Microfluidic Image Analysis System and Method for Pore Utilization Efficiency in Porous Media

Abstract

The present invention relates to the technical fields of oil and gas exploitation and CO2 geological sequestration, and specifically discloses a microfluidic image analysis system and method for the pore utilization efficiency in a porous medium, including the following steps: binarize the image recorded by the image acquisition module R, execute the pore segmentation module P to segment the pore and particle regions in the microfluidic image; execute the multi-level size pore space generation module S to determine the pore distribution regions with different top-down radii; execute the fluid segmentation module F to determine the occurrence and distribution regions of the original (saturated) fluid; execute the characterization module E of the pore utilization efficiency to determine the area of the original fluid in the pores with different top-down radii and determine the utilization efficiency in the pores of different sizes. The present invention can batch-complete the quantitative characterization of the utilization efficiency in the multi-level size pores by using image processing technology, providing important data support for understanding the performance of the oil displacement fluid and the fluid migration characteristics and formulating / optimizing the plan for improving the oil recovery rate.

Description

Technical Field

[0001] The present invention belongs to the technical field of oil and gas exploitation and CO2 geological sequestration, and particularly relates to a microfluidic image analysis system and method for the pore utilization efficiency in porous media. Background Art

[0002] Microfluidics is an advanced technology that can realize in-situ detection and real-time observation of fluid migration characteristics in porous media. In recent years, it has been highly favored in the research of oil and gas exploitation, CO2 geological sequestration, porous media seepage and other (energy geology) fields. However, the quantification of fluid migration phenomena is a key step for the research in the field of energy geology to move from qualitative to quantitative.

[0003] Fluid migration in porous media is a common problem that must be faced in fields such as oil and gas exploitation and CO2 geological sequestration.

[0004] The interface between the CO2 geological sequestration invading fluid (CO2) and the displaced fluid (crude oil or brine) will advance unevenly in the heterogeneous pore space.

[0005] During the enhanced oil recovery process in oil and gas development, viscous oil-displacing fluids will undergo liquid flow diversion in the pore space to displace more crude oil, which will result in great differences in the utilization efficiency of pores of different sizes, directly affecting the oil recovery rate and the efficiency of CO2 geological sequestration. In recent years, with the rise of technologies such as CT and microfluidics, it has been possible to achieve visual detection of fluid migration in porous media. By means of the image processing and characterization technology for the captured images, the research on fluid migration behavior is no longer limited to qualitative description. The quantification of the utilization efficiency in such complex pores plays a key data support role in understanding the fluid migration behavior and transport mechanism, and also plays an important auxiliary role in the design and optimization of oil and gas exploitation and geological sequestration schemes.

[0006] The nuclear magnetic resonance T2 signal inversion method is a common means for evaluating the pore utilization efficiency, and the utilization efficiency is characterized by using the T2 signal area and the converted pore throat radius. Different from the nuclear magnetic signal inversion method, the microfluidic method can directly observe the fluid migration behavior in pores by virtue of the convenience of optical images. In a similar scheme, CN 113740224 A proposed a method combining pore partitioning and tracer particles to realize the identification of microscopic utilized pores in porous media. CN 114607368A proposed a temperature-controlled infrared imaging scanning method to distinguish the water flow path and the swept path by using thermal imaging images. CN111507986B proposed a method for characterizing the pore and utilized pore distribution based on the pore central axis as a judgment to quantify the swept situation in pores.

[0007] The sweep of pores mentioned in the above solution is different from the mobilization of pores. Taking crude oil extraction as an example, the crude oil in the pores will be displaced by the invading fluid, but it will not be completely displaced. The remaining oil that has not been displaced adheres to the particle wall surface in the form of an oil film. Although the pores are swept, the mobilization efficiency is not 100%. Although CN 111507986 B provides a method for quantifying the sweep efficiency in pores, there is currently no solution that can characterize the pore mobilization efficiency on an optical image. Summary of the Invention

[0008] The purpose of the present invention is to provide a microfluidic image analysis system and method for pore mobilization efficiency in a porous medium to solve the problems proposed in the above background technology.

[0009] To achieve the above purpose, the present invention provides the following technical solutions:

[0010] A microfluidic image analysis device for pore mobilization efficiency in a heterogeneous porous medium, comprising:

[0011] Comprising an image acquisition module R, a pore segmentation module P, a multi-level size pore space generation module S, a fluid segmentation module F, and a pore mobilization efficiency characterization module E. The image acquisition module R includes a micro flow injection pump (1), a micro liquid volume pressure sensor (2), a connected computer (3), a microscope and a camera (4), a microfluidic chip (5), and a collection liquid beaker (6).

[0012] A microfluidic image analysis method for pore mobilization efficiency in a porous medium, including placing a micro-scale microfluidic chip (5) under a microscope and a camera (4), injecting a displacement medium using a microfluidic injection pump (1), using the microscope and the camera (4) to photograph and record the microfluidic images in the saturated state, i.e., the initial moment, and different displacement processes, and using the micro liquid volume pressure sensor (2) to record the injection pressure upstream of the chip. It also includes the following steps:

[0013] Running the pore segmentation module P,

[0014] P001: Align the set of pictures to be processed to ensure that the pore regions in different images overlap as much as possible;

[0015] P002: Select a picture at the initial moment of a saturated state as the segmentation object, divide the image into blocks, and the block regions of 4×3 or 5×4 can be used;

[0016] P003: Determine the adaptive segmentation threshold for each block region using the maximum inter-class variance method. Different block regions have different segmentation thresholds. Segment each region according to the corresponding segmentation threshold, and the segmented regions are then aggregated to obtain the segmented pore region;

[0017] Running the multi-level size pore space generation module S,

[0018] S001: Perform distance transformation on the pore segmentation region generated in step P003, which can be specifically implemented by the distanceTransform function in the OpenCV processing package or the bwdist function in Matlab;

[0019] S002: Extract the medial axis skeleton of the pore segmentation region in step P003, which can be specifically implemented by the skeletonize function in the morphology sub-module of OpenCV or the bwmorph function in Matlab;

[0020] S003: Perform an intersection operation on the medial axis skeleton in step S002 and the distance transformation space in step S001 to form a medial axis skeleton with distance dimension fusion, that is, the numerical value of each point on the medial axis skeleton is the pore radius corresponding to that point;

[0021] S004: Count the distance elements on the medial axis skeleton to form a pore size set R p , and ensure that the elements in the pore size set R p are arranged in descending order;

[0022] S005: Traverse the pore size set R in step S004 p ;

[0023] S006: Find the position points with a radius of r i on the medial axis skeleton with distance dimension fusion formed in step S003;

[0024] S007: Perform a circular dilation operation with a radius of r i on the position points found in step S006;

[0025] S008: Change the current radius r i to the next element in the pore size set R p ;

[0026] S009: Determine whether the elements in the pore size set R p have been traversed. If not, re-execute steps S005 - S009; if completed, a multi-level size pore space can be generated;

[0027] Run the fluid segmentation module F,

[0028] F001: Perform a difference operation between the image to be processed and the saturated state image, determine the adaptive segmentation threshold by the maximum inter-class variance method on the difference image, and segment the pore mobilization region;

[0029] Considering that there are fuzzy regions in the original fluid, which will cause insufficient segmentation of the pore mobilization region obtained in step F001, therefore, color image segmentation is performed on the original fluid;

[0030] F002: Label the fuzzy region between the original fluid, i.e., the fluid saturated at the initial moment, and the displacement fluid. Specifically, the roipoly function in Matlab can be used for implementation;

[0031] F003: Calculate the average RGB vector in the labeled region and the covariance matrix C, where the covariance satisfies:

[0032]

[0033] where, n is the number of pixels in the labeled region, R j 、G j and B j are the numerical values in the red, green, and blue spaces of the image respectively;

[0034] F004: The following criterion is used for color image segmentation. The specific criterion is:

[0035]

[0036] where, Th is the segmentation threshold, and z is the RGB vector of the pixel point to be evaluated;

[0037] F005: The fuzzy region obtained by color image segmentation is subjected to a union operation with the pore mobilization region obtained in step F001, and then an intersection operation is performed with the pore region segmented in step P003 to obtain the segmentation region of the original fluid;

[0038] Run the pore mobilization efficiency characterization module E,

[0039] E001: Traverse each multi-level size pore space formed in step S009 one by one according to the pore size;

[0040] E002: Locate the pore space with a radius of r i and measure the area occupied by this pore space

[0041] E003: Locate the region where the original fluid is stored within the pore space with a radius of r i and measure the storage area of this original fluid

[0042] E004: Calculate the pore mobilization efficiency E of the pore with a radius of ri i , and the specific calculation is

[0043] E005: Change the current radius ri to the next element in the pore size set R p ;

[0044] E006: Determine whether the elements in the pore size set R p have been traversed. If not, re-execute steps E001 - E005; if completed, the utilization efficiency in pores with multi-level sizes can be obtained;

[0045] E007: The pore size shown in step E006 is the top-view radius R of the pore t , and the depth information of the microfluidic chip needs to be considered. Convert the top-view radius R t to the hydraulic radius R hyd , and the conversion formula is as follows:

[0046]

[0047] where A is the cross-sectional area of the pore, and I p is the polar moment of inertia;

[0048] The cross-section of the pores in the dry-etched chip is usually rectangular. Considering the influence of the etching depth D p , the calculation formula for the hydraulic radius of the rectangular cross-section is as follows:

[0049]

[0050] Using the above formula, convert the top-view pore radius calculated in step E006 to the hydraulic radius, and then obtain the utilization efficiency of pores with multi-level hydraulic radii.

[0051] Compared with the prior art, the beneficial effects of the present invention are:

[0052] The present invention uses image processing technology to batch-quantitatively characterize the utilization efficiency in pores with multi-level sizes, providing important data support for understanding the performance of displacement fluids and fluid migration characteristics for improving oil recovery, formulating / optimizing plans, significantly improving work efficiency, and also enhancing the practicality of the analysis method. At the same time, this method provides a channel for the quantitative research of microfluidics in the geological field. By using the advantages of convenience, low time cost, and visibility of microfluidic devices, the data acquisition time can be compressed by at least 50% compared with the existing mainstream rock sample analysis methods. At the same time, it is also a verification and technical supplement to other means (such as nuclear magnetic resonance, CT and other geological sample detection means). BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 is the microfluidic image analysis characterization process of pore utilization efficiency;

[0054] Figure 2 is the microfluidic device flow chart of pore utilization efficiency;

[0055] Figure 3 It is an implementation example diagram of pore region segmentation;

[0056] Figure 4 It is a formation flow chart of multi-scale pore space;

[0057] Figure 5 It is a schematic diagram of circular dilation operation on the medial axis skeleton;

[0058] Figure 6 It is an implementation example diagram of multi-scale pore space;

[0059] Figure 7 It is an implementation example diagram of the fluid (crude oil) segmentation area;

[0060] Figure 8 It is a characterization flow chart of the utilization efficiency of multi-scale pores;

[0061] Figure 9 It is a result diagram of the implementation example of the original fluid occurrence area within the pore range of different sizes;

[0062] Figure 10 It is a result diagram of the implementation example of the utilization efficiency of multi-scale pores. Detailed implementation manners

[0063] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0064] Embodiment:

[0065] Please refer to Figure 1 - Figure 10 As shown, a microfluidic image acquisition system for the utilization efficiency of pores in heterogeneous porous media includes an image acquisition module R, a pore segmentation module P, a multi-scale pore space generation module S, a fluid segmentation module F, and a pore utilization efficiency characterization module E. The image acquisition module R includes a micro flow injection pump (1), a micro liquid volume pressure sensor (2), a connected computer (3), a microscope and a camera (4), a microfluidic chip (5), and a collection liquid beaker (6);

[0066] A microfluidic image analysis method for the pore utilization efficiency in a porous medium, including placing a microfluidic chip (5) of micron scale under a microscope and a camera (4), injecting a displacement medium by using a microfluidic injection pump (1), using the microscope and the camera (4) to photograph and record the microfluidic images at the saturation state, i.e., the initial moment, and different displacement processes, using a micro-liquid volume pressure sensor (2) to record the injection pressure upstream of the chip, and further including the following steps:

[0067] Run the pore segmentation module P

[0068] Step P001: Align the set of pictures to be processed to ensure that the pore regions in different images coincide as much as possible;

[0069] Step P002: Select a picture at the initial moment of a saturation state as the segmentation object, divide the image into blocks, and the block regions of 4×3 or 5×4 can be used;

[0070] Step P003: Determine the adaptive segmentation threshold for each block region by using the maximum inter-class variance method. Different block regions have different segmentation thresholds, and each region is segmented according to the corresponding segmentation threshold. The segmented regions are then aggregated to obtain the segmented pore region ( Figure 3 ).

[0071] Run the multi-level size pore space generation module S

[0072] Step S001: Perform distance transformation on the pore segmentation region generated in Step P003, which can be specifically implemented by using the distanceTransform function in the OpenCV processing package or the bwdist function in Matlab;

[0073] Step S002: Extract the medial axis skeleton of the pore segmentation region in Step P003, which can be specifically implemented by using the skeletonize function in the morphology sub-module of OpenCV or the bwmorph function in Matlab;

[0074] Step S003: Perform an intersection operation on the medial axis skeleton in Step S002 and the distance transformation space in Step S001 to form a medial axis skeleton with distance size fusion, that is, the numerical value of each point on the medial axis skeleton is the pore radius corresponding to that point;

[0075] Step S004: Count the distance elements on the medial axis skeleton to form a pore size set R p and ensure that the elements in the pore size set Rp are arranged in descending order;

[0076] Step S005: Traverse the pore size set R in Step S004 p ;

[0077] Step S006: Find the position points with a radius of r on the medial axis skeleton with the distance dimensions fused in Step S003 i ;

[0078] Step S007: Perform a circular dilation operation with a radius of r on the position points found in Step S006 i (); Figure 5

[0079] Step S008: Change the current radius r i to the next element in the pore size set R p ;

[0080] Step S009: Determine whether the elements in the pore size set R p have been traversed. If not, re-execute Steps S005 - S009; if completed, a multi-level pore space can be generated Figure 6 ().

[0081] Run the fluid segmentation module F

[0082] Step F001: Perform a difference operation between the image to be processed and the saturated state image, and use the Otsu method to determine the adaptive segmentation threshold in the difference image to segment the pore mobilization area;

[0083] Considering that there are blurred areas in the original fluid, which may cause insufficient segmentation of the pore mobilization area obtained in Step F001. Therefore, color image segmentation is performed on the original fluid.

[0084] Step F002: Label the blurred area between the original fluid (the fluid saturated at the initial moment) and the displacing fluid. Specifically, the roipoly function in Matlab can be used for implementation;

[0085] Step F003: Calculate the average RGB vector and covariance matrix C in the labeled area, and the covariance satisfies:

[0086]

[0087] where, n is the number of pixels in the labeled area, R j , G j and B j are the numerical values in the red, green, and blue spaces of the image respectively;

[0088] Step F004: Perform color image segmentation using the following criterion. The specific criterion is:

[0089]

[0090] ​Among them, T h is the segmentation threshold. z is the RGB vector of the pixel to be evaluated.

[0091] Step F005: Perform a union operation on the blurred area obtained by color image segmentation and the pore mobilization area obtained in step F001, and then perform an intersection operation with the pore area segmented in step P003 to obtain the segmentation area of the original fluid. In a specific embodiment, the original fluid is saturated crude oil. Figure 7 It is an example diagram of the segmentation area of crude oil. The white area is the blurred area segmented from the color image, and the black is the segmented area where the crude oil is stored.

[0092] Run the characterization module E of the pore mobilization efficiency

[0093] Step E001: Traverse each of the multi-level size pore spaces formed in step S009 one by one according to the pore size ([[]] Figure 8 );

[0094] Step E002: Locate the pore space with a radius of r [[[]] i () and measure the area occupied by this pore space [[[]] Figure 9 a and c),

[0095] Step E003: Locate the area where the original fluid is stored within the pore space with a radius of r [[[]] i () and measure the storage area of this original fluid [[[]] Figure 9 b and d),

[0096] Step E004: Calculate the pore mobilization efficiency E [[[]] i of the pore with a radius of r [[[]] i , and the specific calculation is [[[]]

[0097] Step E005: Change the current radius r [[[]] i to the next element in the pore size set R [[[]] p ;

[0098] Step E006: Determine whether the elements in the pore size set R [[[]] p have been traversed. If not, re-execute steps E001 - E005; if completed, the mobilization efficiency within the multi-level size pores can be obtained [[[]] Figure 10 b).

[0099] Step E007: The pore size shown in step E006 is the top view radius R [[[]] t of the pore. Considering the depth information of the microfluidic chip, convert the top view radius R [[[]] t to the hydraulic radius R [[[]] hyd , and the conversion formula is as follows: [[[]]

[0100]

[0101] Among them, A is the cross-sectional area of the pore, and I p is the polar moment of inertia.

[0102] The cross-section of the pores of the dry-etched chip is usually rectangular. Considering the influence of the etching depth D p the calculation formula of the hydraulic radius of the rectangular cross-section is as follows:

[0103]

[0104] Using the above formula, the top-down pore radius calculated in step E006 is converted into the hydraulic radius, and then the displacement efficiency of the multi-stage hydraulic radius pores is obtained ( Figure 10 c).

[0105] In the present invention, by binarizing the image recorded by the image acquisition module R, the pore segmentation module P is executed to segment the pore and particle regions in the microfluidic image; the multi-stage size pore space generation module S is executed to determine the pore distribution regions with different top-down radii; the fluid segmentation module F is executed to determine the occurrence distribution regions of the original (saturated) fluid; and the pore displacement efficiency characterization module E is executed to determine the area of the original fluid in the pores with different top-down radii and determine the displacement efficiency in the pores with different sizes.

[0106] The dimensions include two types: the top-down radius of the pore and the hydraulic radius, and the displacement efficiency of the multi-stage pores under the two conditions of the top-down radius and the hydraulic radius of the pore is obtained.

[0107] The present invention uses image processing technology to batch-quantitatively characterize the displacement efficiency in multi-stage size pores, providing important data support for understanding the performance of displacement fluids and the characteristics of fluid migration for improving oil recovery, formulating / optimizing plans, significantly improving work efficiency, and also improving the practicality of the analysis method.

[0108] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A microfluidic image analysis system for pore utilization efficiency in porous media, characterized in that: The invention comprises an image acquisition module R, a pore segmentation module P, a multi-level pore space generation module S, a fluid segmentation module F and a pore utilization efficiency characterization module E. The image acquisition module R comprises a micro-flow injection pump (1), a micro-liquid pressure sensor (2), a connection computer (3), a microscope and a camera (4), a microfluidic chip (5) and a collection liquid beaker (6). The invention comprises placing a micron-scale microfluidic chip (5) under the microscope and the camera (4), injecting a displacement medium by using the micro-flow injection pump (1), photographing and recording a saturated state, i.e., an initial moment, and microfluidic images of different displacement processes by using the microscope and the camera (4), and recording the injection pressure upstream of the chip by using the micro-liquid pressure sensor (2). The invention also comprises the following steps: Run the pore segmentation module P, P001: Align the image sets to be processed to ensure that the pore areas in different images overlap as much as possible; P002: Select a saturated initial moment image as the segmentation object and divide the image into blocks. You can use 4×3 or 5×4 block areas. P003: The maximum inter-class variance method is used to determine the adaptive segmentation threshold for each block area. Different block areas have different segmentation thresholds. Each area is segmented according to the corresponding segmentation threshold. The segmented areas are then aggregated to form the segmented pore areas. Run the multi-level pore space generation module S, S001: Perform distance transformation on the pore segmentation area generated in step P003, which can be implemented by the distanceTransform function in the OpenCV processing package or the bwdist function in Matlab; S002: extracting the medial skeleton of the pore segmentation region in step P003, which can be implemented by the skeletonize function of the morphology submodule in OpenCV or the bwmorph function in Matlab; S003: performing an intersection operation on the medial axis skeleton in step S002 and the distance transformation space in step S001 to form a medial axis skeleton with distance and size fusion, that is, the value of each point on the medial axis skeleton is the pore radius corresponding to the point; S004: Count the distance elements on the central axis skeleton to form the pore size set R p , and ensure that the pore size set R p The elements are arranged in order from largest to smallest; S005: For the pore size set R in step S004 p Conduct traversal; S006: Find a radius r on the mid-axis skeleton formed by the distance dimension fusion in step S003. i The location point; S007: For the position point found in step S006, a i Circular expansion operation of ; S008: Current radius r i Change to pore size set R p The next element in ; S009: Determine the pore size set R p Whether the elements in have been traversed, if not, re-execute steps S005-S009; if completed, a multi-level size pore space can be generated; Run the fluid segmentation module F, F001: Perform a difference operation between the image to be processed and the saturated image, and use the maximum inter-class variance method to determine the adaptive segmentation threshold in the difference image to segment the pore activation area; Considering that the original fluid has fuzzy areas, it will cause the pore utilization area segmentation obtained in step F001 to be insufficient, therefore, the original fluid is segmented into color images; F002: marking the fuzzy area between the primary fluid, i.e. the saturated fluid at the initial moment, and the displacement fluid. The specific implementation can be performed by using the roipoly function in Matlab; F003: Calculate the average RGB vector in the marked area And the covariance matrix C, the covariance satisfies: in, n is the number of pixels in the marked area, R j , G j and B j are the values ​​of the red, green and blue spaces in the image respectively; F004: Use the following criteria for color image segmentation: Among them, T h is the segmentation threshold, z is the RGB vector of the pixel to be evaluated; F005: performing a union operation on the fuzzy area obtained by color image segmentation and the pore utilization area obtained in step F001, and then performing an intersection operation with the pore area segmented in step P003 to obtain the segmented area of ​​the primary fluid; Run the pore utilization efficiency characterization module E, E001: Traverse the multi-level pore spaces formed in step S009 one by one according to the pore sizes; E002: Positioning radius is r i The pore space is measured by measuring the area occupied by the pore space. E003: Positioning radius is r i The primary fluid storage area in the pore space is measured E004: Calculate the radius r i The pore utilization efficiency E i , specifically calculated as E005: The current radius ri is changed to the pore size set R p The next element in E006: Determine the pore size set R p Whether the elements in have been traversed, if not, re-execute steps E001-E005; if completed, the utilization efficiency in the multi-level pores can be obtained; E007: The pore size shown in step E006 is the pore top view radius R t , the depth information of the microfluidic chip needs to be considered, and the top-down radius R t Convert to hydraulic radius R hyd , the conversion formula is as follows: Where A is the cross-sectional area of ​​the pore, I p is the polar moment of inertia; The cross section of the pores in dry-etched chips is usually rectangular. Considering the etching depth D p The hydraulic radius of the rectangular section is calculated as follows: Using the above formula, the top-view pore radius calculated in step E006 is converted into a hydraulic radius, and then the utilization efficiency of multi-level hydraulic radius pores is obtained.

Images