Thickened oil combination flooding effect evaluation method based on image processing technology

Abstract

The invention discloses a heavy oil combination flooding effect evaluation method based on an image processing technology. According to the design scheme, firstly, pores and throats are extracted according to casting body sheet data, and a microcosmic visual model is established; secondly, carrying out a heavy oil displacement experiment, and continuously recording image data in an oil-water flowing process through continuous image acquisition; then, extracting each piece of remaining oil at each time point based on a deep learning image segmentation technology, and classifying the remaining oil through shape parameters and a contact relationship; and finally, the recovery ratio is calculated according to the change of the remaining oil area, and the heavy oil combination flooding effect is evaluated according to the recovery ratio, the change of the remaining oil type and the oil drop size distribution. According to the technology, an image processing technology is utilized, the dynamic change and the recovery ratio of the remaining oil are accurately analyzed in real time, a more efficient and more accurate effect evaluation tool is provided for heavy oil combination flooding, oil displacement scheme optimization and recovery ratio improvement are facilitated, and an important decision basis is provided for oil reservoir management.

Description

Technical Field

[0001] The present invention relates to a method for evaluating the effect of heavy oil composite flooding based on image processing technology, and belongs to the technical field of oil and gas field development. Background Art

[0002] As an important energy resource, heavy oil has high viscosity and low fluidity, which poses great challenges to traditional oil production technologies in the exploitation of heavy oil. In order to improve the recovery rate of heavy oil, composite flooding technologies such as thermal recovery and chemical flooding are usually adopted. The composite flooding technology injects displacing agents such as polymers, surfactants, and alkalis to optimize the oil reservoir recovery process based on improving fluid mobility and reducing the oil-water interfacial tension. In heavy oil development, the composite flooding technology has become a major means of cold production, especially when the conventional thermal oil production technology fails to achieve the expected effect, the application of the composite flooding technology has important strategic significance.

[0003] The research and optimization of the composite flooding technology rely on an in-depth understanding of the microscopic mechanism of the oil reservoir. Due to the viscosity and complex flow characteristics of heavy oil, traditional macroscopic experimental methods are difficult to accurately reflect the interaction between residual oil and displacing agents. Therefore, microscopic visualization experiments for studying the mechanism of composite oil displacement have become an important means in this field. The microscopic visualization experiment uses a glass model engraved with pore-throat networks to conduct water flooding experiments, and the oil-water flow process in the model can be directly observed through a high-power microscope, which can be used to study the changes of residual oil, the evolution of the oil-water interface, and the action effects of different displacing agents during the composite flooding process. However, the current quantitative analysis methods for residual oil still have deficiencies: most of the existing analysis methods focus on the single analysis of residual oil, and do not consider the contact situation between residual oil and the rock skeleton. The classification method based on the morphology of crude oil often has certain subjectivity and errors. At the same time, the existing research is usually limited to static analysis, that is, only analyzing the data at a certain moment of the oil-water distribution, lacking real-time monitoring and evaluation of the dynamic evolution of residual oil during the composite flooding process. CN115951022A discloses a method for evaluating the oil displacement effect of a displacing agent for heavy oil chemical flooding. This method obtains the types of microscopic residual oil according to microscopic visualization displacement experiments, and then evaluates the oil displacement effect of the displacing agent used in chemical flooding according to the content changes of various types of microscopic residual oil after chemical flooding. However, this method does not propose specific residual oil classification criteria, has strong subjectivity and will bring certain errors;

[0004] Therefore, there is an urgent need for a more accurate and efficient method for evaluating the effect of composite flooding. Summary of the Invention

[0005] In view of the above problems, the present invention provides a method for evaluating the effect of heavy oil composite flooding based on image processing technology, which accurately evaluates the effect of composite flooding by real-time monitoring the distribution and morphological changes of residual oil, thereby improving the recovery rate.

[0006] The present invention adopts the following technical solutions:

[0007] An evaluation method for the effect of heavy oil composite flooding based on image processing technology, comprising the following steps:

[0008] S1. According to the representative cast thin section data, use image processing technology to extract pores and throats, and use a lithography machine to establish a microscopic visualization model;

[0009] S2. Use a high-temperature and high-pressure microscopic visualization experimental device to conduct a displacement experiment, and through real-time continuous image acquisition, obtain the oil-water flow process and the remaining oil distribution characteristics;

[0010] S3. Based on the deep learning image segmentation technology, extract the rock skeleton and each remaining oil at each time step and number them, and then calculate the area occupied by each remaining oil, the shape parameters, the contact relationship with the rock skeleton, and the corresponding contact line length;

[0011] S4. According to the shape parameters of the remaining oil and the contact relationship with the rock skeleton, judge the type of each remaining oil, including droplet-shaped remaining oil, film-shaped remaining oil, blind-end remaining oil, and cluster-shaped remaining oil, and count the percentage of each type of remaining oil in the total remaining oil;

[0012] S5. Calculate the recovery rate according to the area occupied by the remaining oil at each time step during the displacement process, and evaluate the oil displacement effect of the heavy oil composite flooding by obtaining the change situation of each type of remaining oil during the displacement process.

[0013] Preferably, in step S1, the steps for making the microscopic visualization model include:

[0014] S11. Adjust the contrast and brightness, and switch the color mode of the image to Lab Stack;

[0015] S12. Select the channel with the most obvious contrast between pores and the skeleton, and establish a pore space image through Threshold threshold segmentation;

[0016] S13. Convert the image into a vector file, connect the lithography machine, and lithograph the model on the glass to make a microscopic visualization model.

[0017] Preferably, in step S2, the process of the experiment includes:

[0018] S21. Assemble the model and the kettle cover, and evacuate the model and the pipeline;

[0019] S22. Inject pure water into the high-temperature and high-pressure kettle body, and close the drain valve after it is full;

[0020] S23. Increase the confining pressure inside the autoclave to 2 MPa, then inject crude oil to establish an internal pressure of 1 MPa. Cycle this process, gradually increase the confining pressure of the high-temperature and high-pressure autoclave and the internal pressure of the model to the formation pressure, and keep the confining pressure always higher than the internal pressure during the process;

[0021] S24. After the model is completely saturated with oil, turn on the heating device of the autoclave, heat the whole autoclave to the formation temperature, and then heat and balance for 2 hours;

[0022] S25. Water flooding: Use an injection pump to inject water into the model at a certain speed until no oil comes out at the outlet end of the model;

[0023] S26. Chemical flooding: Use an injection pump to inject 0.3 PV of polymer, surfactant or composite flooding system into the model at a certain speed;

[0024] S27. Subsequent water flooding: Use an injection pump to inject water into the model at a certain speed until no oil comes out at the outlet end of the model;

[0025] S28. The image data of the whole process is collected at high frequency and the dynamic change process of the remaining oil is fed back in real time.

[0026] Preferably, in step S2, the composite flooding experiment includes different types of oil displacement methods such as polymer flooding, surfactant flooding, and alkali flooding. The image data is combined with different oil displacement methods for comparative analysis to optimize the oil displacement effect;

[0027] Preferably, in step S3, the obtaining through deep learning image segmentation and contact lines includes the following steps:

[0028] S51. Prepare a series of images collected during the displacement process, manually label the oil, water, and skeleton in the images using the LabelMe image annotation tool, and use the MaskR-CNN model for training to obtain the bitmap of the remaining oil distribution and number each remaining oil, and at the same time obtain the bitmap of the skeleton distribution;

[0029] S52. Perform dilation operations on each remaining oil in sequence according to the number. After dilation once, perform an AND operation on the dilated remaining oil bitmap and the rock skeleton bitmap. If the return value is 1, it means the remaining oil is in contact with the rock skeleton, and the result of the AND operation is the contact line between the remaining oil and the rock skeleton; if the return value is 0, it means the remaining oil is not in contact with the rock skeleton. Finally, number each remaining oil corresponding to its contact line.

[0030] S53. Calculate the dimensionless length of the contact line of each remaining oil. The dimensionless length of the contact line = contact line length / remaining oil perimeter. If the remaining oil is not in contact with the skeleton, the contact line length is 0.

[0031] 6. The evaluation method for the effect of heavy oil composite flooding based on image processing technology according to claim 1, wherein in step S3, the shape parameters include aspect ratio and shape factor, which are used to judge the type of remaining oil. The aspect ratio = W / H, where W is the width of the remaining oil boundary rectangle and H is the height of the remaining oil boundary rectangle; the shape factor F = 4πS / L 2 , where S is the area of the remaining oil; L is the perimeter of the remaining oil; and π is the ratio of the circumference of a circle to its diameter.

[0032] 7. The evaluation method for the effect of heavy oil composite flooding based on image processing technology according to claim 1, wherein in step S4, the classification of the remaining oil types is carried out according to the following criteria:

[0033] Film-like remaining oil: the contact line length < 0.5, the aspect ratio > 3, and the shape factor < 0.3;

[0034] Cluster-like remaining oil: the contact line length > 0.5, the aspect ratio < 3, and the shape factor < 0.3;

[0035] Blind-end remaining oil: the contact line length > 0.5, the aspect ratio < 1.2, and 0.3 < the shape factor < 0.85;

[0036] Oil-droplet-like remaining oil: the contact line length < 0.25, the aspect ratio < 1.2, and 0.5 < the shape factor < 1.

[0037] Preferably, in step S5, the recovery factor = the total area of the current remaining oil / the total area of the initial crude oil; the evaluation of the oil displacement effect is carried out by the following method: if the film-like remaining oil is significantly reduced, it indicates that the oil displacement system can improve the oil washing efficiency of the aqueous phase; if the blind-end remaining oil is significantly reduced, it indicates that the oil displacement system has viscoelasticity and can increase the shear force of the oil displacement system to improve the oil washing efficiency. The improvement amplitude of the oil washing efficiency of the composite system = (the current content of film-like remaining oil + the content of blind-end remaining oil) / (the content of film-like remaining oil + the content of blind-end remaining oil at the end of water flooding); if the cluster-like remaining oil is significantly reduced, it indicates that the oil displacement system has improved the sweep efficiency of the aqueous phase. The improvement amplitude of the sweep efficiency of the composite system = the current content of cluster-like remaining oil / the content of cluster-like remaining oil at the end of water flooding; if the number of emulsified oil droplets increases, it indicates that the oil displacement system has emulsifying ability and can form an oil-in-water emulsion with the remaining oil. The emulsifying ability of the composite system = the number of emulsified oil droplets / the content of oil-droplet-like remaining oil, and the smaller the value, the stronger the emulsifying ability.

[0038] Compared with the prior art, the beneficial effects of the present invention are as follows: First, pores and throats are extracted from the cast thin section data to establish a microscopic visualization model; Second, heavy oil water flooding and composite flooding experiments are carried out, and image data during the oil-water flow process are continuously recorded through real-time continuous image acquisition; Then, based on the deep learning image segmentation technology, the remaining oil distribution at each time point is extracted, and the remaining oil is classified by shape parameters; Finally, the recovery factor is calculated according to the change of the remaining oil area, and the effect of the composite flooding is evaluated in real time. This technology provides a more efficient and accurate effect evaluation tool for heavy oil composite flooding by analyzing the dynamic changes of the remaining oil and the recovery factor in real time and accurately, which helps to optimize the oil displacement plan, improve the recovery factor, and provide an important decision-making basis for reservoir management.

[0039] The model making method of the present invention can be programmed, is simple and practical, can obtain the oil displacement mechanism, emulsification effect and change characteristics of the remaining oil of the composite system through real-time processing of experimental images, avoids the errors and additional workload caused by manual processing, and can better characterize the oil displacement effect of the composite system, providing a theoretical basis for guiding the formulation of the viscosity reduction composite flooding plan for heavy oil cold production. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings of the embodiments will be briefly introduced below. Obviously, the drawings described below only relate to some examples of the present invention and do not limit the present invention.

[0041] Figure 1 It is the microscopic visualization lithography model constructed according to the cast thin section of the present invention;

[0042] Figure 2 It is the remaining oil distribution map extracted by the present invention;

[0043] Figure 3 It is the remaining oil distribution map of different types extracted by the present invention using an algorithm;

[0044] Figure 4 It is the radius distribution map of droplet-shaped remaining oil obtained by the present invention;

[0045] Figure 5 It is the change diagram of the proportion of different types of remaining oil at different oil displacement stages obtained by the present invention;

[0046] Figure 6 It is the recovery factor change diagram of different oil displacement stages of heavy oil composite flooding obtained by the present invention;

[0047] Figure 7 It is the comparison diagram of the remaining oil types of different oil displacement systems obtained by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0048] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0049] As Figures 1 to 7 shown, a method for fabricating a microscopic visualization lithography model and its experimental process include the following steps:

[0050] Step 1: Based on the technical solution S1, select a representative thin section image of the casting, adjust the contrast and brightness to make the pores and the rock skeleton clear and bright; switch the color model of the image to the Lab mode, save the a channel (red-green channel) with the most obvious contrast between the pores and the skeleton, use Threshold threshold segmentation to distinguish between the pores and the rock skeleton, then extract the contour of the pore image and convert it into a vector file, import it into the Ezcad software, connect the lithography machine, and lithograph the model on the glass to fabricate a microscopic visualization model, as Figure 1 shown.

[0051] Step 2: Based on the technical solution S2, use a high-temperature and high-pressure microscopic visualization experimental device to conduct a microscopic visualization displacement experiment under formation temperature and pressure conditions. Inject 0.3PV of polymer, surfactant, or composite system respectively after water flooding, and then continue water flooding until no oil comes out at the outlet end. The experimental conditions and the properties of the simulated water, polymer, and viscosity reducer used are shown in Table 1. During the experiment, the oil-water flow process in the model and the remaining oil distribution characteristics are obtained through real-time continuous image acquisition.

[0052] Table 1 Experimental conditions and fluid properties of heavy oil composite flooding

[0053] Experimental temperature Experimental pressure Crude oil viscosity Simulated water salinity System viscosity System interfacial tension 60℃ 14 MPa 620 mP·s 5700 mg / L 23 mP·s 0.1 nN / m

[0054] Step 3: Based on the technical solutions S3 and S4, obtain the remaining oil distribution as Figure 2 shown. The area, shape parameters, dimensionless contact line length of the obtained remaining oil, and the remaining oil types judged according to the parameters are shown in Table 2, Figure 3 shown. According to the emulsified oil droplet radius distribution (as Figure 4 shown), it can be found that the oil droplet radius is mainly concentrated in the range of 10 - 20 μm, indicating that the emulsification effect of the composite system is better.

[0055] Table 2 Remaining oil shape parameters and remaining oil types

[0056]

[0057] Table 3 Evaluation table of heavy oil composite flooding effect

[0058]

[0059] Step 4: Based on the technical solution S5, the remaining oil types during the displacement process are as Figure 5As shown, the changes in the composite flooding oil recovery rate at different times are as follows Figure 6 As shown, the remaining oil distributions of polymer flooding, viscosity reducer flooding, and composite system flooding are as follows Figure 7 As shown, the evaluation results of heavy oil composite flooding are shown in Table 3. It can be found that there are a large number of cluster-shaped and film-shaped remaining oils, and a small amount of blind-end and oil-drop-shaped remaining oils after water flooding; surfactant flooding mainly mobilizes the film-shaped remaining oil, and the oil washing efficiency increases by 74.56%. The oil-drop-shaped remaining oil increases to some extent, indicating that the surfactant has good emulsifying ability. Polymer flooding mainly mobilizes the cluster-shaped remaining oil, and the sweep efficiency increases by 57.89%; while the cluster-shaped, film-shaped, and blind-end remaining oils of the composite system flooding are all mobilized to varying degrees, and the oil-drop-shaped remaining oil increases to some extent, indicating that the composite flooding can not only improve the sweep efficiency but also expand the sweep efficiency. The increase in sweep efficiency and oil washing efficiency are 75.84% and 75.83% respectively, and the emulsifying ability is also enhanced compared with that of single surfactant flooding

[0060] The above are only the preferred embodiments of the present invention, and do not impose any form of limitation on the present invention. Although the present invention has been disclosed above with the preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some changes or modifications to the above-disclosed technical content to be equivalent embodiments within the scope of the technical solution of the present invention. However, any simple modification, equivalent change, and modification made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention still fall within the scope of the technical solution of the present invention

Claims

1. A method for evaluating the effect of heavy oil composite flooding based on image processing technology, characterized in that: The following steps are involved: S1. Based on the representative casting thin section data, the pores and throats were extracted using image processing technology, and a microscopic visualization model was established using a photolithography machine; S2. Use high temperature and high pressure microscopic visualization experimental equipment to conduct displacement experiments and obtain the oil-water flow process and residual oil distribution characteristics through real-time continuous image acquisition; S3, based on deep learning image segmentation technology, the rock skeleton and each residual oil at each time step are extracted and numbered, and then the area occupied by each residual oil, shape parameters, contact relationship with the rock skeleton and corresponding contact line length are calculated; S4. According to the shape parameters of the remaining oil and the contact relationship with the rock skeleton, the type of each remaining oil is determined, including oil drop-shaped remaining oil, film-shaped remaining oil, blind-end remaining oil, and cluster-shaped remaining oil, and the percentage of each type of remaining oil in the total remaining oil is counted; S5. Calculate the recovery factor based on the area occupied by the remaining oil at each time step during the displacement process, and evaluate the oil recovery effect of the heavy oil composite flooding by obtaining the changes in each type of remaining oil during the displacement process.

2. The method for evaluating the effect of heavy oil composite flooding based on image processing technology according to claim 1 is characterized in that: In step S1, the microscopic visualization model making step includes: S11, adjust contrast and brightness, and switch the image color mode to Lab Stack; S12, select the channel with the most obvious contrast between pores and skeleton, and establish the pore space image through Threshold segmentation; S13. Convert the image into a vector file, connect it to a photolithography machine, and photolithograph the model on glass to make a microscopic visualization model.

3. The method for evaluating the effect of heavy oil composite flooding based on image processing technology according to claim 1 is characterized in that: In step S2, the experimental process includes: S21, assemble the model and the kettle cover, and evacuate the model and pipelines; S22, injecting pure water into the high temperature and high pressure autoclave body, and closing the drain valve after it is full; S23, increasing the confining pressure in the kettle to 2 MPa, then injecting crude oil to establish an internal pressure of 1 MPa, and repeating this process, gradually increasing the confining pressure of the high-temperature and high-pressure kettle and the internal pressure of the model to the formation pressure, and keeping the confining pressure higher than the internal pressure throughout the process; S24, after the model is completely saturated with oil, the kettle heating device is turned on, the kettle is heated to the formation temperature, and then heated for 2 hours for equilibrium; S25, water drive: use the injection pump to inject water into the model at a certain speed until no oil comes out of the model outlet; S26, chemical flooding: use an injection pump to inject 0.3PV of polymer, surfactant or composite flooding system into the model at a certain speed; S27, subsequent water drive: use the injection pump to inject water into the model at a certain speed until no oil comes out of the model outlet; S28. The image data of the whole process is collected at a high frequency to provide real-time feedback on the dynamic changes of the remaining oil.

4. The method for evaluating the effect of heavy oil composite flooding based on image processing technology according to claim 1 is characterized in that: In step S2, the composite flooding experiment includes different types of oil flooding methods such as polymer flooding, surfactant flooding, alkali flooding and composite flooding. The image data is compared and analyzed in combination with different oil flooding methods to optimize the oil flooding effect.

5. The method for evaluating the effect of heavy oil composite flooding based on image processing technology according to claim 1 is characterized in that: In step S3, the image segmentation and contact line acquisition by deep learning include the following steps: S51, preparing a series of images collected during the displacement process, and manually labeling the oil, water, and skeleton in the images using the LabelMe image annotation tool, and training using the MaskR-CNN model to obtain a bitmap of the remaining oil distribution and number each remaining oil, and at the same time obtain a bitmap of the skeleton distribution; S52, perform expansion operation on each residual oil in sequence according to the number. After one expansion, perform AND operation on the expanded residual oil bit map and the rock skeleton bit map. If the return value is 1, it means that the residual oil is in contact with the rock skeleton, and the result of the AND operation is the contact line between the residual oil and the rock skeleton. If the return value is 0, it means that the residual oil is not in contact with the rock skeleton. Finally, each residual oil is numbered corresponding to its contact line. S53. Calculate the dimensionless length of the contact line of each residual oil. The dimensionless length of the contact line = contact line length / circumference of the residual oil. If the residual oil does not contact the skeleton, the contact line length is 0.

6. The method for evaluating the effect of heavy oil composite flooding based on image processing technology according to claim 1 is characterized in that: In step S3, the shape parameters include aspect ratio and shape factor, which are used to determine the type of residual oil. Aspect ratio = W / H, where W is the width of the residual oil boundary rectangle and H is the height of the residual oil boundary rectangle; shape factor F = 4πS / L 2 , where S is the area of ​​the remaining oil; L is the circumference of the remaining oil; π is the pi.

7. The method for evaluating the effect of heavy oil composite flooding based on image processing technology according to claim 1 is characterized in that: In step S4, the residual oil type classification is performed according to the following standards: Film residual oil: contact line length <0.5, aspect ratio >3, shape factor <0.3; Clustered residual oil: contact line length>0.5, aspect ratio<3, shape factor<0.3; Blind end residual oil: contact line length>0.5, aspect ratio<1.2, 0.3<shape factor<0.85; Droplet-shaped residual oil: contact line length <0.25, aspect ratio <1.2, 0.5<shape factor <1.

8. The method for evaluating the effect of heavy oil composite flooding based on image processing technology according to claim 1 is characterized in that: In step S5, recovery factor = area occupied by current remaining oil / initial total oil-bearing area; oil displacement effect evaluation is performed by the following method: improvement of composite system oil washing efficiency = (area occupied by current film-like remaining oil + area occupied by blind-end remaining oil) / (area occupied by film-like remaining oil at the end of water flooding + area occupied by blind-end remaining oil); improvement of composite system sweep efficiency = area occupied by current cluster-like remaining oil / area occupied by cluster-like remaining oil at the end of water flooding; emulsification capacity of composite system = number of oil droplets of remaining oil / area occupied by oil droplets of remaining oil, the smaller the value, the stronger the emulsification capacity.

Images