Reservoir fracture prediction method and device

Abstract

A reservoir fracture prediction method and apparatus, the method comprising: preprocessing seismic data of a predicted area, and determining a plurality of seismic attributes based on the preprocessed seismic data; preprocessing well logging data of the predicted area to obtain well logging information; and obtaining reservoir fracture density data of the predicted area based on the seismic attributes, the well logging information, and a pre-established discrete fracture network model; wherein the discrete fracture network model is composed of a combination of a large fracture discrete random network model, a medium fracture discrete random network model, and a small fracture discrete random network model.

Description

Technical Field

[0001] This article relates to the field of geophysical exploration technology, and in particular to a reservoir fracture prediction method and device. Background Art

[0002] Fractures are important pathways for oil and gas migration in low-permeability reservoirs and have a significant impact on oil and gas field development. Low-permeability reservoirs have poor porosity and permeability, resulting in high displacement pressures, making it difficult for oil to undergo significant secondary migration in such reservoirs. The combined transport pattern of reservoirs and fractures allows for better connectivity between the reservoir and source rock, allowing oil and gas to migrate and adjust along fracture-developed zones. Furthermore, during waterflooding, fractures can easily lead to uneven water absorption in injection wells, water channeling along the fractures, and rapid increases in water content in oil wells in sandstone fracture-developed zones. This leads to lower water drive levels in lateral wells along fractures and significant declines in oil well productivity. Therefore, accurately identifying reservoir fractures and objectively understanding their distribution characteristics is key to efficient production of low-permeability reservoirs.

[0003] In the prior art, seismic methods are widely used to detect fractures, which generally include the following aspects: ① fracture prediction based on longitudinal wave anisotropy; ② special technologies for special cases, such as identifying fracture-cavity reservoirs through seismic response characteristics of fracture-cavity reservoirs; ③ numerical simulation technology of tectonic stress; ④ fracture prediction research based on logging-constrained wave impedance inversion; for example, an invention patent discloses a method for predicting underground fractures, which obtains the radial component R(t) and the lateral component T(t) of the converted shear wave; obtains the time difference function Δt(t) between the fast shear wave component S1(t) and the slow shear wave component S2(t) based on the radial component R(t) and the lateral component T(t); derives the time difference function Δt(t) between the fast shear wave component S1(t) and the slow shear wave component S2(t) to obtain the derivative function dt(t) of the time difference function; and uses the derivative function dt(t) to predict the intensity of fracture development.

[0004] However, the above-mentioned fracture identification and prediction are performed on seismic data volumes, and the resolution and accuracy are affected to a certain extent; therefore, there is an urgent need for a reservoir fracture prediction method that can improve the resolution and accuracy. Summary of the Invention

[0005] The present application provides a reservoir fracture prediction method and apparatus, which predicts reservoir fractures based on seismic data and well logging data. Furthermore, during the prediction, the fractures are divided into three levels of fractures, and different levels of fracture models are established based on the identification characteristics of the different levels of fractures. Finally, the models are integrated into a discrete fracture network model for predicting fracture density, thereby improving the prediction accuracy of fracture density.

[0006] In a first aspect, the present application provides a reservoir fracture prediction method, the method comprising:

[0007] Preprocessing the seismic data of the prediction area, and determining multiple seismic attributes based on the preprocessed seismic data;

[0008] Preprocess the logging data of the predicted area to obtain logging information;

[0009] Reservoir fracture density data of the area to be predicted is obtained based on the seismic attributes, the well logging information, and a pre-established discrete fracture network model; wherein the discrete fracture network model is composed of a combination of a large fracture discrete random network model, a medium fracture discrete random network model, and a small fracture discrete random network model.

[0010] In an exemplary embodiment, the large fracture discrete random network model is established based on artificial earthquake fault interpretation information;

[0011] The medium crack discrete random network model is a model established based on the properties of ant bodies;

[0012] The small fracture discrete random network model is a model established based on the inter-well fracture probability model.

[0013] In an exemplary embodiment, preprocessing of seismic data includes one or more of the following: resampling, strong axis removal, and median filtering.

[0014] In an exemplary embodiment, determining a plurality of seismic attributes based on the preprocessed seismic data includes:

[0015] Conducting stratigraphic structural interpretation based on pre-processed seismic data;

[0016] Extract seismic attributes based on the results of stratigraphic structure interpretation.

[0017] In an exemplary embodiment, the seismic attributes include one or more of the following: variance volume, ant volume, structural dip, structural azimuth, and curvature.

[0018] In an exemplary embodiment, preprocessing the well logging data of the area to be predicted to obtain the well logging information includes:

[0019] De-noising the dual laterolog data in the prediction area;

[0020] The deep laterolog data after denoising is subtracted from the shallow laterolog data to obtain the amplitude difference of the dual laterolog data.

[0021] Logging fracture density data is determined according to the amplitude difference of the dual laterolog data, and the logging fracture density data is used as logging information.

[0022] In an exemplary embodiment, the process of establishing the small crack discrete random network model is as follows:

[0023] Step 1. Determine the correlation coefficient between the probability of fracture development and the seismic attribute, and establish a seismic attribute fusion body based on the correlation coefficient;

[0024] Step 2. Establish a 3D formation brittleness model based on the well logging mineral composition;

[0025] Step 3. Generate a fault distance attribute volume based on the large fracture discrete random network model;

[0026] Step 4. Establishing an inter-well fracture probability model based on the seismic attribute fusion body, the formation brittleness three-dimensional model and the fault distance attribute body;

[0027] Step 5: Establish the small fracture discrete random network model based on the inter-well fracture probability model, small fracture geometric parameters, and logging fracture density in the logging information.

[0028] In an exemplary embodiment, establishing an inter-well fracture probability model based on the seismic attribute fusion volume, the formation brittleness three-dimensional model, and the fault distance attribute volume includes:

[0029] Based on the seismic attribute fusion body, the 3D formation brittleness model and the fault distance attribute body, a multivariate fusion method is used to establish the interwell fracture probability body model.

[0030] In an exemplary embodiment, establishing an inter-well fracture probability model based on the seismic attribute fusion volume, the formation brittleness three-dimensional model, and the fault distance attribute volume includes:

[0031] Setting the small fracture geometry parameters and the logging fracture density as parameters of the small fracture discrete random network model;

[0032] The inter-well fracture probabilistic model is used as an objective function to generate the small fracture discrete random network model.

[0033] In an exemplary embodiment, determining the correlation coefficient between the fracture development probability and the seismic attribute and establishing the seismic attribute fusion body based on the correlation coefficient includes:

[0034] Normalize the seismic attributes;

[0035] Based on the correlation analysis between the probability of fracture development and each normalized seismic attribute, the correlation coefficient between the probability of fracture development and each seismic attribute is determined;

[0036] The correlation coefficient is used as a weight coefficient to perform weighted averaging on the normalized seismic attribute volume to obtain a seismic attribute fusion volume.

[0037] In an exemplary embodiment, establishing a three-dimensional formation brittleness model based on well logging mineral composition includes:

[0038] Calculate the logging mineral composition using a multi-mineral model based on the logging data;

[0039] The brittleness index value of a single well is calculated using the mineral composition ratio method;

[0040] Under the control of lithofacies, a three-dimensional formation brittleness model is established based on the brittleness index value of a single well.

[0041] In a second aspect, an embodiment of the present invention provides a device for predicting reservoir fractures, the device comprising: a memory and a processor; the memory is used to store a program for predicting reservoir fractures, and the processor is used to read and execute the program for predicting reservoir fractures, and perform any one of the methods described in the above embodiments.

[0042] In a third aspect, an embodiment of the present invention provides a computer-readable storage medium having a data processing program stored thereon, and the data processing program is executed by a processor to perform a reservoir fracture prediction method according to any one of the above embodiments.

[0043] Compared to related technologies, the present application provides a reservoir fracture prediction method and apparatus. The method comprises: preprocessing seismic data in the prediction area and determining multiple seismic attributes based on the preprocessed seismic data; preprocessing well logging data in the prediction area to obtain well logging information; and obtaining reservoir fracture density data for the prediction area based on the seismic attributes, the well logging information, and a pre-established discrete fracture network model. The discrete fracture network model is composed of a large fracture discrete random network model, a medium fracture discrete random network model, and a small fracture discrete random network model. The present application predicts reservoir fractures based on seismic data and well logging data. Furthermore, during the prediction, fractures are divided into three grades, and different grade fracture models are established based on their identification characteristics. Finally, these models are merged into a discrete fracture network model for predicting fracture density, thereby improving the accuracy of fracture density prediction.

[0044] Other features and advantages of the present application will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the present application. Other advantages of the present application can be realized and obtained by the solutions described in the description and the drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The accompanying drawings are used to provide an understanding of the technical solution of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the technical solution of the present application and do not constitute a limitation on the technical solution of the present application.

[0046] Figure 1 This is a flow chart of a reservoir fracture prediction method according to an embodiment of the present application;

[0047] Figure 2 This is a schematic diagram of a reservoir fracture prediction device according to an embodiment of the present application;

[0048] Figure 3 A flow chart is established for a small crack discrete random network model in some exemplary embodiments. DETAILED DESCRIPTION

[0049] This application describes multiple embodiments, but this description is exemplary rather than restrictive, and it will be apparent to those skilled in the art that there may be more embodiments and implementations within the scope of the embodiments described herein. Although many possible feature combinations are shown in the drawings and discussed in the detailed description, many other combinations of the disclosed features are also possible. Unless specifically limited, any feature or element of any embodiment may be used in combination with any other feature or element in any other embodiment, or may replace any other feature or element in any other embodiment.

[0050] This application includes and contemplates combinations of features and elements known to those of ordinary skill in the art. The embodiments, features, and elements disclosed in this application may also be combined with any conventional features or elements to form a unique inventive solution defined by the claims. Any features or elements of any embodiment may also be combined with features or elements from other inventive solutions to form another unique inventive solution defined by the claims. Therefore, it should be understood that any feature shown and / or discussed in this application may be implemented individually or in any appropriate combination. Therefore, except for the limitations made according to the appended claims and their equivalents, the embodiments are not subject to other limitations. In addition, various modifications and changes may be made within the scope of protection of the appended claims.

[0051] In addition, when describing representative embodiments, the specification may have presented the method and / or process as a specific sequence of steps. However, to the extent that the method or process does not rely on the specific order of the steps described herein, the method or process should not be limited to the steps in the specific order described. As will be understood by those skilled in the art, other orders of steps are also possible. Therefore, the specific order of the steps set forth in the specification should not be interpreted as a limitation to the claims. In addition, the claims for the method and / or process should not be limited to performing their steps in the order written, and those skilled in the art can readily understand that these orders can be changed and still remain within the spirit and scope of the embodiments of the present application.

[0052] The embodiment of the present invention provides a reservoir fracture prediction method, such as Figure 1 As shown, the method includes steps S100-S120, which are specifically as follows:

[0053] Step S100: pre-processing the seismic data of the prediction area, and determining a plurality of seismic attributes based on the pre-processed seismic data;

[0054] Step S110: pre-processing the well logging data of the predicted area to obtain well logging information;

[0055] Step S120: Obtaining reservoir fracture density data of the area to be predicted based on the seismic attributes, the well logging information, and a pre-established discrete fracture network model.

[0056] In this embodiment, the discrete fracture network model is composed of a large fracture discrete random network model, a medium fracture discrete random network model, and a small fracture discrete random network model. The discrete fracture random network model defines parameters such as fracture location, occurrence, and size as random variables and assigns them a probability distribution function. This represents the fracture system using discrete fracture data.

[0057] In an exemplary embodiment, preprocessing of the seismic data of the prediction area includes one or more of the following: resampling, strong axis removal, median filtering, and edge detection. In this embodiment, the specific implementation of each preprocessing step is not specifically limited, and those skilled in the art may select a corresponding processing method.

[0058] For example, the resampling process can resample the original seismic data sampling time according to the reservoir thickness in the target area.

[0059] Strong axis removal can be performed on seismic reflection cycles using wavelet demarcation. This is achieved by decomposing strong reflection wavelets into wavelets of varying amplitudes and frequencies. Wavelets with large amplitudes are removed, and reconstructed with wavelets with smaller amplitudes, thereby suppressing the strong axis.

[0060] The median filter process is used to enhance the continuity of seismic reflection and reduce the impact of noise.

[0061] The detection of the edge of the seismic data volume is mainly to find the discontinuity point in the data volume, and to enhance this discontinuity with the help of attribute extraction techniques such as dip and azimuth attributes, chaos attributes and variance attributes; among them, the variance volume technology uses the similarity between adjacent seismic signals to describe the lateral heterogeneity of strata, lithology, etc., and by calculating the variance value of the sample points, it reveals the discontinuity information in the data volume, thereby identifying faults and lithology. On regular layer planes, the amplitude along the layer does not change much and the variance is small; while near the fault, the amplitude changes greatly and the variance is also large.

[0062] In an exemplary embodiment, a plurality of seismic attributes are determined based on the preprocessed seismic data; a stratigraphic structure interpretation is performed based on the preprocessed seismic data; and seismic attributes are extracted based on the stratigraphic structure interpretation results.

[0063] The implementation process of extracting earthquake attributes in this embodiment includes:

[0064] Step 1: Seismic data horizon interpretation;

[0065] Before analyzing seismic attributes, seismic data must first be interpreted at the horizon level, including synthetic seismograms, horizon calibration, and horizon tracking. Synthetic seismograms and horizon calibration are fundamental to seismic data interpretation and serve as the link between well logging and seismic data.

[0066] Specifically, step 1 includes:

[0067] Step 1.1: Synthetic seismic record production;

[0068] The reflection coefficient sequence is calculated using acoustic time difference logging data and density logging data. Then, the seismic wavelet extracted from the sub-side channel is convolved with the reflection coefficient sequence to obtain the synthetic seismic record of a single well.

[0069] Step 1.2: Seismic layer calibration;

[0070] The synthetic seismic record of the single well obtained in step 1.1 is compared with the seismic trace near the well. First, a rough calibration is performed using the marker formation, and then a fine calibration is performed using the wave group and wave system comparison until the correlation between the synthetic seismic record and the seismic trace near the well reaches a predetermined threshold.

[0071] Step 1.3: Horizon tracing interpretation;

[0072] According to the seismic waveform characteristics and reflection characteristics of the seismic traces near the well, gradually refined interpretation grids are used to track and compare on the seismic profile, establish an isochronous stratigraphic framework, and complete the horizon interpretation of the target area.

[0073] Step 2: Extract seismic attributes based on the horizon structure interpretation results;

[0074] Seismic attributes include one or more of the following: variance volume, ant volume, structural dip, structural azimuth, and curvature.

[0075] In one exemplary embodiment, based on well logging and seismic data, reservoir fractures are classified into three levels: large, medium, and small fractures, according to fracture size, the criticality of fracture identification methods, and the main controlling factors of fracture development. Corresponding discrete random network models for large, medium, and small fractures are then established. The large fracture discrete random network model is based on artificial seismic fault interpretation information; the medium fracture discrete random network model is based on ant body properties; and the small fracture discrete random network model is based on the interwell fracture probability body model.

[0076] In an exemplary embodiment, a large fracture discrete random network model is established based on artificial earthquake fault interpretation information:

[0077] Based on the structural interpretation results of the layers and faults of the three-dimensional original seismic data, the plane distribution law of the fault interpretation corresponding to each layer can be obtained, based on which a discrete random network model of large fractures is established.

[0078] In an exemplary embodiment, a discrete random network model for medium fractures is established based on ant body attributes. Ant body attributes can be used to establish the discrete random network model for medium fractures, or multiple three-dimensional seismic attribute volumes can be extracted for fracture detection, with seismic attributes sensitive to fractures being prioritized. Seismic attributes sensitive to fractures can be prioritized by creating along-layer attribute slices and interlayer attribute slices of the target layer, conducting structural and fault element analysis, and selecting seismic attributes sensitive to fractures to reveal the planar distribution pattern of fractures. For example, a discrete random network model for medium fractures can be established by prioritizing structural dip or curvature attributes.

[0079] Taking the selection of ant volume attributes to establish a discrete random network model of fractures as an example, the process of generating ant volumes from 3D raw seismic data consists of three main steps: first, structural smoothing is performed on the raw seismic data volume to reduce noise interference and enhance the continuity of seismic reflection wave events; second, variance analysis techniques are used on the structurally smoothed seismic data volume to obtain a variance volume, highlighting the discontinuities in the seismic data; third, multiple experiments are conducted on ant tracking parameters to determine the optimal parameter values ​​suitable for the study area. Ant tracking is then used to track discontinuities in the variance volume that meet the preset fracture conditions to generate ant volumes. The ant volume parameters include: defining the seed point, offset, and ant search step size.

[0080] In one exemplary embodiment, well logging data primarily includes acoustic waves, density, natural gamma, spontaneous potential, resistivity, and wellbore diameter. Conventional well logging curves are relatively comprehensive. However, due to factors such as the logging instrument, logging environment, and measurement time during field logging, the amplitudes of each logging curve are subject to interference from non-formation measurement factors, necessitating preprocessing of the logging data. Specifically, this preprocessing includes standardization of the logging data, reconstruction of acoustic wave curves, denoising, and calculation of logging fracture density data.

[0081] In an exemplary embodiment, before establishing the small fracture discrete random network model, it is necessary to pre-process the well logging data of the prediction area to obtain well logging information, including:

[0082] Step 1. De-noising the dual laterolog data in the prediction area;

[0083] Step 2: Subtract the denoised deep laterolog data from the shallow laterolog data to obtain the amplitude difference of the dual laterolog data;

[0084] Step 3: Determine logging fracture density data based on the amplitude difference of the dual laterolog data, and use the logging fracture density data as logging information.

[0085] In an exemplary embodiment, the process of establishing a small crack discrete random network model is as follows: Figure 3 As shown below:

[0086] Step S31. Determine the correlation coefficient between the probability of fracture development and the seismic attribute, and establish a seismic attribute fusion body based on the correlation coefficient;

[0087] In this step, the fracture development probability is obtained by statistically calculating the occurrence data of the target study area. For example, the fracture development probability is obtained by statistically calculating the ratio of the thickness of the fracture development section of a certain formation to the formation thickness of multiple wells in the target study area.

[0088] The implementation process of establishing a seismic attribute fusion body according to the correlation coefficient includes:

[0089] Step 311: normalize the earthquake attributes;

[0090] Step 312: Determine the correlation coefficient between the fracture development probability and each seismic attribute based on the correlation analysis between the fracture development probability and each normalized seismic attribute;

[0091] Step 313: Use the correlation coefficient as a weight coefficient to perform weighted averaging on the normalized seismic attribute volume to obtain a seismic attribute fusion volume.

[0092] The processing of this step can reduce the problem of multiple solutions for a single seismic attribute.

[0093] Step S32: Establishing a three-dimensional formation brittleness model based on the well logging mineral composition;

[0094] In this step, a 3D formation brittleness model is established based on the well logging mineral composition, including:

[0095] Step 321. Calculate the well logging mineral composition using a multi-mineral model based on the well logging data;

[0096] Step 322. Calculate the single well brittleness index value using the mineral composition ratio method;

[0097] Step 323. Under the control of lithofacies, a three-dimensional formation brittleness model is established according to the brittleness index value of the single well.

[0098] Step S33. Generate a fault distance attribute body based on a large fracture discrete random network model;

[0099] Faults are one of the important factors controlling fracture development. Generally speaking, the degree of fracture development is correlated with the distance to the fault; that is, the closer to the fault, the more developed the fracture. Therefore, based on the large fracture discrete random network model, a distance-to-fault attribute volume is generated and used as the constraint data for establishing the fracture probability volume model.

[0100] Step S34. Establishing an inter-well fracture probability model based on the seismic attribute fusion body, the formation brittleness three-dimensional model and the fault distance attribute body;

[0101] In this step, a multivariate fusion method may be used to establish an inter-well fracture probability body model based on the seismic attribute fusion body, the formation brittleness three-dimensional model, and the fault distance attribute body.

[0102] Step S35: establishing the small fracture discrete random network model according to the inter-well fracture probability model, small fracture geometric parameters, and the logging fracture density in the logging information.

[0103] In this step, a probability model of inter-well fractures is established based on the seismic attribute fusion body, the formation brittleness three-dimensional model and the fault distance attribute body, including:

[0104] Step 351: Setting the small fracture geometry parameters and the well logging fracture density as the small fracture discrete random network model parameters;

[0105] In this step, the geometric parameters of the small fractures include: fracture occurrence, fracture length, fracture aperture, and fracture permeability; the logging fracture density is obtained based on the well logging preprocessing.

[0106] Step 352: Using the inter-well fracture probabilistic model as the objective function, a small fracture discrete random network model is established using the Gaussian simulation method.

[0107] Based on the above technical solution, the reservoir fracture prediction method provided by this application has the following technical effects:

[0108] 1. The present invention uses a fusion of multiple data such as seismic data and well logging data to predict reservoir fractures. In addition, during the prediction, the fractures are divided into three levels of fractures, namely large fractures, medium fractures and small fractures, and discrete random network models of large fractures, medium fractures and small fractures are established respectively, which are fused into a discrete fracture network model. According to the identification characteristics of fractures of different levels, a deterministic method of artificial seismic fault interpretation is used to establish a discrete random network model of large fractures; a deterministic method of ant body tracking is used to establish a discrete random network model of medium fractures; a probability body model of inter-well fractures is established by integrating multiple information such as seismic information, geological information, and well logging information, and a fracture density model is established based on this as a constraint. Then, a discrete random network model of small fractures is established using the geometric morphology of small fractures and the fracture density model as constraint data, thereby greatly improving the prediction accuracy of fractures.

[0109] 2. When establishing a small fracture discrete random network model, the fracture density obtained from well logging data and the fracture density model obtained from seismic data are used to perform dual constraints on the small fracture discrete random network model, which greatly improves the accuracy of small fracture prediction.

[0110] The present disclosure also provides a device for predicting reservoir fractures. Figure 2 As shown, it includes: a memory S200 and a processor S210; the memory S200 is used to store a program for reservoir fracture prediction, and the processor S210 is used to read and execute the program for reservoir fracture prediction, and execute any one of the methods in the above embodiments.

[0111] An embodiment of the present disclosure further provides a computer-readable storage medium, on which a data processing program is stored. The data processing program is used by a processor to execute the method according to any one of the above embodiments.

[0112] Example 1

[0113] This example demonstrates a reservoir fracture prediction method. The specific process is as follows:

[0114] Step 1: Preprocess the seismic data of the prediction area; wherein the preprocessing includes one or more of the following: resampling, strong axis removal, median filtering, and edge detection.

[0115] Step 2: pre-process the well logging data of the prediction area;

[0116] This step preprocesses logging data from 195 wells in the study area. This logging data includes various types of data: acoustic wave, density, natural gamma, spontaneous potential, resistivity, and wellbore diameter. This preprocessing includes standardization and acoustic curve reconstruction. To predict fractures, preprocessing also includes calculating logging fracture density data. This calculation involves denoising the dual laterolog data in the prediction area, subtracting the denoised deep laterolog data from the shallow laterolog data to determine the amplitude difference between the dual laterolog data, and determining the logging fracture density data based on this amplitude difference.

[0117] Step 3: determining multiple seismic attributes based on the preprocessed seismic data;

[0118] Perform stratigraphic structural interpretation based on pre-processed seismic data; extract seismic attributes based on stratigraphic structural interpretation results.

[0119] Step 4: Establish a discrete crack network model based on the large crack discrete random network model, the medium crack discrete random network model and the small crack discrete random network model.

[0120] Based on logging, seismic and other data, the natural fractures in the reservoir are mainly divided into three levels of fractures, namely large fractures, medium fractures and small fractures according to the fracture scale, the limit of the fracture identification method and the main controlling factors of fracture development; a discrete fracture network model is established for each level of fractures.

[0121] 41. A large fracture discrete random network model is established based on artificial earthquake fault interpretation information.

[0122] 42. A medium crack discrete random network model is established based on the properties of ant bodies.

[0123] 43. A probability model of inter-well fractures is established by integrating seismic information, geological information, logging information, etc., and a fracture density model is established based on this as a constraint. Then, a discrete random network model of small fractures is established using the target-based characteristic point process simulation method, with the small fracture geometry and fracture density model as constraint data.

[0124] 44. The discrete random network models of large cracks, medium cracks and small cracks are integrated into a discrete crack network model, and the equivalent physical property parameter model of the cracks is obtained through coarsening.

[0125] Step 5: Use the discrete fracture network model to predict reservoir fracture density.

[0126] In this example, dual laterolog data is used to constrain the small fracture discrete random network model. In addition, when generating the small fracture discrete random network model, the logging fracture density obtained from the logging data and the fracture density model obtained from the seismic data are used to double constrain the small fracture discrete random network model, thereby improving the accuracy of small fracture prediction.

[0127] It will be appreciated by those skilled in the art that all or some of the steps, systems, and functional modules / units in the methods disclosed above may be implemented as software, firmware, hardware, and appropriate combinations thereof. In hardware implementations, the division between the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed by several physical components in cooperation. Some or all components may be implemented as software executed by a processor, such as a digital signal processor or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or temporary medium). As is well known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable, and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, it is well known to those skilled in the art that communication media generally embodies computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

Claims

1. A reservoir fracture prediction method, characterized in that: The method comprises: Preprocessing the seismic data of the prediction area, and determining multiple seismic attributes based on the preprocessed seismic data; Preprocess the logging data of the predicted area to obtain logging information; Obtaining reservoir fracture density data for the area to be predicted based on the seismic attributes, the well logging information, and a pre-established discrete fracture network model; wherein the discrete fracture network model is composed of a combination of a large fracture discrete random network model, a medium fracture discrete random network model, and a small fracture discrete random network model; The large fracture discrete random network model is established based on artificial earthquake fault interpretation information; The medium crack discrete random network model is a model established based on the properties of ant bodies; The small fracture discrete random network model is a model established based on the inter-well fracture probability model; The establishment process of the small crack discrete random network model is as follows: Step 1. Determine the correlation coefficient between the probability of fracture development and the seismic attribute, and establish a seismic attribute fusion body based on the correlation coefficient; Step 2. Establish a 3D formation brittleness model based on the well logging mineral composition; Step 3. Generate a fault distance attribute volume based on the large fracture discrete random network model; Step 4. Establishing an inter-well fracture probability model based on the seismic attribute fusion body, the formation brittleness three-dimensional model and the fault distance attribute body; Step 5: Establish the small fracture discrete random network model based on the inter-well fracture probability model, small fracture geometric parameters, and logging fracture density in the logging information.

2. The reservoir fracture prediction method according to claim 1, characterized in that: The preprocessing of seismic data includes one or more of the following: resampling, strong axis removal and median filtering.

3. The reservoir fracture prediction method according to claim 1, characterized in that: The determining of multiple seismic attributes based on the pre-processed seismic data includes: Conducting stratigraphic structural interpretation based on pre-processed seismic data; Extract seismic attributes based on the results of stratigraphic structure interpretation.

4. The reservoir fracture prediction method according to claim 3, characterized in that: The seismic attributes include one or more of the following: variance volume, ant volume, structural dip, structural azimuth, and curvature.

5. The reservoir fracture prediction method according to claim 2, characterized in that: The pre-processing of the well logging data of the prediction area to obtain the well logging information includes: De-noising the dual laterolog data in the prediction area; The deep laterolog data after denoising is subtracted from the shallow laterolog data to obtain the amplitude difference of the dual laterolog data. Logging fracture density data is determined according to the amplitude difference of the dual laterolog data, and the logging fracture density data is used as logging information.

6. The reservoir fracture prediction method according to claim 1, characterized in that: The establishing of an inter-well fracture probability model based on the seismic attribute fusion body, the formation brittleness three-dimensional model and the fault distance attribute body includes: A multivariate fusion method is used to establish an interwell fracture probability model based on the seismic attribute fusion body, the formation brittle three-dimensional model and the fault distance attribute body.

7. The reservoir fracture prediction method according to claim 1, characterized in that: The establishing of an inter-well fracture probability model based on the seismic attribute fusion body, the formation brittleness three-dimensional model and the fault distance attribute body includes: Setting the small fracture geometry parameters and the logging fracture density as parameters of the small fracture discrete random network model; The inter-well fracture probabilistic model is used as an objective function to generate the small fracture discrete random network model.

8. The reservoir fracture prediction method according to claim 1, characterized in that: Determining the correlation coefficient between the probability of fracture development and the seismic attribute, and establishing a seismic attribute fusion body based on the correlation coefficient, includes: Normalize the seismic attributes; Based on the correlation analysis between the probability of fracture development and each normalized seismic attribute, the correlation coefficient between the probability of fracture development and each seismic attribute is determined; The correlation coefficient is used as a weight coefficient to perform weighted averaging on the normalized seismic attribute volume to obtain a seismic attribute fusion volume.

9. The reservoir fracture prediction method according to claim 1, characterized in that: The method of establishing a three-dimensional formation brittleness model based on well logging mineral composition includes: Calculate the logging mineral composition using a multi-mineral model based on the logging data; The brittleness index value of a single well is calculated using the mineral composition ratio method; Under the control of lithofacies, a three-dimensional formation brittleness model is established based on the brittleness index value of a single well.

10. A device for predicting reservoir fractures, characterized in that: The device includes: a memory and a processor; the memory is used to store a program for reservoir fracture prediction, and the processor is used to read and execute the program for reservoir fracture prediction, and perform the method according to any one of claims 1-9.

11. A computer-readable storage medium having a data processing program stored thereon, wherein the data processing program is executed by a processor to perform the reservoir fracture prediction method according to any one of claims 1 to 9.

Images