eigenvalue, with the isotropic component dominating), shear (two positive and one negative eigenvalue, with the deviatoric tensor component dominating), and composite (a combination of features reflecting complex fracture mechanisms). The system also calculates the energy of each microseismic event using a spectral integration method, avoiding noise interference in time-domain calculations and improving the accuracy of energy estimates. The high-precision 3D microseismic information simulation software (VisGMD) provides a 3D visualization of information such as earthquake source location, event type, and energy, forming a basic event dataset that allows personnel to intuitively understand the development of hydraulic fractures.

[0060] Microseismic activity analysis: First, the monitoring area is divided into fracturing areas and mining areas. Microseismic events in the fracturing area are considered to be events generated by fracturing operations, while microseismic events in the mining area are mainly affected by coal mining activities. Feature extraction is performed on the fracturing area: Based on the timestamps in the basic event dataset, the number of microseismic events in different time periods is counted; the energy values ​​of the events in the fracturing area are extracted and the energy distribution is calculated; and the spatial coordinates of the events in the fracturing area are obtained to reflect the geometric characteristics of the fractures. Combined with the fracturing operation pressure and flow parameters, a temporal correspondence is established with the microseismic event characteristics. For example, during a certain day and a certain shift, different microseismic events were observed, such as growth and clustering, when the pressure and flow were stable or changing. This indicates a corresponding relationship between the two. However, due to the complex geology, the growth rates are not completely consistent, and the event clustering area is not necessarily in the construction drilling area, but within the fracturing area.

[0061] For the mining area, events were divided into the roof area (Z coordinate greater than the coal seam roof height) and the non-roof area based on the elevation of the earthquake source. The time series of microseismic events in the roof area were extracted as frequency features, and the energy was extracted as the magnitude feature. Through 24-hour uninterrupted monitoring, the number of microseismic events and energy release differences on different dates were analyzed. It was found that microseismic events can remain active even during non-construction periods. This indicates that although construction is the main cause, cracks will continue to expand due to stress redistribution after construction. Therefore, the evaluation of fracturing effectiveness should include microseismic events in the period before and after construction.

[0062] Study the law of roof rupture activity: According to the characteristics of microseismic events in each area, calculate the microseismic activity parameter set. For the fracturing area, calculate the energy release rate by counting microseismic events of different energy levels to obtain the microseismic potential (in the observation period ), reflecting the energy release state of the rock mass under fracturing; the apparent stress (average value 32MPa) was calculated using the focal mechanism solution algorithm, reflecting the stress distribution state of the rock mass in the fracturing area; the energy index (b value is 1.24) was calculated by constructing the b-value analysis, following the modified Gutenberg-Richter law, reflecting the degree of rock mass rupture caused by fracturing; the apparent volume parameter (affected range is approximately 2800m³) was calculated using the kernel density estimation algorithm, reflecting the extension range of the hydraulic fracture.

[0063] For the mining area, microseismic potential was calculated using energy accumulation analysis, apparent stresses were calculated using principal stress direction analysis, energy indices were calculated based on frequency and magnitude characteristics, and apparent volume parameters were calculated using a volume expansion algorithm. It was found that the number of large-magnitude events was positively correlated with the microseismic event rate, and that the fracturing operation areas generally coincided with the active roof areas. The study showed that in areas with sufficient fracturing, while roof rupture activity was active, the proportion of large-magnitude events was significantly reduced, indicating that hydraulic fracturing can effectively reduce the risk of large-scale sudden roof ruptures.

[0064] The spatial distance from each microseismic event in the basic event dataset to the small coal pillar working face of the coal mine was calculated. Combined with the frequency characteristics and magnitude characteristics of the mining area, the critical value of the roof pressure distance was obtained to be 34m and the critical value of the pressure relief distance was 17m. Based on this, the mining area was further divided into the pressure area and the pressure relief area, providing a basis for the regional adjustment of fracturing parameters.

[0065] Fracturing Effect Evaluation: Based on the microseismic activity parameter set and regional delineation results, a kernel density estimation algorithm was used to generate a microseismic event density cloud map in the reference coordinate system. This algorithm calculates the weighted density of surrounding microseismic events at each point in space, breaking through the limitations of traditional geometric simplification models and capable of describing complex and irregular fracture propagation patterns. Vertical sections of the density cloud map were extracted along the roadway to analyze the dynamic changes in the microseismic event area during coal cutter advancement. Horizontal sections were extracted perpendicular to the coal seam to analyze the differences in microseismic event density between different inter-roadways.

[0066] The energy exceeds The microseismic events were marked as large-energy events and were specially marked in the density cloud map. The calculated energy proportion coefficient of large-energy events (the ratio of large-energy event energy to total energy) was 0.67, exceeding the preset threshold of 0.65 for moderately hard roof. The spatial dispersion of large-energy events (expressed as the product of 1-Moran's index and the actual impact volume ratio) was 0.74, exceeding the preset threshold of 0.65 for large-area uniform fracturing. Both indicators exceeded the threshold. Combined with the spatial distribution characteristics of microseismic events shown in the density cloud map, it was determined that the hydraulic fracturing had caused orderly roof collapse and the fracturing effect was good.

[0067] Principal component analysis was used to further analyze the correlation between fracturing parameters and microseismic events. Extracting two principal components, one for fracturing design and the other for fracturing operation, revealed a significant positive correlation between pressure parameters and the number of microseismic events (correlation coefficient r = 0.78), and a moderate positive correlation between flow parameters and microseismic event energy (correlation coefficient r = 0.62). This analysis indicates that, under the geological conditions of this region, increasing fracturing parameters is more effective in improving fracturing results. In subsequent construction, appropriate increases in fracturing parameters and adjustments to drillhole locations based on density contours will optimize fracturing results.

[0068] The above schematically describes the invention of the present application and its implementation methods. This description is not restrictive. Without departing from the spirit or basic features of the present application, the present application can be implemented in other specific forms. What is shown in the accompanying drawings is only one of the implementation methods of the invention of the present application. The actual structure is not limited to this, and any figure marks should not limit the claims involved. Therefore, if a person of ordinary skill in the art is inspired by it, without departing from the purpose of the present invention, a structural method and embodiment similar to the technical solution are designed without creativity, which should fall within the scope of protection of the present application. In addition, the word "including" does not exclude other elements or steps, and the word "one" before an element does not exclude the inclusion of "multiple" elements. Words such as first and second are used to indicate names and do not indicate any specific order.

Claims

1. A hydraulic fracture monitoring method based on downhole microseismic sensor networking, characterized in that: include: Establish a microseismic monitoring coordinate system, use Kijko's D-value criterion to analyze and adjust the spatial positions of multiple microseismic sensors, and obtain a reference coordinate system; According to the reference coordinate system, the microseismic sensors are arranged in a two-layer structure to form a spatial network to collect microseismic signals generated by rock fractures. The microseismic signals are converted into optical signals. Based on the optical signals, waveform recognition, phase extraction and source location calculation are performed to obtain a basic event dataset containing event type, energy and source location in the reference coordinate system. Based on the basic event data set, the monitoring area is divided into the fracturing area and the production area in the reference coordinate system; And extract the microseismic event characteristics of the fracturing area and the production area respectively; According to the characteristics of microseismic events in each region, the microseismic potential, apparent stress, energy index and apparent volume parameters of each region are calculated as a set of microseismic activity parameters to characterize the characteristics of rock mass rupture; Based on the microseismic activity parameter set and regional division results, a microseismic event density cloud map in the reference coordinate system is generated, and hydraulic fracturing monitoring is carried out based on the density cloud map.

2. The method for monitoring hydraulic fractures based on a downhole microseismic sensor network according to claim 1, characterized in that: The upper layer depth is greater than the preset threshold , the depth of the lower layer is greater than the preset threshold ; yes N times, where N is a positive integer greater than 10; Microseismic event characteristics include the number of microseismic events based on timestamp statistics, energy values ​​calculated based on spectral integration methods, spatial distribution characteristics based on three-dimensional coordinates, and event type distribution based on moment tensor inversion.

3. The hydraulic fracture monitoring method based on downhole microseismic sensor networking according to claim 1 or 2, characterized in that: Establish a microseismic monitoring coordinate system, including: An initial sensor array is deployed around the working face of a small coal pillar in a coal mine, and a rectangular coordinate system is established with the center of the coal pillar as the origin, the working face direction as the X-axis, the direction perpendicular to the working face as the Y-axis, and the direction perpendicular to the coal seam as the Z-axis. The geometric configuration parameter D value of the initial sensor array is calculated using Kijko's D value criterion, where D value is the determinant of the sensor geometric arrangement matrix; When the D value is less than the preset threshold, it indicates that the spatial arrangement of the sensors is unreasonable. The spatial position of each sensor is adjusted until the D value is greater than the preset threshold. According to the adjusted sensor position, a calibration test is carried out to verify the positioning accuracy. When the positioning accuracy meets the preset accuracy requirements, the established rectangular coordinate system is used as the reference coordinate system.

4. The method for monitoring hydraulic fractures based on a downhole microseismic sensor network according to claim 3, characterized in that: Obtain a basic event dataset containing event type, energy, and source location in a reference coordinate system, including: According to the optical signal, the effective event waveform is identified through the short-time energy ratio algorithm; According to the recognized effective event waveforms, the arrival times of P-wave and S-wave phases are identified and extracted by the AIC information criterion method; The earthquake source position is calculated in the reference coordinate system using the Levenberg-Marquardt nonlinear inversion algorithm based on the time difference between the extracted P-wave and S-wave phases. According to the earthquake source location, the type of microseismic event is determined by moment tensor inversion, wherein the microseismic event type includes tension type, shear type and composite type; According to the identified effective event waveform, the energy of microseismic events is calculated using the spectrum integration method. The earthquake source location, microseismic event type and microseismic event energy are used as the basic event data set.

5. The method for monitoring hydraulic fractures based on downhole microseismic sensor networking according to claim 1, characterized in that: Extract microseismic event characteristics of the fracturing area and the production area respectively, including: According to the earthquake source location in the basic event data set, the monitoring area is divided into the fracturing area and the production area in the reference coordinate system; For the fracturing area, the number of microseismic events, the energy of microseismic events and the spatial coordinates of the corresponding microseismic events in different periods are extracted according to the timestamps to obtain the microseismic event characteristics of the fracturing area; The mining area is divided into roof area and non-roof area according to the Z coordinate of the earthquake source position; For the roof area, the number of microseismic events, the energy of microseismic events and the spatial coordinates of the corresponding microseismic events in different periods are extracted according to the timestamps as the microseismic event characteristics of the mining area.

6. The method for monitoring hydraulic fractures based on downhole microseismic sensor networking according to claim 5, characterized in that: Obtain microseismic event characteristics in the fracturing area, including: According to the timestamps in the basic event dataset, the number of microseismic events in the fracturing area at different times is counted; Extract the energy of microseismic events in the fracturing area from the basic event data set and calculate the energy distribution; Obtain the spatial coordinates of microseismic events in the fracturing area in the basic event dataset; The number, energy distribution and spatial coordinates are combined as the characteristics of microseismic events in the fractured area; A time correspondence is established between the pressure and flow parameters of the fracturing operation and the extracted microseismic event characteristics of the fracturing area.

7. The method for monitoring hydraulic fractures based on downhole microseismic sensor networking according to claim 6, characterized in that: Obtain microseismic event characteristics in the mining area, including: According to the Z coordinate of the earthquake source position in the basic event data set, the microseismic events in the mining area are divided into roof area and non-roof area; Extract the occurrence time series of microseismic events in the roof area from the basic event dataset as the frequency feature; The energy of microseismic events in the roof region is extracted from the basic event dataset as the magnitude feature; The frequency characteristics and magnitude characteristics are used as the characteristics of microseismic events in the mining area.

8. The method for monitoring hydraulic fractures based on downhole microseismic sensor networking according to claim 7, characterized in that: The microseismic activity parameter set that characterizes the rock mass rupture characteristics is obtained, including: For the fracturing area, the energy release rate is calculated by counting the number of microseismic events at different energy levels to obtain the microseismic potential parameter; the apparent stress parameter is calculated using the focal mechanism solution algorithm; the energy index parameter is calculated by b-value analysis based on the Gutenberg-Richter law; and the apparent volume parameter is calculated using the kernel density estimation algorithm. Among them, the b-value is the energy index parameter, which reflects the energy-frequency distribution relationship of the microseismic event. For the mining area, the microseismic potential parameters are calculated by energy accumulation analysis; the apparent stress field parameters are calculated by principal stress direction analysis; the energy index parameters are calculated by b-value analysis; and the apparent volume parameters are calculated by volume expansion algorithm based on the spatial distribution of events in the roof area. The microseismic potential, apparent stress, energy index and apparent volume parameters of the fracturing area and the production area are combined to obtain a set of microseismic activity parameters that characterize the rock mass failure characteristics.

9. The method for monitoring hydraulic fractures based on downhole microseismic sensor networking according to claim 8, characterized in that: Hydraulic fracturing monitoring based on density cloud maps, including: Based on the microseismic activity parameter set and the division results of the fracturing area and the production area, a microseismic event density cloud map in the reference coordinate system is generated using the kernel density estimation algorithm; Extract vertical and horizontal sections of the density cloud map. The vertical section is used to analyze the dynamic changes of the microseismic event area, and the horizontal section is used to analyze the planar distribution of microseismic events. Based on the earthquake source location in the basic event data set, the spatial distance from each microseismic event to the small coal pillar working face of the coal mine is calculated. Combined with the frequency characteristics and magnitude characteristics of the mining area, the critical values ​​of the roof pressure distance and the pressure relief distance are calculated. Based on the critical values, the monitoring area is divided into the pressure area and the pressure relief area. The microseismic events whose energy exceeds the preset threshold in the basic event dataset are marked as high-energy events; Calculate the ratio of the energy of large energy events to the total energy as the energy ratio coefficient; According to the large energy level events, the spatial dispersion coefficient of the monitoring area is calculated; When both the energy proportion coefficient and the spatial dispersion coefficient exceed the threshold, it is determined that hydraulic fracturing causes orderly roof collapse.

10. A hydraulic fracture monitoring device based on the method according to any one of claims 1 to 9, characterized in that: include: A plurality of microseismic sensors (1) for collecting microseismic signals generated by rock fracture; Multiple data acquisition substations (3) to receive and process microseismic signals; A data transmission substation (7) converts the microseismic signal processed by the data acquisition substation (3) into an optical signal; A microseismic server (11) performs microseismic event analysis based on the optical signal; A mining network switch (9) is connected to the data transmission substation (7) and the microseismic server (11) for network data exchange; The ring network cabinet (10) in the mine is connected to the mine network switch (9) and the micro-seismic server (11) respectively, forming a communication network of the system; The microseismic sensor (1), the data acquisition substation (3) and the data transmission substation (7) are connected via a mining sensor extension line (2) and a mining cable (6); The data transmission substation (7) and the microseismic server (11) are connected via a mining optical fiber (5).

Images