Deep fracture network description and three-dimensional reconstruction method based on multi-source information fusion

Abstract

The invention relates to the technical field of geothermal development, and discloses a multi-source information fusion deep fracture network description and three-dimensional reconstruction method, which comprises the following steps: S1, carrying out fusion analysis on surface dominant fractures and deep hidden fractures; s2, determining a rock fracture mechanism by using a laboratory method; s3, fracture parameters are obtained after engineering fracturing; s4, establishing a multi-field coupled fracture expansion model to describe a deep reservoir fracture network; and S5, carrying out model verification. According to the invention, by integrating geological data of different scales and types, the integrity and accuracy of fracture network description are improved. And a rock fracture mechanism is researched, and the accuracy of fracture network formation is determined from the aspect of mechanical causes. In addition, three-dimensional visualization of the crack network is realized, and visual crack network structure representation is provided. The method provided by the invention is simple, the depicting precision is improved by a comprehensive method, and the cost can be reduced.

Description

Technical Field

[0001] The present invention relates to the field of geothermal development technology, and more specifically discloses a method for deep fracture network characterization and three-dimensional reconstruction by multi-source information fusion. Background Art

[0002] With the growth of global energy demand and the concern for environmental sustainability, the development and utilization of renewable energy has become an important part of the energy strategy of countries around the world. Geothermal energy, as a clean and sustainable form of energy, has the characteristics of good stability, low environmental impact and wide applicability. It is one of the important energy sources to alleviate the energy crisis and reduce greenhouse gas emissions. The total amount of hydrothermal geothermal resources in my country with a burial depth of less than 3 kilometers is equivalent to 1.25 trillion tons of standard coal, and the deep geothermal resources with a burial depth of 3 kilometers to 10 kilometers are equivalent to 856 trillion tons of standard coal. Deep geothermal resources belong to a new field with larger resources, higher temperatures, better resource endowments, and wider application scenarios, and have great development potential.

[0003] Deep heat reservoirs have complex structures and variable scales of fracture networks, and interact with various underground physical fields (such as stress field, seepage field, temperature field, etc.), which affect the migration of underground fluids and the exploitation of geothermal resources. Traditional fracture network characterization methods mostly rely on a single information source, such as geophysical exploration, direct field observation, or downhole imaging. These methods may work well when dealing with fracture networks of a single scale and simple structure, but they are obviously insufficient when facing multi-scale and highly complex fracture systems. A single information source is often difficult to fully reflect the three-dimensional structure of the fracture network, resulting in limited accuracy and reliability of reconstruction. The use of multi-source information fusion technology provides new possibilities for the study of fracture networks. By integrating multiple information sources such as ground and underground geophysical exploration data, geological drilling data, and microseismic monitoring data, the fracture network can be comprehensively understood at a larger scale and higher resolution. Accurately characterizing and reconstructing the fracture network of deep heat reservoirs is of great significance for revealing the distribution law of heat reservoirs and the mechanism of fracture extension, reservoir transformation, safe production increase, comprehensive evaluation, and comprehensive development of power generation, heating, and energy storage. Summary of the invention

[0004] The present invention mainly provides a method for deep fracture network characterization and three-dimensional reconstruction by multi-source information fusion, which can solve the problem that the reconstruction accuracy and reliability of traditional fracture network characterization methods are relatively limited.

[0005] To solve the above technical problems, according to one aspect of the present invention, more specifically, a method for deep fracture network characterization and three-dimensional reconstruction based on multi-source information fusion, comprising the following steps:

[0006] S1. Conduct fusion analysis of surface explicit fractures and deep implicit fractures: Collect surface outcrops, drilling data, and seismic data in the study area, integrate these data, compare the formation evolution and distribution law of explicit fractures with the development law of deep implicit fractures, and infer and characterize the deep fracture network in combination with the geological structure evolution process;

[0007] S2. Determine the rock fracture mechanism using laboratory methods: simulate the crack morphology and formation mechanism formed by rock compression, shearing and splitting under reservoir temperature and pressure conditions, and test various mechanical parameters of rocks; conduct true triaxial hydraulic fracturing tests, simulate the artificial water injection fracturing process, and determine the crack evolution process from the perspective of mechanical causes;

[0008] S3. Acquisition of fracture parameters after engineering fracturing: Acquisition of fracture parameters after fracturing through microseismic monitoring data combined with remote detection sonic logging and conventional logging data;

[0009] S4. Establish a multi-field coupled fracture propagation model to characterize the fracture network in deep reservoirs;

[0010] S5. Model verification: Use the generated fracture network model to perform seepage simulation, compare the calculated results with the actual single and paired well injection and production test data, and verify the accuracy of the deep fracture network model.

[0011] Furthermore, the step S1 specifically includes:

[0012] S11. Select the outcrops in the surrounding study area with the same lithology or similar structure as the development site, conduct natural fracture surveys, collect and statistically analyze the direction, dip angle and trace length of natural fractures, and determine the development and distribution patterns of surface explicit fractures;

[0013] S12. Collect the natural fracture inclination and dip angle information of the cores drilled in the study area, and perform XRD analysis on the altered mineral composition of the cores to determine the well sections with dense natural fracture development in the vertical range;

[0014] S13. Collect logging data, identify natural fractures, and analyze information such as the inclination, dip, density, aperture, porosity, etc. of natural fractures, and draw fracture dip and dip rosette diagrams;

[0015] S14. Collect 3D seismic data, extract ant volume, coherence volume, and curvature volume data according to reservoir characteristics, and determine the spatial distribution law of deep hidden fractures;

[0016] S15. Compare and analyze multiple groups of fracture statistics from different sources, and preliminarily infer the characteristics of the fracture network in deep three-dimensional space by combining the shallow obvious fractures with the hidden fractures in the vertical range around the well.

[0017] Furthermore, the step S2 specifically includes:

[0018] S21. Carry out multiple sets of uniaxial compression tests at room temperature to test the Young's modulus, Poisson's ratio, and compressive strength of the rock, and record the development morphology of the cracks when the rock is damaged;

[0019] S22. On the basis of uniaxial test, the influence of confining pressure, temperature and the combined effect of temperature and pressure are added respectively to test and analyze the changes of rock mechanical properties, record the rock failure morphology, and summarize the change law of rock mechanical properties;

[0020] S23. Conduct true triaxial hydraulic fracturing tests to evaluate the expansion of hydraulic fractures by combining the acoustic emission characteristics of rock samples with the injection pressure.

[0021] Furthermore, the step S3 specifically includes:

[0022] S31. Collect microseismic monitoring related data, including seismic waveform data, earthquake source location and seismic wave propagation path information;

[0023] S32, randomly selecting a certain number of sample data points from the microseismic data to initialize the fracture parameter estimation;

[0024] S33, using the selected sample data points, performing parameter estimation according to the crack model to obtain a fitting model of the crack;

[0025] S34. Statistically calculate the average distance μ and variance σ of all sample data points to the fitting model, and calculate the distance d of each sample data point to the fitting model. The sample data points that meet the following formula are set as internal points, which have a good fit with the model. Otherwise, they are outliers, which have a poor fit with the model.

[0026] |d-μ|≤σ;

[0027] S35. If the number of internal points of the current fitting model exceeds the set threshold, the current model is considered to be valid and the iteration is stopped; otherwise, steps S32-S33 are repeated;

[0028] S36, refitting the model using all the internal point data to obtain the final fracture parameter estimation and generate the optimal solution for this fracture;

[0029] S37, removing the data points where the cracks are successfully generated in the microseismic data set, and continuing steps S32-S36 until the next crack parameter is successfully obtained.

[0030] Furthermore, the step S4 specifically includes:

[0031] S41, establishing a natural fracture network model by Monte Carlo method based on the fracture characteristic parameters in steps S1 and S2;

[0032] S42, coupling rock physical parameters, performing geomechanical coarsening, and simulating reservoir in-situ stress distribution;

[0033] S43, performing numerical simulation of fracture extension according to actual engineering parameters to obtain fracture reconstruction volume after engineering fracturing;

[0034] S44, using the method described in step S3, reconstructing the post-fracturing fracture network through microseismic monitoring data observed in actual engineering, and calculating the fracture volume parameters;

[0035] S45, performing error analysis by comparing step S43 with step S44, and adjusting model parameters to optimize the fracture network model;

[0036] S46. Generate a final deep reservoir fracture network model.

[0037] The beneficial effects of the method for deep fracture network characterization and three-dimensional reconstruction of multi-source information fusion of the present invention are as follows: the present invention improves the integrity and accuracy of fracture network description by integrating geological data of different scales and types. In addition, the rock fracture mechanism is studied to determine the accuracy of fracture network formation from the perspective of mechanical causes. In addition, three-dimensional visualization of the fracture network is achieved to provide an intuitive fracture network structure representation. The method provided by the present invention is simple, the comprehensive method improves the characterization accuracy, and can reduce costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The present invention is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0039] Figure 1 is a schematic flow chart of step S1;

[0040] Figure 2 is a schematic diagram of the process of step S3;

[0041] Figure 3 is a schematic flow chart of step S4;

[0042] Figure 4 This is the depiction and reconstruction of deep fractures according to the present invention. DETAILED DESCRIPTION

[0043] The present invention will be described in detail below with reference to the accompanying drawings and in combination with embodiments. It should be noted that the embodiments and features in the embodiments of the present application can be combined with each other without conflict.

[0044] According to one aspect of the present invention, a method for deep fracture network characterization and three-dimensional reconstruction based on multi-source information fusion is provided. Taking a hot dry rock development and utilization site as an example, the deep reservoir range is between 3500-4000 meters. The site has carried out hydraulic fracturing field tests, and the reservoir lithology is granite. To characterize the deep fracture network, the following steps are included:

[0045] S1. Conduct a survey of natural fissures, collect and compile information on the direction, dip angle, and trace length of natural fissures, and determine the development and distribution patterns of surface (shallow) explicit fissures. Figure 1 The specific steps include:

[0046] S11 conducted a natural fracture investigation around the development site. The density of natural fractures was found to be between 2-3 / m and 7-10 / m. The fractures were mainly oriented in the northeast and northwest directions with an inclination of 70-90 degrees, and were mainly high-angle fractures.

[0047] S12 analysis of the core fractures in the well showed that the core was highly developed at high angles, with multiple stages of hydrothermal alteration relatively developed, and some biotite-chlorite and epidote-induced changes, indicating that the fractures were densely developed there.

[0048] Through statistical analysis of logging data, it is found that the natural fractures in S13 are mainly oriented in the northeast and northwest directions, the bottom hole fractures are mainly of medium to high inclination, the spatial development is dense, and the proportion of open fractures is relatively large.

[0049] Based on 3D seismic data, S14 extracts the attribute information of ant bodies, coherence bodies, and curvature bodies to determine the spatial distribution pattern of deep fractures.

[0050] S15 comprehensively compares the statistical data of multiple groups of fractures from different sources. The distribution patterns of fracture development are generally consistent, and the characteristics of the fracture network in deep three-dimensional space can be preliminarily inferred.

[0051] S2. Conduct indoor tests to determine the fracture mechanism of reservoir granite: Obtain rock samples from the study area, simulate the crack morphology and formation mechanism of granite compression, shearing and splitting under reservoir temperature and pressure conditions, and test various mechanical parameters of granite. Conduct true triaxial hydraulic fracturing tests to simulate the artificial water injection fracturing process and determine the crack evolution process from the perspective of mechanical causes. Specific steps include:

[0052] S21 conducted a uniaxial compression test on granite at room temperature. The rock failure surface formed was wide and rough, showing obvious crushing damage, which is a typical tensile failure.

[0053] After increasing the temperature and pressure conditions in S22, the granite sample transitioned from crushing to relatively complete shear failure. The main reason may be the closure of microcracks inside the granite caused by the increase in confining pressure. Therefore, it is inferred that shear failure is the main type of rock failure in deep reservoirs. The Young's modulus of granite was measured to be between 16-45Gpa, and the Poisson's ratio was between 0.15-0.3.

[0054] S23 uses multiple groups of 400×400×400 granite specimens to carry out true triaxial hydraulic fracturing tests, and evaluates the expansion of hydraulic fractures by combining the acoustic emission characteristics of rock samples with injection pressure. The test shows that the fracture initiation pressure is between 25-35Mpa. As the confining pressure increases, the fracture initiation pressure is greater. After the fracture is formed, the pressure to maintain the fracture expansion under high confining pressure conditions is smaller.

[0055] S3. Acquisition of fracture parameters after engineering fracturing: The fracture parameters after fracturing are obtained by combining microseismic monitoring data with remote detection acoustic logging (near the wellbore) and conventional logging (surface) data. The main process is as follows: Figure 2 The specific steps include:

[0056] S31 collects microseismic monitoring related data, including seismic waveform data, source location and seismic wave propagation path information.

[0057] S32 randomly selects a certain number of sample data points from the microseismic data to initialize the fracture parameter estimation.

[0058] S33 uses the selected sample data points to perform parameter estimation according to the crack model to obtain a fitting model of the crack.

[0059] S34 statistically calculates the average distance μ and variance σ of all sample data points to the fitting model, and calculates the distance d of each sample data point to the fitting model. Sample data points that satisfy the following formula are set as internal points (fitting the model better), otherwise they are outliers (fitting the model poorly).

[0060] |d-μ|≤σ

[0061] S35 If the number of internal points of the current fitting model exceeds the set threshold, the current model is considered to be valid and the iteration is stopped; otherwise, steps S32-S34 are repeated.

[0062] S36 uses all the internal point data to refit the model and obtain the final fracture parameter estimate, generating the optimal solution for this fracture.

[0063] S37 removes the data points that successfully generate fractures in the microseismic data set, and continues with steps S32-S36 until the next fracture parameter is successfully obtained.

[0064] S4. Establish a multi-field coupled fracture propagation model to characterize the fracture network in deep reservoirs. The main process is as follows: Figure 3 The specific steps include:

[0065] S41 establishes a natural fracture network model through the Monte Carlo method based on the fracture characteristic parameters in steps S1 and S2.

[0066] S42 couples rock physical parameters to simulate reservoir in-situ stress distribution.

[0067] S43 performs numerical simulation of fracture extension according to actual engineering parameters to obtain the fracture reconstruction volume after engineering fracturing.

[0068] S44 utilizes the method described in step S3 to reconstruct the post-fracturing fracture network through microseismic monitoring data observed in actual engineering projects, and calculates the fracture volume parameters.

[0069] S45 performs error analysis by comparing step S43 with step S44, and adjusts model parameters to optimize the fracture network model.

[0070] S46 generates the final deep reservoir fracture network model, such as Figure 4 shown.

[0071] S5. Model verification: The generated fracture network model is used to perform seepage simulation. The calculated results are well compared with the actual single and paired well injection and production test data, verifying the accuracy of the deep fracture network model.

[0072] Of course, the above description is not a limitation of the present invention, and the present invention is not limited to the above examples. Changes, modifications, additions or substitutions made by ordinary technicians in this technical field within the essential scope of the present invention also fall within the protection scope of the present invention.

Claims

1. A method for deep crack network characterization and three-dimensional reconstruction based on multi-source information fusion, characterized in that: The following steps are involved: S1. Conduct fusion analysis of surface explicit fractures and deep implicit fractures: Collect surface outcrops, drilling data, and seismic data in the study area, integrate these data, compare the formation evolution and distribution law of explicit fractures with the development law of deep implicit fractures, and infer and characterize the deep fracture network in combination with the geological structure evolution process; S2. Determine the rock fracture mechanism using laboratory methods: simulate the crack morphology and formation mechanism formed by rock compression, shearing and splitting under reservoir temperature and pressure conditions, and test various mechanical parameters of rocks; conduct true triaxial hydraulic fracturing tests, simulate the artificial water injection fracturing process, and determine the crack evolution process from the perspective of mechanical causes; S3. Acquisition of fracture parameters after engineering fracturing: Acquisition of fracture parameters after fracturing through microseismic monitoring data combined with remote detection sonic logging and conventional logging data; S4. Establish a multi-field coupled fracture propagation model to characterize the fracture network in deep reservoirs; S5. Model verification: Use the generated fracture network model to perform seepage simulation, compare the calculated results with the actual single and paired well injection and production test data, and verify the accuracy of the deep fracture network model.

2. The method for deep fracture network characterization and three-dimensional reconstruction based on multi-source information fusion according to claim 1 is characterized by: The step S1 specifically includes: S11. Select the outcrops in the surrounding study area with the same lithology or similar structure as the development site, conduct natural fracture surveys, collect and statistically analyze the direction, dip angle and trace length of natural fractures, and determine the development and distribution patterns of surface explicit fractures; S12. Collect the natural fracture inclination and dip angle information of the cores drilled in the study area, and perform XRD analysis on the altered mineral composition of the cores to determine the well sections with dense natural fracture development in the vertical range; S13. Collect logging data, identify natural fractures, and analyze information such as the inclination, dip, density, aperture, porosity, etc. of natural fractures, and draw fracture dip and dip rosette diagrams; S14. Collect 3D seismic data, extract ant volume, coherence volume, and curvature volume data according to reservoir characteristics, and determine the spatial distribution law of deep hidden fractures; S15. Compare and analyze multiple groups of fracture statistics from different sources, and preliminarily infer the characteristics of the fracture network in deep three-dimensional space by combining the shallow obvious fractures with the hidden fractures in the vertical range around the well.

3. The method for deep fracture network characterization and three-dimensional reconstruction based on multi-source information fusion according to claim 1 is characterized by: The step S2 specifically includes S21. Carry out multiple sets of uniaxial compression tests at room temperature to test the Young's modulus, Poisson's ratio, and compressive strength of the rock, and record the development morphology of the cracks when the rock is damaged; S22. On the basis of uniaxial test, the influence of confining pressure, temperature and the combined effect of temperature and pressure are added respectively to test and analyze the changes of rock mechanical properties, record the rock failure morphology, and summarize the change law of rock mechanical properties; S23. Conduct true triaxial hydraulic fracturing tests to evaluate the expansion of hydraulic fractures by combining the acoustic emission characteristics of rock samples with the injection pressure.

4. The method for deep fracture network characterization and three-dimensional reconstruction based on multi-source information fusion according to claim 1, characterized in that: The step S3 specifically includes: S31. Collect microseismic monitoring related data, including seismic waveform data, earthquake source location and seismic wave propagation path information; S32, randomly selecting a certain number of sample data points from the microseismic data to initialize the fracture parameter estimation; S33, using the selected sample data points, performing parameter estimation according to the crack model to obtain a fitting model of the crack; S34. Statistically calculate the average distance μ and variance σ of all sample data points to the fitting model, and calculate the distance d of each sample data point to the fitting model. The sample data points that meet the following formula are set as internal points, which have a good fit with the model. Otherwise, they are outliers, which have a poor fit with the model. |d-μ|≤σ; S35. If the number of internal points of the current fitting model exceeds the set threshold, the current model is considered to be valid and the iteration is stopped; otherwise, steps S32-S33 are repeated; S36, refitting the model using all the internal point data to obtain the final fracture parameter estimation and generate the optimal solution for this fracture; S37, removing the data points where the cracks are successfully generated in the microseismic data set, and continuing steps S32-S36 until the next crack parameter is successfully obtained.

5. The method for deep fracture network characterization and three-dimensional reconstruction based on multi-source information fusion according to claim 1, characterized in that: The step S4 specifically includes: S41, establishing a natural fracture network model by Monte Carlo method based on the fracture characteristic parameters in steps S1 and S2; S42, coupling rock physical parameters, performing geomechanical coarsening, and simulating reservoir in-situ stress distribution; S43, performing numerical simulation of fracture extension according to actual engineering parameters to obtain fracture reconstruction volume after engineering fracturing; S44, using the method described in step S3, reconstructing the post-fracturing fracture network through microseismic monitoring data observed in actual engineering, and calculating the fracture volume parameters; S45, performing error analysis by comparing step S43 with step S44, and adjusting model parameters to optimize the fracture network model; S46. Generate a final deep reservoir fracture network model.

Images