A machine learning-based quantitative characterization method for lamellae development in shale reservoirs
Through CT scanning and machine learning technology, combining micron and micron nano CT data, a digital model of shale theory is constructed, which solves the problem of three-dimensional characterization of shale theory, improves the accuracy and efficiency of shale oil and gas exploration and development, and reduces the cost.
Patent Information
- Application Number
- CN202410931151.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-12
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2044-07-12
AI Technical Summary
The prior art cannot accurately characterize the three-dimensional spatial conditions of shale physics, resulting in inaccuracy and efficiency of shale oil and gas exploration and development.
Using CT scanning and reconstruction technology to obtain grayscale images of the shale core oil and gas reservoir slab structure, build a shale slab digital model through machine learning, combine micron and micron CT scanning data for multi-scale fusion to establish a shale slab structure machine learning standard model, and then obtain the digital result model of page pedicure development.
It improves the quantitative prediction accuracy of shale pedicure development, enhances the scientificity and economicality of shale oil and gas exploration and development, reduces costs, and improves the development efficiency and success rate of shale oil and gas resources.
Smart Images

Figure CN118470022B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of shale oil and gas exploration and development, and specifically relates to a machine learning-based quantitative characterization method for shale reservoir lamellae development. Background Art
[0002] The quantitative characterization of shale lamellae development has always been a world-class technical problem that has long been needed to be solved in shale oil and gas exploration and development projects but has not been fundamentally solved.
[0003] At present, the main technologies for characterizing shale foliation are manual qualitative measurement and scanning electron microscope qualitative detection. Manual measurement can only identify foliation above the millimeter level, and can only measure foliation displayed on a two-dimensional section. Scanning electron microscopes can only detect foliation below micrometers, and due to the limited field of view of scanning electron microscopes, they can only characterize the foliation development characteristics within a very small range of two-dimensional slices, and cannot better characterize the lateral changes in shale foliation. The current method can only obtain random probability values and cannot effectively characterize the true three-dimensional spatial condition of shale foliation, resulting in low accuracy of the obtained shale foliation development status, which in turn leads to technical problems of low accuracy in shale oil and gas exploration and development.
[0004] Therefore, there is an urgent need to provide a three-dimensional quantitative characterization method for shale lamellae development to improve the accuracy of shale oil and gas exploration and development. Summary of the invention
[0005] The purpose of the present invention is to overcome the shortcomings of the above-mentioned background technology and provide a machine learning-based quantitative characterization method for shale reservoir foliation development, so as to solve the technical problem that the measurement inaccuracy in the prior art leads to low accuracy of the obtained shale foliation development status, and then leads to low accuracy of shale oil and gas exploration and development prediction. The present invention improves the accuracy of quantitative prediction of shale oil and gas foliation development, improves the efficiency of shale oil and gas exploration and development, and reduces the cost of shale oil and gas exploration and development.
[0006] In order to solve the above technical problems, the present invention is implemented by the following technical solutions:
[0007] A machine learning-based quantitative characterization method for shale reservoir lamellae development, comprising:
[0008] Obtain grayscale images of the lamellar structure of shale core oil and gas reservoirs based on CT scanning and reconstruction, and construct a digital model of shale core lamellar structure based on the grayscale images of the lamellar structure of shale core oil and gas reservoirs;
[0009] Based on the digital model of shale core foliation, a standard machine learning model of shale core oil and gas reservoir foliation structure was constructed through machine learning of CT scan images.
[0010] Based on the machine learning standard model of shale core oil and gas reservoir lamellae structure, a digital result model of lamellae development in oil and gas layer core reservoirs is obtained through machine learning;
[0011] Based on the digital result model of lamellae development in core reservoirs of oil and gas layers, the shale lamellae evaluation standard and lamellae evaluation optimization model are established through the random judgment method, and then the lamellae-dominant target intervals of oil and gas layers are constructed.
[0012] Preferably, the step of obtaining a grayscale image of the lamella structure of a shale core oil and gas reservoir comprises:
[0013] Obtaining the geological background of the target shale layer, and determining the shale core CT scanning parameters based on the geological background;
[0014] The core is scanned based on the shale core CT scanning parameters to obtain a grayscale image of the lamella structure of the shale core oil and gas reservoir.
[0015] Further preferably, scanning the core based on the shale core CT scanning parameters to obtain the grayscale image of the lamella structure of the shale core oil and gas reservoir includes:
[0016] Perform micron CT scanning on the full-diameter core based on micron CT scanning parameters to obtain a micron grayscale image;
[0017] Perform micro-nano CT scanning on the plunger core based on micro-nano CT scanning parameters to obtain a micro-nano grayscale image;
[0018] The micron grayscale image and the micro-nano grayscale image are multi-scale fused to obtain the lamella structure grayscale image of the shale core oil and gas reservoir.
[0019] Preferably, the method of constructing a shale core lamellae digital model based on the grayscale image of the lamellae structure of the shale core oil and gas reservoir comprises:
[0020] Identify the grayscale image of the lamella structure of the shale core oil and gas reservoir to obtain lamella and structure characterization parameters;
[0021] Acquire screening criteria for the foliation characterization parameters, and determine a plurality of target foliation and structure characterization parameters from the foliation and structure characterization parameters based on the screening criteria;
[0022] The shale core foliation digital model is constructed based on the multiple target foliation and structural characterization parameters.
[0023] Preferably, the method of constructing a machine learning standard model of shale core oil and gas reservoir foliation structure based on the shale core foliation digital model through machine learning of CT scan images includes:
[0024] Based on machine learning of micron CT scan images, a standard machine learning model for micron grayscale images of foliation structures is obtained;
[0025] Based on the machine learning of micro-nano CT scanning images, a standard model for machine learning of micro-nano grayscale images of foliation structure is obtained.
[0026] Preferably, the method of obtaining a digitalized result model of lamellar development in oil and gas layer core reservoirs by machine learning based on a machine learning standard model of lamellar structure in shale core oil and gas reservoirs includes:
[0027] Identify the grayscale image of the lamella structure of shale core oil and gas reservoirs to obtain the lamella and structure characterization parameters;
[0028] Acquire screening criteria for the foliation characterization parameters, and determine a plurality of target foliation and structure characterization parameters from the foliation and structure characterization parameters based on the screening criteria;
[0029] A digital model of shale core foliation is constructed based on multiple target foliation and structural characterization parameters.
[0030] Preferably, the method of constructing a target interval with dominant lamellae in the oil and gas layer based on a digital model of reservoir lamellae development includes:
[0031] Based on the digital result model of lamellae development in core reservoirs of oil and gas layers, the shale lamellae evaluation standard and lamellae evaluation optimization model are established through the random judgment method, and then the lamellae-dominant target intervals of oil and gas layers are constructed.
[0032] Compared with the prior art, the method proposed in the present invention has the following significant advantages and beneficial effects in the quantitative characterization of shale reservoir lamellae development:
[0033] (1) The present invention uses CT scanning and reconstruction technology to capture shale foliation structure in three-dimensional space for the first time, overcoming the limitation of traditional methods that are limited to two-dimensional analysis. It can more accurately reflect the true development state of shale foliation and improve the depth of geological understanding of shale oil and gas exploration and development.
[0034] (2) By using machine learning technology to construct a standard model of shale foliation structure, it is not only able to identify subtle foliation structures at the micrometer or even nanometer level, but also able to capture the lateral variation characteristics of foliation through big data analysis and processing, greatly improving the accuracy and comprehensiveness of foliation characterization.
[0035] (3) The use of machine learning methods to automatically analyze CT scan images reduces the errors caused by human subjective judgment, making the shale characterization process more objective and efficient, and helps to standardize the evaluation process of shale oil and gas resources.
[0036] (4) The present invention provides a quantifiable and comparable data basis for shale oil and gas exploration through digital model construction, which helps to achieve refined management of shale reservoirs and optimize development strategies.
[0037] (5) Through the digital result model of comprehensive foliation development, it is possible to more accurately identify and locate the target strata with dominant foliation, provide scientific guidance for subsequent drilling and mining activities, and improve the success rate and economy of oil and gas resource development.
[0038] (6) Combining micron- and micro-nano-scale CT scanning data, a fusion analysis of multi-scale foliation structure is achieved. This method can more comprehensively cover the foliation characteristics at different scales, thus enhancing the comprehensiveness and accuracy of shale property analysis.
[0039] In summary, the machine learning-based quantitative characterization method proposed in the present invention effectively solves the problem of three-dimensional spatial characterization of shale strata by integrating high-precision CT imaging technology and advanced data analysis methods, significantly improves the scientific nature and economic benefits of shale oil and gas exploration and development, and is of great significance for promoting the effective development of shale oil and gas resources. BRIEF DESCRIPTION OF THE DRAWINGS
[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without paying any creative work.
[0041] Figure 1 This is a grayscale image of the lamellae development in the shale core oil and gas reservoir obtained based on CT scanning and reconstruction.
[0042] Figure 2 To obtain a digital quantitative model map of the foliation development of oil and gas layer core reservoirs through machine learning.
[0043] Figure 3 To obtain digital quantitative result curves of reservoir lamellae development in oil and gas layer cores through machine learning. DETAILED DESCRIPTION
[0044] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.
[0045] The flowchart used in the present invention shows the operations implemented according to some embodiments of the present invention. It should be understood that the operations of the flowchart can be implemented out of order, and the steps without logical context can be reversed or implemented simultaneously. Under the guidance of the content of the present invention, those skilled in the art can add one or more other operations to the flowchart, and can also remove one or more operations from the flowchart. These functional entities can be implemented in software form.
[0046] Reference to "embodiments" in the present invention means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.
[0047] like Figure 1-3 As shown, the present invention provides a machine learning-based quantitative characterization method for shale reservoir lamellae development, the method comprising:
[0048] S101. Obtain grayscale images of the lamination structure of shale core oil and gas reservoirs based on CT scanning and reconstruction, and construct a digital model of shale core lamination based on the grayscale images of the lamination structure of shale core oil and gas reservoirs. The digital model of shale core lamination is a complete three-dimensional digital model of the core including lamination, which highlights the lamination by improving the resolution and other methods, so that subsequent machine learning can improve learning efficiency. Figure 1 As shown in FIG. 1 , the most original model image after CT scanning reconstruction is a grayscale image, including foliation, minerals, pores, cracks, etc., and foliation, etc. are all displayed in grayscale images.
[0049] Among them, the grayscale image of the lamella structure of the shale core oil and gas reservoir is obtained, including:
[0050] Obtain the geological background of the target shale layer and determine the CT scanning parameters of the shale core based on the geological background;
[0051] The core is scanned based on the shale core CT scanning parameters to obtain the grayscale image of the lamella structure of the shale core oil and gas reservoir.
[0052] Furthermore, the core is scanned based on the shale core CT scanning parameters to obtain a grayscale image of the lamella structure of the shale core oil and gas reservoir, including:
[0053] Perform micro-nano CT scanning on the plunger core based on micro-nano CT scanning parameters to obtain a micro-nano grayscale image;
[0054] The micron grayscale image and the micro-nano grayscale image are fused at multiple scales to obtain the grayscale image of the lamella structure of the shale core oil and gas reservoir.
[0055] S102. Based on the digital model of shale core foliation, a standard machine learning model of shale core oil and gas reservoir foliation structure is constructed through machine learning of CT scan images, including:
[0056] Based on machine learning of micron CT scan images, a standard machine learning model for micron grayscale images of foliation structures is obtained;
[0057] Based on the machine learning of micro-nano CT scanning images, a standard model for machine learning of micro-nano grayscale images of foliation structure is obtained.
[0058] S103. Based on the standard machine learning model of lamella structure, a digital result model of lamella development in oil and gas layer core reservoir is obtained through machine learning, including:
[0059] Identify the grayscale image of the lamella structure of shale core oil and gas reservoirs to obtain the lamella and structure characterization parameters;
[0060] Acquire screening criteria for the foliation characterization parameters, and determine a plurality of target foliation and structure characterization parameters from the foliation and structure characterization parameters based on the screening criteria;
[0061] A digital model of shale core foliation is constructed based on multiple target foliation and structural characterization parameters.
[0062] like Figure 2 As shown in the figure, the digital quantitative model diagram of the lamellar development of the oil and gas layer core reservoir is obtained through machine learning. It is a three-dimensional digital quantitative result diagram reflecting the degree of lamellar development extracted from the shale grayscale image. Figure 3 As shown in the figure, the digital quantitative result curve of lamellar development in the core reservoir of the oil and gas layer obtained by machine learning is a numerical curve of the digital quantitative model of lamellar development, which quantitatively reflects the degree, change and law of lamellar development. The quantitative model can only be displayed visually, and cannot directly give people the numerical value and change characteristics. Only images cannot be used for quantitative comprehensive evaluation.
[0063] S104. Based on the digital model of reservoir lamellae development, construct the target interval with advantageous lamellae in the oil and gas layer, including:
[0064] Based on the digital result model of lamellae development in core reservoirs of oil and gas layers, the coupling relationship between lamellae and oil and gas saturation is studied through random judgment method, and the shale lamellae evaluation standard and lamellae evaluation optimization model are established. Then, the lamellae-dominant target layer sections of oil and gas layers are constructed, providing a scientific basis for the in-depth exploration and development of shale oil and gas.
[0065] The embodiment of the present invention provides a machine learning-based quantitative characterization method for shale reservoir foliation development. First, grayscale images of shale core oil and gas reservoir foliation structures are obtained based on CT scanning and reconstruction, and a shale core foliation digital model is constructed based on the grayscale images of shale core oil and gas reservoir foliation structures, thereby improving the accuracy of the constructed three-dimensional digital model of shale foliation structure characteristics, thereby improving the accuracy and efficiency of shale oil and gas exploration and development predictions. The comprehensiveness and reliability of the shale foliation development evaluation and optimization model are improved, so that the accuracy of the target foliation layer determined based on the shale foliation development evaluation and optimization model can be further improved, thereby improving the efficiency of shale oil and gas exploration and development and reducing the cost of oil and gas exploration and development.
[0066] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.
[0067] The contents not described in detail in the specification of the present invention belong to the prior art known to the professional and technical personnel in this field.
Claims
1. A machine learning-based quantitative characterization method for shale reservoir lamellae development, characterized in that: include: S1. Obtaining grayscale images of the lamellar structure of shale core oil and gas reservoirs based on CT scanning and reconstruction, and constructing a digital model of the shale core lamellar structure based on the grayscale images of the lamellar structure of shale core oil and gas reservoirs; The method of obtaining a grayscale image of the lamellar structure of a shale core oil and gas reservoir comprises: Obtaining the geological background of the target shale layer, and determining the shale core CT scanning parameters based on the geological background; Scanning the core based on the shale core CT scanning parameters to obtain a grayscale image of the lamella structure of the shale core oil and gas reservoir; Scanning the core based on the shale core CT scanning parameters to obtain a grayscale image of the lamella structure of the shale core oil and gas reservoir includes: Perform micron CT scanning on the full-diameter core based on micron CT scanning parameters to obtain a micron grayscale image; Perform micro-nano CT scanning on the plunger core based on micro-nano CT scanning parameters to obtain a micro-nano grayscale image; Performing multi-scale fusion of the micron grayscale image and the micro-nano grayscale image to obtain a grayscale image of the lamella structure of the shale core oil and gas reservoir; The method of constructing a shale core lamellae digital model based on the grayscale image of the lamellae structure of the shale core oil and gas reservoir comprises: Identify the grayscale image of the lamella structure of the shale core oil and gas reservoir to obtain lamella and structure characterization parameters; Acquire screening criteria for the foliation characterization parameters, and determine a plurality of target foliation and structure characterization parameters from the foliation and structure characterization parameters based on the screening criteria; Constructing the shale core lamellae digital model based on the multiple target lamellae and structural characterization parameters; S2. Based on the digital model of shale core foliation, a standard machine learning model of shale core oil and gas reservoir foliation structure is constructed through machine learning of CT scan images; The method of constructing a machine learning standard model of shale core oil and gas reservoir foliation structure based on the shale core foliation digital model and through machine learning of CT scan images includes: Based on machine learning of micron CT scan images, a standard machine learning model for micron grayscale images of foliation structures is obtained; Based on the machine learning of micro-nano CT scanning images, a standard machine learning model for micro-nano grayscale images of foliation structures is obtained; S3, based on the machine learning standard model of shale core oil and gas reservoir lamellae structure, obtain the digital result model of lamellae development in oil and gas layer core reservoir through machine learning; S4. Based on the digital model of reservoir lamellae development, construct the target interval with dominant lamellae in the oil and gas layer.
2. The machine learning-based quantitative characterization method for shale reservoir lamellae development according to claim 1 is characterized in that: The method of constructing a target interval with advantageous lamellae in the oil and gas layer based on a digital model of reservoir lamellae development includes: Based on the digital result model of lamellae development in core reservoirs of oil and gas layers, the shale lamellae evaluation standard and lamellae evaluation optimization model are established through the random judgment method, and then the lamellae-dominant target intervals of oil and gas layers are constructed.
