Method for determining size of three-dimensional digital core modeling of tight sandstone based on rock physics experiment constraint

Abstract

The present application relates to a kind of based on the determination method of the size of three-dimensional digital core modeling of tight sandstone of rock physical experiment constraint, belong to three-dimensional digital core modeling technical field.The determination method of three-dimensional digital core modeling size provided by the present application is constrained by rock physical experiment, can ensure that modeling size can both identify intergranular pore, and highly consistent with rock physical parameter;And the representativeness of modeling size is made accurate evaluation by comparing the porosity of three-dimensional digital core, mineral component content and the result of conventional rock physical experiment, provides strong basis for the construction of tight sandstone digital model.

Description

Technical Field

[0001] This invention relates to a method for determining the dimensions of a three-dimensional digital core model of tight sandstone based on rock physics experimental constraints, belonging to the field of three-dimensional digital core modeling technology. Background Technology

[0002] As unconventional reservoirs such as tight sandstone become the focus of oil and gas exploration and development, their characteristics—such as tight lithology, low porosity and permeability, strong heterogeneity, and high clay content—lead to the following problems with conventional rock physics experiments: sample saturation and displacement are difficult, experimental operations are complex, and it is difficult to capture microscopic pore structure characteristics, resulting in insufficient characterization accuracy. Moreover, conventional rock physics experiments cannot quantitatively measure and control pore structure and fluid distribution; therefore, it is difficult to accurately study the influence of various microscopic factors on macroscopic physical properties of reservoirs using only conventional rock physics experiments.

[0003] With the development of related disciplines, numerical simulation of rock physics has become one of the important methods in rock physics research, helping to reveal the variation laws of rock physical properties. Based on three-dimensional digital cores, numerical simulation algorithms can be used to calculate the acoustic, electrical, nuclear magnetic resonance responses, and seepage characteristics of rocks. Compared with conventional rock physics experiments, digital core physics experiments are faster and less expensive, and can simulate different rock physical properties based on the same three-dimensional digital core, making it easier to analyze the correlation between different physical properties. However, the accuracy and applicability of numerical simulations largely depend on the established rock microscopic model. Only when the porosity of the model can reflect the pore structure characteristics of the real rock sample will the simulation results have theoretical and applied value. The accuracy and representativeness of digital core modeling are the foundation of numerical simulation of rock response characteristics, while the scanning resolution and scanning size of various imaging methods are mutually restrictive. The larger the size of the three-dimensional digital core (i.e., the number of pixels it contains), the more accurate the simulation results of the macroscopic physical properties of the rock. However, due to the limited storage capacity and processing speed of computers, the size of the three-dimensional digital core cannot be too large. Moreover, the X-CT scanning resolution depends on the size of the scanned sample, and the resolution determines the ability to distinguish the pores and mineral microstructures of the rock.

[0004] For core samples composed of intergranular pores and mineral components, the connectivity and porosity of intergranular pores are core parameters for evaluating reservoir permeability. Revealing micro-mineral structures and micropores using imaging techniques at different scales is of great research significance. Although current three-dimensional digital core technology can construct a visual model of pore space and mineral components through imaging techniques, it still has the following limitations: (1) There is a lack of unified standards for the selection of sample size and resolution, which leads to deviations between modeling results and experimental data. Scanning resolution and sample size are mutually restrictive. For example, if the sample size is large and the resolution is insufficient, micropores cannot be identified; if the resolution is too high and the sample size is small, the amount of data is huge and it is difficult to represent the macroscopic properties of the sample; (2) Existing research focuses on multi-scale and multi-component modeling, and there is still a lack of methods for selecting the optimal size of a single core sample. Summary of the Invention

[0005] To address the problems existing in the prior art, this invention provides a method for determining the size of a three-dimensional digital core model of dense sandstone based on rock physics experimental constraints. Through comparative analysis of conventional rock physics experiments and digital modeling, it ensures that the modeling size can both identify intergranular porosity and highly match the rock physics parameters.

[0006] The present invention achieves the above objectives by adopting the following technical solutions:

[0007] This invention provides a method for determining the dimensions of a three-dimensional digital core model of tight sandstone based on rock physics experimental constraints, comprising the following steps:

[0008] S1. Obtain core samples and obtain experimental porosity, mineral composition content, and nuclear magnetic resonance T2 spectrum of the core samples in saturated water state through rock physics experiments.

[0009] S2. Drill multiple core samples of different diameters from the core sample;

[0010] S3. Perform X-ray CT scans on each core sample to obtain X-CT three-dimensional grayscale images of each core sample at their respective scanning resolutions.

[0011] S4. Based on the mineral composition characteristics of dense sandstone, the required component categories for segmentation are established, including five components: iron ore minerals, potassium feldspar and calcite, quartz and plagioclase, clay, and pores. Machine learning image segmentation algorithms are used to divide the X-CT three-dimensional grayscale images of each core sample into components, and multi-mineral component three-dimensional digital cores of each core sample are constructed.

[0012] S5. Calculate the CT-identified porosity Φ of the three-dimensional digital core for each core sample. CT and CT mineral component content V ii represents iron ore minerals, potassium feldspar and calcite, quartz and plagioclase, clay, and pores, respectively.

[0013] S6. Use experimental porosity to calibrate the nuclear magnetic resonance T2 spectrum of saturated water so that the sum of the integrated signal intensity after calibration is equal to the experimental porosity. The modeling size of the X-ray CT scan should be the size that can accurately identify the interparticle pores in the sample. In the calibrated T2 spectrum, the right peak is the larger pore, that is, the peak corresponding to the interparticle pores. The pore corresponding to the left trough position adjacent to the peak is the smallest pore diameter of the interparticle pores.

[0014] S7. On the T2 spectrum, accumulate the signal intensity after calibration from right to left, and mark the accumulated signal intensity corresponding to the trough position as Φ. v ;

[0015] Porosity Φ was identified using CT scans of each core sample. CT To constrain the process, the signal strength after the scale is accumulated from right to left. When the accumulated signal strength Φ SUM equal to Φ CT Stop accumulating and calculate. When P is less than a set value, it is determined that the diameter and scanning resolution of the corresponding core sample meet the modeling requirements.

[0016] Specifically, in step S2, the diameter of the core sample is 0.8mm-10mm, and the diameter of each core sample is evenly distributed between the maximum and minimum diameters.

[0017] Typically, in step S1, the experimental porosity is measured using the helium displacement method, and the mineral component content is measured using the XRD method.

[0018] Specifically, in step S5, the CT porosity is equal to the ratio of the number of voxels classified as pores to the total number of voxels in the three-dimensional digital core; the CT mineral component content is equal to the ratio of the number of voxels classified as the corresponding mineral component to the total number of voxels.

[0019] Furthermore, the method for determining modeling dimensions provided by the present invention also includes optimizing the selection of modeling dimensions based on their representativeness. The method for evaluating the representativeness of modeling dimensions includes the following steps:

[0020] P1. Obtain the SEM image of the core sample end face from step S1. Based on the mineral composition characteristics of the dense sandstone, establish the component categories to be segmented, including five components: iron ore minerals, potassium feldspar and calcite, quartz and plagioclase, clay, and pores. Use a machine learning image segmentation algorithm to segment the SEM image into components, identify the micropores within each mineral component, and calculate the proportion R of intra-group micropores in each mineral component. i ;

[0021] P2, The proportion of microporosity R within the group obtained in step P1. i The corresponding CT mineral component content V obtained in step S5 i The product of these is denoted as the microporosity Φ in the 3D digital core. i The microporosity Φ i The CT recognition porosity Φ obtained in step S5 CT The sum of these values ​​is used as the total porosity Φ in the model. 总 .

[0022] P3. Calculate the relative error between the total porosity of the model obtained in step P2 and the experimental porosity obtained in step S1. If the relative error is less than the specified value, the determined modeling size is considered to meet the representativeness requirement of porosity.

[0023] Furthermore, the representativeness assessment method also includes:

[0024] P4. While meeting the representativeness requirement in step P3, when the X-CT scan image shows that iron ore minerals are distributed in a dispersed manner, select the size corresponding to the core sample with the larger diameter as the modeling size.

[0025] Specifically, in step P1, the SEM image is acquired using a large field-of-view, high-resolution SEM image automatic acquisition system with a scanning resolution of 10-100 nm; the X-ray CT scan has a scanning resolution of 1 μm-10 μm. Here, 1 μm is the limit resolution of most current instruments.

[0026] Furthermore, in step P1, the proportion of microporosity R within the group i The calculation method includes the following steps:

[0027] From each mineral component obtained by segmentation, N rectangular regions are randomly selected. The ratio of the number of microporous voxels in each rectangular region to the total number of voxels in that rectangular region is taken as the microporous proportion R within the group. i The proportion of micropores within a group of N rectangular regions, R i The average value is used as the result.

[0028] The beneficial effects of this application include, but are not limited to:

[0029] The method for determining the modeling size of a three-dimensional digital core of tight sandstone based on rock physics experimental constraints provided by this invention can ensure that the modeling size can both identify intergranular porosity and closely match rock physics parameters. Moreover, by comparing the porosity and mineral composition content of the three-dimensional digital core with the results of conventional rock physics experiments, the representativeness of the modeling size is accurately evaluated, providing a strong basis for the construction of digital models of tight sandstone. Attached Figure Description

[0030] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0031] Figure 1 The T2 NMR spectrum is after calibration.

[0032] Figure 2 This represents the overall field of view of the MAPS image.

[0033] Figure 3 A schematic diagram for obtaining X-CT scan core samples of different diameters.

[0034] Figure 4 Image segmentation results of a 1mm diameter core sample and a schematic diagram of a three-dimensional digital core with multiple mineral components. Detailed Implementation

[0035] The present invention will be further described in detail below. However, it should be noted that the following specific embodiments are merely exemplary examples of the invention, and the scope of protection of the invention is not limited thereto. The scope of protection of the invention is defined only by the claims. It will be apparent to those skilled in the art that various other modifications and substitutions can be made to the embodiments of the invention within the scope of protection defined by the claims, and the same technical effects can still be achieved, thus achieving the ultimate technical objective of the invention.

[0036] The following will provide a detailed description of the method for determining the size and representativeness evaluation of three-dimensional digital core modeling of tight sandstone based on rock physics experimental constraints provided by the present invention, using specific implementation methods.

[0037] 1. Methods for determining modeling dimensions

[0038] S1. Obtain a standard plunger core sample with a diameter of 25.4 mm and conduct routine rock physics experiments on the core sample;

[0039] Specifically, the porosity of the core sample was measured using the helium displacement method, and the experimental porosity was found to be 16.7%. Then, samples of a certain size were taken from the core sample, and the content of each mineral component was obtained using an XRD diffractometer, typically as a volume fraction. The results are shown in Table 1. It should be noted that the pores in the sample are destroyed during the crushing and grinding process during XRD analysis; therefore, the XRD experimental results do not include porosity information. The volume fractions of mineral components in the table are corrected results obtained using a helium porosimeter.

[0040]

[0041] Nuclear magnetic resonance (NMR) spectroscopy was used to measure the attenuation curve of magnetic enhancement intensity in core samples in saturated water, obtaining the NMR T2 spectrum of the core samples in saturated water. The NMR T2 spectrum in saturated water exhibits a bimodal distribution, indicating that the sample is composed of pores of two different pore sizes. The larger the pore size distribution and the larger the pore size in the sample, the longer the transverse relaxation time T2.

[0042] S2. Drill multiple cylindrical core samples of different diameters from the core sample, such as... Figure 2 As shown, the diameters of the various core samples are 9 mm, 7 mm, 5 mm, 3 mm, and 1 mm, respectively.

[0043] S3. Perform X-ray CT scans on each core sample to obtain X-CT three-dimensional grayscale images of each core sample at their respective scanning resolutions; specifically, the X-CT scanning resolutions corresponding to core samples with diameters of 9mm, 7mm, 5mm, 3mm, and 1mm are 5.85μm, 4.66μm, 3.19μm, 2.35μm, and 1.1μm, respectively.

[0044] S4. Potassium feldspar and calcite have similar densities, as do quartz and plagioclase, resulting in similar grayscale values ​​in X-CT images. Therefore, as... Figure 3 As shown, based on the mineral composition characteristics of dense sandstone, the required component categories for segmentation are established, including five components: iron ore minerals, potassium feldspar and calcite, quartz and plagioclase, clay, and pores. Machine learning image segmentation algorithms are used to divide the X-CT three-dimensional grayscale images of each core sample into components, and multi-mineral component three-dimensional digital cores of each core sample are constructed.

[0045] S5. Calculate the CT-identified porosity Φ of the three-dimensional digital core for each core sample. CT and CT mineral component content V i , i represents iron ore minerals, potassium feldspar and calcite, quartz and plagioclase, clay, and porosity, respectively. The results are shown in Table 2. Among them, CT-identified porosity Φ CT Equals the ratio of the number of voxels classified as pores to the total number of voxels in a 3D digital core; CT mineral component content V i It is equal to the ratio of the number of voxels that are divided into the corresponding mineral components to the total number of voxels.

[0046]

[0047] As shown in Table 2, there are differences in CT-identified porosity and CT-identified mineral component content among core samples of different diameters. The volumetric mineral component content identified by CT is higher than that obtained by XRD testing. This is because CT cannot identify micropores developed between mineral crystals or within grains, and instead segments these micropores into their corresponding minerals. The difference in clay content between CT and XRD testing is mainly due to the unique layered structure and large specific surface area of ​​clay, resulting in the most developed micropores in the clay component.

[0048] S6. The nuclear magnetic resonance T2 spectrum in saturated water was calibrated using experimental porosity, so that the sum of the integrals of the calibrated signal intensity equaled the experimental porosity. The calibrated T2 spectrum is as follows: Figure 1 As shown, Figure 1 The area of ​​the closed region formed by the T2 spectrum distribution curve and the x-axis is the porosity.

[0049] The modeling size for X-ray CT scans should be the size that allows for accurate identification of intergranular pores in the sample. In the calibrated T2 spectrum, the peak on the right represents larger pores, i.e., the peak corresponding to intergranular pores. The pore corresponding to the trough on the left side of this peak at T2=5.2ms is the smallest pore size among the intergranular pores.

[0050] S7. On the T2 spectrum, accumulate the signal intensity after calibration from right to left, and mark the accumulated signal intensity corresponding to the trough position as Φ. v ;

[0051] Porosity Φ was identified using CT scans of each core sample. CT To constrain the process, the signal strength after the scale is accumulated from right to left. When the accumulated signal strength Φ SUM equal to Φ CT Stop accumulating and calculate. When P is less than a set value, the corresponding core sample diameter and scanning resolution are determined to meet the modeling requirements.

[0052] In this embodiment, based on the relationship between the porosity identified by CT and the nuclear magnetic resonance T2 spectrum of each core sample, the core samples with diameters of 1 mm and 3 mm that meet the modeling requirements are located near the trough T2=5.2 ms.

[0053] Furthermore, the modeling size determination method provided by the present invention also includes a step of optimizing the selection of modeling sizes based on their representativeness.

[0054] 2. Representativeness assessment method

[0055] P1. Before drilling core samples, the core samples from step S1 are cold-mounted, and the end faces are mechanically polished. A carbon layer is then sprayed onto the end faces to improve their conductivity. SEM images of the end faces are acquired using a large field-of-view, high-resolution SEM image acquisition system (MAPS). Figure 4 As shown.

[0056] Based on the mineral composition characteristics of dense sandstone, five component categories were established for segmentation: iron ore minerals, potassium feldspar and calcite, quartz and plagioclase, clay, and pores. Machine learning image segmentation algorithms were used to segment the SEM image into these components, identifying intra-group micropores within each mineral component. The proportion (R) of intra-group micropores within each mineral component was calculated. i Then use R i Correct the porosity and mineral composition content of three-dimensional digital cores.

[0057] The proportion of microporosity within the group R i The calculation method includes the following steps:

[0058] Ten rectangular regions were randomly selected from each of the quartz and plagioclase, potassium feldspar and calcite, and clay components in the SEM images. Micropores in each region were extracted using a machine learning image segmentation algorithm. The ratio of the number of micropore voxels in a rectangular region to the total number of voxels in that region was taken as the micropore percentage R within the group. i The proportion of micropores within a group of N rectangular regions, R i The average value was used as the result. The intra-group microporosity R of quartz and plagioclase, potassium feldspar and calcite, and clay minerals in the core sample was calculated. i They were 4.94%, 8.56%, and 29.98%, respectively.

[0059] P2, The proportion of microporosity within the group obtained in step P1, R i The corresponding CT mineral component content V obtained in step S5 i The product of these is denoted as the microporosity Φ in the 3D digital core. i The results are shown in Table 3.

[0060]

[0061] Iron ore minerals have low volume content and very low micropore content in their mineral components, therefore micropores in iron ore minerals are ignored in Table 3.

[0062] The microporosity Φ of each part of the three-dimensional digital rock core i The CT recognition porosity Φ obtained in step S5 CT The sum of these values ​​is used as the total porosity Φ in the model. 总 .

[0063] The CT-identified porosity, total porosity of the model, and experimental porosity of each core sample are shown in Table 4.

[0064]

[0065] P3. Calculate the relative error between the total porosity of the model obtained in step P2 and the experimental porosity obtained in step S1. According to the standard for estimating oil and gas reserves (DZ / T 0217-2020), if the relative error between the two porosities is less than 8%, it is considered to meet the representativeness requirement of porosity.

[0066] Furthermore, the core skeleton is mainly composed of felsic minerals, with a relatively low content of iron ore minerals. However, iron ore minerals have a high density and strong absorption of X-rays, resulting in the highest brightness in X-CT scan images and high identification accuracy. Both X-CT and scanning electron microscopy test images show that the iron ore minerals in this sample are dispersed, with large individual clusters and low spatial distribution density. Therefore, their volumetric content can be used to assist in evaluating the representativeness of the three-dimensional digital core. Comparing the CT mineral composition content of core samples of different diameters in Table 2 with the XRD content of iron ore minerals in Table 1, it can be seen that the closest core sample diameter is 3 mm. For this type of sample, while ensuring the ability to identify pores, the core sample size should be appropriately increased. The increased modeling size should meet the above modeling requirements and representativeness requirements.

[0067] In this invention, the Trainable Weka Segmentation plugin from the open-source software FIJI is used for image segmentation.

[0068] The above specific embodiments should not be construed as limiting the scope of protection of the present invention. For those skilled in the art, any alternative improvements or modifications made to the embodiments of the present invention shall fall within the scope of protection of the present invention.

[0069] Any aspects of this invention not described in detail are well-known to those skilled in the art.

Claims

1. A method for determining the dimensions of a three-dimensional digital core model of tight sandstone based on rock physics experimental constraints, characterized in that, Includes the following steps: S1. Obtain core samples and obtain experimental porosity, mineral composition content, and nuclear magnetic resonance T2 spectrum of the core samples in saturated water state through rock physics experiments. S2. Drill multiple core samples of different diameters from the core sample; S3. Perform X-ray CT scans on each core sample to obtain X-CT three-dimensional grayscale images of each core sample at their respective scanning resolutions. S4. Based on the mineral composition characteristics of dense sandstone, the required component categories for segmentation are established, including five components: iron ore minerals, potassium feldspar and calcite, quartz and plagioclase, clay, and pores. Machine learning image segmentation algorithms are used to divide the X-CT three-dimensional grayscale images of each core sample into components, and multi-mineral component three-dimensional digital cores of each core sample are constructed. S5. Calculate the CT-identified porosity Φ of the three-dimensional digital core for each core sample. CT and CT mineral component content V i i represents iron ore minerals, potassium feldspar and calcite, quartz and plagioclase, clay, and pores, respectively. S6. Use experimental porosity to calibrate the nuclear magnetic resonance T2 spectrum of saturated water so that the sum of the integrated signal intensity after calibration is equal to the experimental porosity. The modeling size of the X-ray CT scan should be the size that can accurately identify the interparticle pores in the sample. In the calibrated T2 spectrum, the right peak is the larger pore, that is, the peak corresponding to the interparticle pores. The pore corresponding to the left trough position adjacent to the peak is the smallest pore diameter of the interparticle pores. S7. On the T2 spectrum, accumulate the signal intensity after calibration from right to left, and mark the accumulated signal intensity corresponding to the trough position as Φ. v ; Porosity Φ was identified using CT scans of each core sample. CT To constrain the process, the signal strength after the scale is accumulated from right to left. When the accumulated signal strength Φ SUM equal to Φ CT Stop accumulating and calculate. When P is less than the set value, it is determined that the diameter and scanning resolution of the corresponding core sample meet the modeling requirements, and the diameter and scanning resolution of the core sample are the modeling dimensions. The determination method further includes optimizing the selection of modeling dimensions based on their representativeness. The method for evaluating the representativeness of the modeling dimensions includes the following steps: P1. Obtain the SEM image of the core sample end face from step S1. Based on the mineral composition characteristics of the dense sandstone, establish the component categories to be segmented, including five components: iron ore minerals, potassium feldspar and calcite, quartz and plagioclase, clay, and pores. Use a machine learning image segmentation algorithm to segment the SEM image into components, identify the micropores within each mineral component, and calculate the proportion R of intra-group micropores in each mineral component. i ; P2, The proportion of microporosity R within the group obtained in step P1. i The corresponding CT mineral component content V obtained in step S5 i The product of these is denoted as the microporosity Φ in the 3D digital core. i The microporosity Φ i The CT recognition porosity Φ obtained in step S5 CT The sum of these values ​​is used as the total porosity Φ in the model. 总 ; P3. Calculate the relative error between the total porosity of the model obtained in step P2 and the experimental porosity obtained in step S1. If the relative error is less than the specified value, the determined modeling size is considered to meet the representativeness requirement of porosity.

2. The determination method according to claim 1, characterized in that, In step S1, the experimental porosity is measured by helium displacement method, and the mineral component content is measured by XRD method.

3. The determination method according to claim 1, characterized in that, In step S2, the diameter of the core sample is 0.8mm-10mm, and the diameter of each core sample is evenly distributed between the maximum and minimum diameters.

4. The determination method according to claim 1, characterized in that, In step S5, the CT porosity is equal to the ratio of the number of voxels classified as pores to the total number of voxels in the three-dimensional digital core; the CT mineral component content is equal to the ratio of the number of voxels classified as the corresponding mineral component to the total number of voxels.

5. The determination method according to claim 1, characterized in that, Also includes: P4. While meeting the representativeness requirement in step P3, when the X-CT scan image shows that iron ore minerals are distributed in a dispersed manner, select the size corresponding to the core sample with the larger diameter as the modeling size.

6. The determination method according to claim 1, characterized in that, In step P1, the SEM image is acquired using a large field-of-view, high-resolution SEM image automatic acquisition system with a scanning resolution of 10-100 nm; the X-ray CT scan has a scanning resolution of 1 μm-10 μm.

7. The determination method according to claim 1, characterized in that, In step P1, the proportion of microporosity within the group is R. i The calculation method includes the following steps: From each mineral component obtained by segmentation, N rectangular regions are randomly selected. The ratio of the number of microporous voxels in each rectangular region to the total number of voxels in that rectangular region is taken as the microporous proportion R within the group. i The proportion of micropores within a group of N rectangular regions, R i The average value is used as the result.

Images