Karst fracture-cave reservoir intelligent identification method and device based on seismic geological data

Abstract

The invention relates to the technical field of oil and gas geology exploration and development, and discloses a karst fracture-cavity reservoir intelligent identification method based on seismic geological data and a karst fracture-cavity reservoir intelligent identification device based on the seismic geological data. According to the method, logging data, logging data, seismic data, rock core data, test data, production data and other data of the karst fracture-cavity reservoir in the exploration stage are fully utilized, the geologic feature model and the seismic forward modeling model are constructed on the basis, the constructed models are more reasonable and reliable, and the construction efficiency is improved. And a machine learning means in artificial intelligence is fused into the method technology, and karst fracture-cavity reservoir identification of the whole work area is performed through a result obtained through data and model training, so that the interpretation efficiency and the prediction precision can be greatly improved, and the time cost is saved.

Description

Technical Field

[0001] The present invention relates to the technical field of oil and gas geological exploration and development, and particularly relates to an intelligent identification method and device for karst fracture-vug reservoirs based on seismic geological data. Background Art

[0002] Carbonate karst fracture-vug reservoirs are one of the main reservoir types in the three major basins in the western part of China. After multiple stages of tectonic fractures, paleoweathering, dissolution and other effects, the reservoirs show a carbonate fracture-vug system with strong heterogeneity in space, providing an important place for oil and gas migration and accumulation.

[0003] Currently, for the identification of reservoirs, conventional methods only consider using core data, logging data and seismic data to identify karst reservoirs. The types of data sets used in the identification process are single, resulting in a low accuracy of the identification results. For example, the invention patent with the publication number CN113156505A discloses a method for identifying reef bank reservoirs in fault-depressed lake basins with progressive constraints of "basement structure-paleogeomorphology-seismic facies", and the invention patent with the publication number CN116027419A discloses a method for identifying and determining gas-bearing sandstone reservoirs using seismic waveform characteristics. The above existing technologies have the problems described in the background art. Summary of the Invention

[0004] In order to solve the problems and deficiencies existing in the above-mentioned prior art, the present invention proposes an intelligent identification method and device for karst fracture-vug reservoirs based on seismic data. This method makes full use of various data such as logging data, logging data, seismic data, core data, test data, production data, etc. of karst fracture-vug reservoirs in the exploration stage, and uses this as basic data to construct a geological feature model and a seismic forward model. The constructed models are more reasonable and reliable and more in line with the actual geological situation.

[0005] In order to achieve the above invention purpose, the technical solution of the present invention is specifically as follows:

[0006] An intelligent identification method for karst fracture-vug reservoirs based on seismic geological data, the method mainly includes the following steps:

[0007] S1. Obtain the seismic geological data of karst fracture-vug reservoirs in the exploration stage and establish a database.

[0008] In the present invention, the data includes logging data, logging data, seismic data, core data, test data and production data, and the established database is as follows:

[0009] D = Sta(L, C, T, S, P);

[0010] Among them, D represents the established database, Sta(·) represents statistical data collection, L represents logging and well logging related data, C represents core data, T represents test data, S represents seismic data, and P represents production data.

[0011] In this step, a variety of data are fully utilized to establish a database, which can provide rich and reliable data support for subsequent model establishment.

[0012] S2. According to the seismic geological data of the karst fracture-cavity reservoir obtained in step S1, construct a geological feature model of the karst reservoir, establish a forward seismic model, and clarify the seismic response characteristics of the karst reservoir.

[0013] In the present invention, the establishment of the geological feature model mainly includes: reservoir microscopic feature analysis, reservoir type analysis, and stratigraphic division and correlation analysis; among them,

[0014] For the stratigraphic division and correlation sub-analysis, through the stratigraphic division and correlation results of the gas reservoir, obtain the connection with the characteristics in the data (logging and well logging data), and through marking and multi-data analysis, divide the characteristics with more correlation into weight "1", and divide the characteristics with less correlation or low occurrence frequency into weight "0";

[0015] For the reservoir microscopic feature analysis, mainly through the analysis of the sedimentary microfacies types and typical features of the gas reservoir, obtain the connection with the characteristics in the data (core data and test data), and through marking and multi-data analysis, divide the characteristics with more correlation into weight "1", and divide the characteristics with less correlation or low occurrence frequency into right "0";

[0016] For the reservoir type analysis, mainly through the connection between the reservoir characteristics and distribution characteristics of the gas reservoir in the existing gas reservoirs and the characteristics of the corresponding data (core data, test data, well logging, logging data), and through marking and multi-data analysis, divide the characteristics with more correlation into weight "1", and divide the characteristics with less correlation or low occurrence frequency into weight "0".

[0017] Among them, considering that the two are related, that is, the original beach facies is the basis for reservoir development, and tectonic fracture and supergene dissolution are the promoting factors for reservoir formation;

[0018] Through the above processing methods and means, finally establish a geological feature model including reservoir microscopic feature analysis, reservoir type analysis, and stratigraphic division and correlation analysis.

[0019] Furthermore, in the present invention, the establishment of the seismic forward model mainly includes: fine classification of structural fractures, fracture-vug analysis, reservoir characteristic analysis, and geological characteristic analysis; among which, for the above, similarly, through the analysis of the porosity, permeability, and saturation of the gas reservoir, and the connection with the corresponding logging, seismic, and reservoir characteristics; through marking and the analysis of logging, seismic, and geological data, the characteristics with high correlation are assigned a weight of "1", and the characteristics with low correlation or low occurrence frequency are assigned a weight of "0".

[0020] In this step, when establishing the geological model and the seismic forward model, the data analysis of logging, seismic, structure, sedimentary facies, hydrocarbon accumulation, etc. is fully utilized to establish the geological model and the seismic forward model during training, which can make the established model more accurate and thus obtain a more reliable training result.

[0021] S3. The characteristics related to the karst reservoir can be obtained through the above geological characteristic model and seismic forward model, and then combined with the constructed model for machine learning training, so as to establish an intelligent recognition model for the karst fracture-vug reservoir and standardize it.

[0022] M = AI m [GM, SM];

[0023] Among them, GM is the established geological model; SM is the established seismic forward model; AI[·] is for machine learning; the subscript m represents the method during machine learning, and a suitable method can be selected according to the actual data situation, such as model training methods like convolutional neural network and deep learning; M is the calculated training result related to the karst fracture-vug reservoir.

[0024] This step uses machine learning for the intelligent recognition of karst reservoirs, which can improve the recognition efficiency, save time costs, and avoid the cumbersome process of using manual interpretation for the recognition of karst fracture-vug reservoirs.

[0025] S4. Input the geological data of the target area to be recognized into the model for recognition, so as to obtain the fracture-vug distribution recognition results of each well and the overall block.

[0026] Based on the same inventive concept, the present invention also provides an intelligent recognition device for karst fracture-vug reservoirs based on seismic geological data. The device is used to implement the above intelligent recognition method. The device includes a data acquisition module, a model construction module, a model training module, and a reservoir intelligent recognition output module; among which,

[0027] The data acquisition module acquires the seismic geological data of the karst fracture-vug reservoir during the exploration stage.

[0028] A model construction module that constructs a geological feature model and a forward seismic model of a karst reservoir based on the acquired seismic geological data of a karst fractured-vuggy reservoir;

[0029] A model training module that conducts machine learning training by combining the constructed geological feature model and forward seismic model, thereby establishing a seismic geological identification model for a karst fractured-vuggy reservoir and standardizing it;

[0030] A reservoir intelligent identification output module that inputs the geological data of the target area to be identified into the identification model for identification, and obtains the identification results of the fracture-vug distribution of each well and the overall block.

[0031] A computer device includes a memory, a processor, and a computer program stored on the memory and executable in the processor. When the processor executes the computer program, the method steps of the intelligent identification method for a karst fractured-vuggy reservoir of the present invention are implemented.

[0032] A computer-readable storage medium stores a computer program. When the computer program is executed in a computer processor, the method steps of the intelligent identification method for a karst fractured-vuggy reservoir of the present invention are implemented.

[0033] Advantages of the present invention:

[0034] 1. The present invention makes full use of various data such as logging data, well logging data, seismic data, core data, test data, and production data of a karst fractured-vuggy reservoir in the exploration stage. Based on this basic data, the acquired data is used to construct a geological feature model and a forward seismic model, and the constructed models are more reasonable and reliable.

[0035] 2. When establishing the model, the present invention conducts various studies such as fine stratigraphic correlation, microscopic reservoir characteristics, reservoir types, analysis of main controlling factors of storage, fine interpretation of structural fractures, seismic response analysis, and seismic inversion, which comprehensively restrict the establishment of the model. The obtained model is more in line with the actual geological situation and more accurate.

[0036] 3. The present invention integrates the machine learning method in artificial intelligence into this method technology. When identifying the karst fractured-vuggy reservoir in the entire work area through the results obtained from data and model training, the interpretation efficiency can be greatly improved and the time cost can be saved.

[0037] 4. The present invention first integrates the machine learning method into the identification of karst fractured-vuggy reservoirs. By training a preset machine learning model with a data set, a seismic geological identification model for karst fractured-vuggy reservoirs is obtained and standardized; finally, the geological data of the target area to be identified is input into the model for identification, thereby obtaining geological identification results, which can greatly improve the interpretation efficiency and save time costs. Brief Description of the Drawings

[0038] The foregoing and following specific descriptions of the present invention become clearer when read in conjunction with the following drawings, in which:

[0039] Figure 1 is the flow chart of the method of the present invention;

[0040] Figure 2 is the structural composition diagram of the device of the present invention;

[0041] Figure 3 are the core data, field profiles, logging data, and seismic data obtained in the embodiments of the present invention.

[0042] Figure 4 are the geological model and seismic forward model established in the embodiments of the present invention. Detailed Description of the Invention

[0043] In order to enable those skilled in the art to better understand the technical solutions in the present invention, the following will further illustrate the technical solutions for achieving the objectives of the present invention through several specific embodiments. It should be noted that the technical solutions claimed by the present invention include but are not limited to the following embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts shall fall within the protection scope of the present invention.

[0044] Carbonate karst fracture-vuggy reservoirs are one of the main reservoir types in the western part of China. After multiple stages of tectonic fractures, paleo-weathering, dissolution and other effects, the reservoirs show a carbonate fracture-vuggy system with strong heterogeneity in space, providing an important place for oil and gas migration and accumulation.

[0045] Currently, for the identification of reservoirs, conventional methods only consider using core data, logging data and seismic data to identify karst reservoirs. The types of data sets used in the identification process are single, resulting in a low accuracy of the identification results.

[0046] Based on this, the present invention proposes an intelligent identification method and device for karst fracture-vuggy reservoirs based on seismic geological data. The method of the present invention makes full use of various data such as logging data, logging data, seismic data, core data, test data, production data, etc. of karst fracture-vuggy reservoirs in the exploration stage, and uses this as basic data to construct a geological feature model and a seismic forward model. The constructed models are more reasonable and reliable and more in line with the actual geological situation.

[0047] This embodiment discloses an intelligent identification method for karst fracture-vuggy reservoirs based on seismic geological data. The following further illustrates the implementation of the present invention with reference to the drawings, mainly including the following steps:

[0048] (1) Input data such as petrology, paleontology, well logging, etc. of field profiles, core observations, cast thin sections, and drilling cores, determine the development types, vertical distributions of regional reservoir spaces, and reservoir control factors, and establish a geological data database.

[0049] (2) Input relevant seismic data such as seismic horizon calibration, amplitude characteristics, forward and inverse modeling results of different horizons of a single well, and establish a seismic database and a comprehensive seismic model.

[0050] (3) Establish a geological feature model based on main geological parameters such as single-well stratification, lithological characteristics, reservoir characteristics, sedimentary microfacies, etc.

[0051] (4) Conduct numerical calculations by combining the geological feature model and the comprehensive seismic model to perform a standardized template for fracture-vug characteristics of each layer series.

[0052] (5) According to the standardized template, obtain the fracture-vug distribution areas of each single well in different layer series, and finally calculate the distribution law of karst fracture-vugs in the entire area.

[0053] The specification appendix Figure 1 is the core data, field profiles, well logging data, and seismic data obtained in the embodiments of the present invention. A model is constructed by importing the research layer series as historical training data.

[0054] The specification appendix Figure 2 is a certain geological model and seismic forward modeling model established in the embodiments of the present invention. From Figure 2 it can be clearly determined the geological distribution characteristics and seismic reflection characteristics of karst reservoirs. Irregular karsts form irregular seismic axis pull-downs on seismic data and produce phase mutations. Further, by constructing various types of models for training and learning, finally, the distribution of karst fracture-vug reservoirs in the entire research area is predicted using the learning results.

[0055] The present invention can directly input the results after learning and training into the data of the research area to obtain the predicted distribution results of karst reservoirs in the entire work area. Usually, it may take dozens to hundreds of hours of workload to complete the interpretation by manual interpretation method, while only dozens of seconds or even seconds are required by this method, which greatly improves the interpretation work efficiency and can save time costs.

[0056] Compared with the prior art, the present invention proposes a more accurate and rapid technical idea. Since a variety of data are fully utilized to establish a data set, and the data set is constructed by using a variety of data of seismic geology of karst fracture-vug reservoirs, that is to say, the data set has data diversity. Therefore, the data set is used as the data source for training a preset machine learning model into a karst fracture-vug reservoir identification model, ensuring the reliability and accuracy of the karst fracture-vug reservoir identification model and preparing for the data requiring karst fracture-vug reservoir identification. In addition, the technology of the present invention applies machine learning to the identification of karst fracture-vug reservoirs, which has distinct creativity. After model training, the identification results of karst fracture-vug reservoirs in the study area can be obtained quickly, with extremely high efficiency.

[0057] The above is only a preferred embodiment of the present invention, and does not constitute any form of obstruction to the present invention. Any simple modification or equivalent change made to the above embodiments based on the technical essence of the present invention shall fall within the protection scope of the present invention.

Claims

1. An intelligent identification method for karst fracture-cave reservoirs based on seismic geological data, characterized in that, The method includes the following steps: Step S1. Obtain the seismic geological data of karst fracture-vug reservoirs in the exploration stage; Step S2. According to the obtained seismic geological data of karst fracture-vug reservoirs, construct a geological feature model and a seismic forward modeling of karst reservoirs; Step S3. Combine the constructed models for machine learning training to establish a seismic geological identification model for karst fracture-vug reservoirs and standardize it; Step S4. Input the geological data of the target area to be identified into the model for identification, so as to obtain the identification results of the fracture-vug distribution of each well and the overall block.

2. The intelligent identification method for karst fracture-vug reservoir based on seismic geological data according to claim 1, wherein The construction of the geological feature model and the seismic forward modeling of karst reservoirs according to the obtained seismic geological data of karst fracture-vug reservoirs includes at least: Analyze the correlation of the elements mainly related to the characteristics of karst fracture-vug reservoirs in the seismic geological data, divide the characteristics with more correlation into weight "1", and divide the characteristics with less correlation or low occurrence frequency into weight "0", and then establish a geological feature model and a seismic forward modeling.

3. The intelligent identification method for karst fracture-cave reservoir based on seismic geological data according to claim 1, characterized in that The geological feature model includes stratigraphic division and correlation analysis, microscopic reservoir feature analysis and reservoir type analysis.

4. The intelligent identification method for karst fracture-vug reservoir based on seismic geological data according to claim 1, characterized in that, The seismic forward modeling includes fine classification of structural faults, fracture-vug analysis, reservoir feature analysis and geological feature analysis.

5. The intelligent identification method for karst fracture-vug reservoir based on seismic geological data according to claim 1, characterized in that The seismic geological data includes logging data, well logging data, core data, test data, production data and seismic data.

6. An intelligent identification device for karst fracture-vug reservoirs based on seismic and geological data, the device is used to implement the intelligent identification method for karst fracture-vug reservoirs based on seismic and geological data according to any one of the above claims 1-5, characterized in that, The device includes a data acquisition module, a model construction module, a model training module and a reservoir intelligent identification output module; among them, The data acquisition module obtains the seismic geological data of karst fracture-vug reservoirs in the exploration stage; The model construction module constructs a geological feature model and a seismic forward modeling of karst reservoirs according to the obtained seismic geological data of karst fracture-vug reservoirs; The model training module combines the constructed geological feature model and seismic forward modeling for machine learning training to establish a seismic geological identification model for karst fracture-vug reservoirs and standardize it; The reservoir intelligent identification output module inputs the geological data of the target area to be identified into the identification model for identification, and obtains the identification results of the fracture-vug distribution of each well and the overall block.

7. A computer device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method steps described in any one of claims 1-5 above.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, and when the computer program is executed in a computer processor, it implements the method described in any one of claims 1-5 above.

Images