Characterization method for glutenite distribution in glutenite tight reservoir

Abstract

The invention discloses a characterization method for glutenite distribution in a glutenite tight reservoir, which comprises the following steps of: (1) acquiring an actual sample, collecting gravel distribution data of the glutenite reservoir, analyzing a picture by using a neural network algorithm, and identifying to obtain the distribution of the gravel content, the gravel size and the gravel position in the sample; (2) obtaining gravel particle size and content parameters by adopting a particle size grading curve, constructing a rock model, and importing gravel particle distribution characteristics; and (3) setting the part except the gravel particles as a matrix, and endowing the matrix and the gravel particles with different physical property parameters and rock mechanical parameters. According to the method, glutenite distribution pattern characterization in various forms can be realized, and glutenite simulation prediction under various stratum conditions is realized.

Description

Technical Field

[0001] The invention belongs to the technical field of hydraulic fracturing of conglomerate oil and gas reservoirs, and in particular relates to a method for characterizing the distribution of conglomerate in a conglomerate tight reservoir. Background Art

[0002] Gravel reservoirs are widely distributed in my country. Compared with conventional oil and gas reservoirs, gravel reservoirs contain gravel components. Due to factors such as the size of gravel, the non-uniformity of gravel content, and the strength of gravel, gravel reservoirs generally have poor porosity and permeability characteristics and serious heterogeneity, so the natural production capacity is generally not high. In order to meet the requirements of industrial production, hydraulic fracturing is needed to increase production. During the hydraulic fracturing process, the distribution of fracture morphology often determines the effect of fracturing. In gravel reservoirs, due to the presence of gravel and the extremely irregular arrangement of gravel, the fractures may be affected by the size of gravel, the content of gravel, and the strength of gravel during the extension process, resulting in a complex fracture network structure, thereby affecting the fracturing effect, and at the same time, higher requirements are put forward for the embedding of proppants. If the proppant is severely embedded, the conductivity of the reservoir will be weakened, making the overall fracturing effect worse. In order to solve this problem, it is necessary to establish a characterization method for the distribution of gravel in gravel tight reservoirs, accurately simulate the distribution law of gravel fractures, and then guide the fracturing design of gravel reservoirs. Summary of the invention

[0003] The present invention is proposed to fully simulate the extremely irregular arrangement of gravel, reflect the final distribution morphology of bedrock and conglomerate in conglomerate, and provide guidance for the fracturing design of conglomerate reservoirs. Its purpose is to provide a characterization method for the distribution of conglomerate in conglomerate tight reservoirs.

[0004] The present invention is achieved through the following technical solutions:

[0005] A method for characterizing the distribution of conglomerate in a conglomerate tight reservoir comprises the following steps:

[0006] (I) collecting gravel distribution data of the sandy conglomerate reservoir in actual samples; the gravel distribution data of the sandy conglomerate reservoir includes gravel content, gravel size and gravel location distribution;

[0007] (II) substituting the gravel distribution data of the conglomerate reservoir obtained in step (I) into the particle size gradation curve to generate the gravel distribution data of the conglomerate reservoir after modeling;

[0008] The particle size gradation curve is used to generate the distribution of gravel particles and describe the distribution of gravel particles in different particle size ranges. The particle size gradation curve can be used to generate the proportion and distribution of particles of different particle sizes, which can reflect the size ratio, content distribution and other parameters of gravel particles. For example, the Fuller gradation curve will distribute gravels of different sizes in a certain proportion to the simulation area according to the formula, thereby simulating the real distribution state of gravel particles in the reservoir. The particle size gradation curve is an important tool for simulating the particle distribution characteristics in the real reservoir.

[0009] The data of step (I) are the gravel content, size and position directly measured from the actual sample, which is a kind of original observation data and may contain local deviation or incompleteness. Step (II) is to substitute the directly measured data of step (I) into the formula of the particle size distribution curve (such as the normal distribution formula, the Fuller grading formula, etc.) to calculate the proportion and distribution of gravels of different particle sizes. The data obtained at this time is the gravel distribution data of the sandstone reservoir after modeling. This data is generated by processing the data of step (I) and substituting it into the mathematical model. The gravel particle size and content parameters are more standardized and regularized, and can reflect the gravel distribution of the entire reservoir when used for simulation. These parameters are usually fitted or corrected to make them more in line with the requirements of numerical simulation.

[0010] (III) Establishing a two-dimensional or three-dimensional numerical simulation model according to simulation requirements;

[0011] The boundary is divided according to the size of the established two-dimensional or three-dimensional numerical simulation model, and the boundary is the wall; a circular wall is constructed in the center of the area to simulate the wellbore as needed, and the gravel particle distribution generated by the particle size grading curve is introduced around the area; finally, a rock simulation model with different gravel particle sizes including the wellbore is obtained;

[0012] (IV) The part other than the gravel particles in the rock simulation model in step (III) is set as the matrix, and the matrix and gravel particles are given different physical properties and rock mechanical parameters to achieve the reproduction of the data from the actual sample to the rock simulation model data, and finally the rock mechanical analysis is carried out according to the simulation steps.

[0013] 2. The method for characterizing the distribution of conglomerate in a conglomerate tight reservoir according to claim 1, characterized in that: the specific method of collecting gravel distribution data of the conglomerate reservoir in the actual sample in step (I) comprises the following steps:

[0014] (i) Drilling samples

[0015] Select a location on the actual sample where the gravel particles have uniform size to drill a standard columnar rock sample;

[0016] (ii) Image Recognition

[0017] The standard columnar rock sample is separated from the center along the axial direction to obtain a rock sample section, a high-definition camera is used to take a picture of the rock sample section, and a neural network image recognition algorithm is used to analyze the picture to obtain gravel distribution data of the sandy conglomerate reservoir, wherein the gravel distribution data of the sandy conglomerate reservoir includes the size of gravel particles, the distribution of gravel particle sizes, and the distribution of gravel positions;

[0018] (iii) Manual Identification

[0019] Manually identify and color the gravel particles in the rock sample section, outline the size of the gravel particles, and obtain the size of the gravel particles, the distribution of gravel particle size, and the distribution of gravel locations;

[0020] (iv) Compare the gravel distribution data of the conglomerate reservoir obtained by image recognition in step (ii) with the gravel distribution data of the conglomerate reservoir obtained by manual identification in step (iii). If the error between the two is less than 10%, the image recognition result is considered to be credible, and the result of the image recognition in step (ii) is used as the basis for calculating the particle size distribution curve in step (ii). If the error is greater than 10%, adjust the training parameters of the neural network image recognition algorithm until the error is less than 10%.

[0021] In the above technical solution, the actual sample is a full-diameter core sample with a core diameter of 12 cm; the standard columnar rock sample has a diameter of 2.5 cm and a length of 3 cm to 5 cm; the drilling of the standard columnar rock sample is carried out in accordance with standard GB / T29172-2012.

[0022] In the above technical solution, the simulation requirements are the conditions and constraints that need to be met when performing numerical simulation modeling to ensure that the model can accurately simulate the distribution and mechanical behavior of gravel in the sandy conglomerate reservoir. The simulation requirements include:

[0023] (i) Design requirements for geometric structures: Based on experimental data and rock characteristics, a two-dimensional or three-dimensional numerical simulation model is established. The model's boundary, wellbore and other structures must be accurately constructed according to simulation requirements;

[0024] (ii) Requirements for the generation of particle size distribution curves: Generate particle size distribution curves based on the distribution data of gravel to ensure that the size, distribution and shape of gravel in the model are consistent with the actual situation and meet the set accuracy requirements;

[0025] (iii) Material property assignment requirements: During simulation, different physical properties (such as permeability and porosity) and mechanical parameters (such as Young's modulus and compressive strength) need to be assigned to the matrix and gravel particles to ensure that the simulation results are consistent with the physical performance of the actual reservoir;

[0026] (iv) Accuracy and error control requirements: The accuracy requirement of the simulation is that the error between the gravel particle distribution in the model and the actual sample should be less than 10%; if the error is greater than this value, the simulation parameters need to be adjusted to ensure the credibility of the results;

[0027] (v) Design requirements for the simulation environment: Set loading conditions (such as pressure and stress) according to the simulation objectives and divide the mesh to ensure the accuracy and effectiveness of the mechanical analysis.

[0028] In the above technical solution, the two-dimensional or three-dimensional numerical simulation model refers to a computational model that simulates the distribution, mechanical properties and other phenomena of gravel in a sandstone reservoir based on actual physical and geometric parameters through mathematics and computer technology. The model is used to analyze rock mechanical behavior, crack expansion and other processes to provide a basis for designing fracturing schemes; the two-dimensional numerical simulation model only considers two dimensions (usually X and Y axes) and is mainly used to simplify complex three-dimensional problems. The two-dimensional model is suitable for analysis on cross sections or planes and is usually used for some approximate analysis or when the three-dimensional effect is small; the three-dimensional numerical simulation model considers all three dimensions (X, Y, Z axes) and can more comprehensively simulate three-dimensional phenomena in the real world, especially when studying complex geological structures (such as reservoirs, crack expansion, etc.).

[0029] In the above technical solution, the method for establishing a two-dimensional or three-dimensional numerical simulation model comprises the following steps:

[0030] (i) Data collection:

[0031] Gravel distribution, particle size, content and location data are obtained from actual samples; this data can be obtained through high-definition image recognition or manual annotation and analyzed using a neural network algorithm.

[0032] (ii) Define the geometry:

[0033] Construct the reservoir geometry from the data: in a 2D model this is usually a plan view; in a 3D model this is a 3D view of the entire reservoir;

[0034] (iii) Meshing:

[0035] Divide the geometry into many small computational cells or grids;

[0036] This step is very important because it determines the accuracy of the simulation. Finer grids can bring higher accuracy, but the amount of calculation will also increase accordingly. The 2D model divides the grid on the plane, while the 3D model divides the grid in space.

[0037] In the above technical solution, the area around the area is the scope of the study area defined in the simulation. The size and shape of this range will directly affect the accuracy and calculation cost of the simulation. When defining the area around the area, it is usually necessary to consider the scale of the reservoir, the size of the wellbore, the range of fracture extension, and the setting of boundary conditions, which can be set to 1m*1m, 2m*2m, etc.

[0038] In the above technical solution, the particle size gradation curve includes a normal distribution gradation; the normal distribution gradation is calculated using the following formula:

[0039]

[0040] Where, μ is the diameter of the gravel particles, in mm; σ is the deviation factor, without unit;

[0041] x is the gravel distribution data of the sandy conglomerate reservoir;

[0042] f(x) is the distribution function of gravel.

[0043] In the above technical solution, the types of aggregates of the gravel particles include two-dimensional regular polygonal aggregates, random polygonal aggregates, two-dimensional circular aggregates, two-dimensional elliptical aggregates and three-dimensional polyhedral aggregates.

[0044] In the above technical solution, when the aggregate type of the particles is a two-dimensional polygonal aggregate or a three-dimensional polyhedral aggregate, the number of minimum edges is ≥3, the maximum number of edges is ≥the minimum number of edges; the radius random coefficient is between 0 and 1; the angle random coefficient is between 0 and 1; and the shape coefficient is ≥1.2.

[0045] In the above technical solution, the particle size distribution of the particles in step (II) is developed using any one of Fuller grading, Fuller grading + Walaven or Taybo grading.

[0046] In the above technical solution, when D max / d min When ≤5, the particle size distribution of the particles in step (II) is expanded using Fuller gradation, and the Fuller gradation expansion formula is formula (2):

[0047]

[0048] When 5<D max / d min When ≤10, the particle size distribution of the particles in step (II) is expanded by Fuller gradation + Walawen gradation, and the expansion formula of Fuller gradation + Walawen gradation is formula (3),

[0049] P(D<d i )=P k [1.065(d i / Dmax ) 0.5 -0.053(d i / D max ) 4 -0.012(d i / D max ) 6 -0.0045(d i / D max ) 8 -0.0025(d i / D max ) 10 ] Formula (3);

[0050] When D max / d min When >10, the particle size distribution of the particles in step (II) is expanded using the Taybo gradation, and the Taybo gradation expansion formula is formula (4):

[0051] P(D<d i )=P k (d i / D max ) n , n∈(0.3~0.7) Formula (4)

[0052] In formulas (2) to (4), D is the particle diameter in mm; d min D is the minimum particle diameter, in mm; max is the maximum particle diameter, in mm; P k For a particle size d i The passing percentage of the skeleton, in %; d i For the skeleton i The diameter of the pipe is in mm; n is the exponential coefficient, 0.3<n<0.7.

[0053] In the above technical solution, if the error between the particle size gradation parameters generated in step (II) and the gravel particle size and gravel content obtained in step (I) is less than 10%, it is considered that the simulation accuracy meets the requirements, otherwise the gradation parameters are further adjusted until they meet the requirements.

[0054] In the above technical solution, the calculation area size in step (II) is larger than the external wall size; the generated particle size should be small enough to ensure simulation accuracy; and there are enough particles around the wellbore to eliminate the influence of geometric heterogeneity.

[0055] The beneficial effects of the present invention are:

[0056] The present invention provides a method for characterizing the distribution of conglomerate in a conglomerate tight reservoir, which can realize the characterization of various forms of conglomerate distribution morphology and realize the simulation prediction of conglomerate under various formation conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 is a distribution diagram of sandstone based on two-dimensional normal distribution in the present invention;

[0058] Figure 2 It is the distribution map of the sandstone based on the two-dimensional circle in the present invention;

[0059] Figure 3 It is a distribution map of sandstone based on two-dimensional polygons in the present invention;

[0060] Figure 4 This is a photo of sandstone identified based on a neural network in Example 1 of the present invention.

[0061] For ordinary technicians in this field, other relevant drawings can be obtained based on the above drawings without any creative work. DETAILED DESCRIPTION

[0062] In order to enable those skilled in the art to better understand the technical solution of the present invention, the technical solution of the present invention is further described below with reference to the accompanying drawings and through specific implementation methods.

[0063] A method for characterizing the distribution of conglomerate in a conglomerate tight reservoir comprises the following steps:

[0064] (I) Obtaining actual samples to collect gravel distribution data of sandstone conglomerate reservoirs, taking pictures of different sections of actual samples using a high-definition camera, and using a neural network algorithm to analyze the pictures to identify the distribution of gravel content, gravel size and location in the samples. The specific steps are as follows:

[0065] (i) Drilling samples: According to the industry standard GB / T29172-2012, on the full-diameter (core diameter is 12 cm) core sample, select a position with relatively uniform particle size and drill a standard columnar rock sample with a diameter of 2.5 cm and a length of 3 cm to 5 cm;

[0066] (ii) Section splitting: the rock sample is split along the axis from the center, photographed with a high-definition camera, and uploaded to a computer for analysis using a neural network image recognition algorithm to obtain the distribution of gravel;

[0067] The specific steps of neural network image recognition algorithm analysis are:

[0068] ① Select the training group and select N data from the sample as the training group;

[0069] ② Initialize accuracy control parameters and learning rate;

[0070] ③ Select the calculation mode and determine the target vector;

[0071] ④Calculate the intermediate output vector;

[0072] ⑤Compare the error between the output vector and the target vector;

[0073] ⑥ Calculate the weight of each vector;

[0074] ⑦ Determine whether the indicator meets the accuracy requirements. If not, continue to iterate. If so, proceed to the next step;

[0075] ⑧Training is over;

[0076] (iii) manually distinguish and color different gravel particles, and outline the size, particle size distribution and location distribution of gravel particles;

[0077] (iv) Compare the image recognition result in (ii) with the manual recognition result in (iii). If the error is less than 10%, the recognition result is considered reliable; the computer recognition result is used as the basis for calculation.

[0078] If the error is greater than 10%, adjust the parameters and continue training until the error is less than 10%.

[0079] (Ⅱ)

[0080] A. Based on the calculation results in (I), the information on conglomerate particle size content, distribution, particle size shape, etc. is further improved in the computer to carry out virtual simulation, specifically:

[0081] ① Modeling: Establish an object structure of a certain geometric size (can be two-dimensional or three-dimensional) according to design requirements;

[0082] ②Add attributes: set the material properties, elastic modulus, deformation factor and other parameters of the object;

[0083] ③ Grid division: divide the collection structure into multiple grids;

[0084] ④ Create analysis steps: set the time step and divide the overall process into multiple time steps for calculation;

[0085] ⑤Apply load: apply load to the model;

[0086] ⑥ Calculation and analysis: After all pre-processing processes (①~⑤) are completed, simulation and calculation analysis are started to obtain the calculation results;

[0087] ⑦ Check the results: Check the stress, strain, pressure field and other parameters based on the data obtained by calculation and analysis;

[0088] B. Constructing a two-dimensional characterization method: according to the requirements of the simulation calculation, the calculation area is set, particles are generated in the wall according to the distribution of particle sizes and positions obtained in (I), and a circular wall is constructed in the center of the area to simulate the wellbore;

[0089] The particle size gradation curve includes a normal distribution gradation;

[0090] The normal distribution gradation is calculated using formula (1):

[0091]

[0092] Where: μ is the diameter of the particle, in mm; σ is the deviation factor, without unit;

[0093] x is a random variable, unit is mm;

[0094] f(x) is the probability density function and has no units.

[0095] In order to obtain more conglomerate particle sizes and shapes, the types of conglomerate aggregates include two-dimensional regular polygonal aggregates, random polygonal aggregates, two-dimensional circular aggregates, two-dimensional elliptical aggregates and three-dimensional polyhedral aggregates, and the ellipticity coefficient is increased;

[0096] When the aggregate type is a two-dimensional polygonal aggregate or a three-dimensional polyhedral aggregate, the minimum number of edges must be ≥ 3 to form an aggregate; the maximum number of edges must be ≥ the minimum number of edges; the radius random coefficient must be between 0 and 1; the angle random coefficient must be between 0 and 1; the shape coefficient must be ≥ 1.2;

[0097] When calculating the particle size distribution, any of the following three grading methods is used.

[0098] (1) When the particle diameter is relatively regular (Dmax / dmin≤5), the Fuller gradation is used, and the following formula (2) is used:

[0099]

[0100] P k : A particle size d i Percentage of skeleton passing, %;

[0101] d i : Skeleton d i Diameter, mm;

[0102] D max : Maximum particle diameter, mm;

[0103] In actual example, when Dmax = 10mm and dmin = 5mm, the graph is as shown in the attached figure. Figure 1 shown.

[0104] (2) When 5<Dmax / dmin≤10, when Fuller gradation + Walaven is used, the following formula (3) is used:

[0105] P(D<d i )=P k [1.065(d i / D max ) 0.5 -0.053(d i / D max ) 4 -0.012(d i / D max ) 6 -0.0045(d i / D max ) 8 -0.0025(d i / D max ) 10 Formula (3);

[0106] (3) When the particle diameter difference is large (Dmax / dmin>10) and Taibo grading is used, the following formula (4) is used:

[0107] P(D<d i )=P k (d i / D max ) n , n∈(0.3~0.7) Formula (4);

[0108] In formulas (2) to (4), dmin is the minimum particle diameter, in mm; D max is the maximum particle diameter, in mm; P k For a particle size d i The passing percentage of the skeleton, in %; d i For the skeleton i The diameter of the pipe is in mm; n is the exponential coefficient, 0.3<n<0.7.

[0109] In actual example, when Dmax = 10mm and dmin = 5mm, for a two-dimensional circular skeleton, the graph is as shown in the attached figure. Figure 2 As shown; for a two-dimensional polygonal skeleton, its graphics are as shown in the attached Figure 3 shown.

[0110] (III) The part other than the gravel particles in the rock simulation model in step (II) is set as the matrix, and the matrix and gravel particles are given different physical properties (permeability, porosity) and rock mechanics parameters (Young's modulus, Poisson's ratio, compressive strength, etc.), and finally rock mechanics analysis is carried out according to the simulation steps;

[0111] (IV) In step (II), the calculation area size should be larger than the external wall size; if the error between the generated particle size gradation parameters and the gravel particle size and gravel content obtained in (I) is less than 10%, the simulation accuracy is considered to meet the requirements. If the parameter error is greater than 10%, first adjust the percentage of large particle size (ratio to the minimum particle size>10) so that its accuracy meets the requirements, and then adjust the accuracy of small particles step by step until all particles meet the accuracy requirements.

[0112] Example 1

[0113] In order to analyze the distribution characteristics of the X-stratum sandstone in a certain area, three cores with a size of 300mm*300mm*300mm were taken from the ground. The six faces of each rock were photographed and analyzed, and a total of 18 rock faces were collected. The results of manual analysis are shown in Table 1, where the average content of large-grained gravel is 14.3%, the average content of medium-gravel is 47.3%, and the average content of small gravel is 38.4%. The results of the neural network recognition method are as follows: the average content of large-grained gravel is 15.7%, the average content of medium-gravel is 49.2%, and the average content of small gravel is 35.1%. The evaluation error is less than 10%, and it is believed that the neural network recognition method meets the accuracy requirements and can be used for calculation and analysis.

[0114] According to the identification results, the information of conglomerate particle size content, distribution, particle size shape, etc. is further improved for virtual simulation. The ratio of the diameter of the largest particle to the diameter of the smallest particle is greater than 10, so the Fuller gradation is used. According to the formula of the Fuller series, the proportion of each type of particle is calculated separately: the proportion of particles larger than 1cm is 14.64%, the proportion of particles smaller than 0.5cm is 35.35, and the proportion of particles 0.5-1cm is 50.01%.

[0115] Table 1 Statistics of artificially identified conglomerate data

[0116]

[0117] The applicant declares that the above is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Those skilled in the art should understand that any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present invention shall fall within the protection scope and disclosure scope of the present invention.

Claims

1. A method for characterizing the distribution of conglomerate in a conglomerate tight reservoir, characterized by: The following steps are involved: (I) collecting gravel distribution data of the sandy conglomerate reservoir in actual samples; the gravel distribution data of the sandy conglomerate reservoir includes gravel content, gravel size and gravel location distribution; (II) substituting the gravel distribution data of the conglomerate reservoir obtained in step (I) into the particle size gradation curve to generate the gravel distribution data of the conglomerate reservoir after modeling; (III) establishing a two-dimensional or three-dimensional numerical simulation model according to the simulation requirements; dividing the boundary according to the size of the established two-dimensional or three-dimensional numerical simulation model, where the boundary is the wall; constructing a circular wall to simulate the wellbore at the center of the area as required, and introducing the gravel particle distribution generated by the particle size grading curve around the area; finally obtaining a rock simulation model with different gravel particle sizes including the wellbore; (IV) The part other than the gravel particles in the rock simulation model in step (III) is set as the matrix, and the matrix and gravel particles are given different physical properties and rock mechanical parameters to achieve the reproduction of the data from the actual sample to the rock simulation model data, and finally the rock mechanical analysis is carried out according to the simulation steps.

2. The method for characterizing the distribution of conglomerate in a conglomerate tight reservoir according to claim 1, characterized in that: The specific method of collecting gravel distribution data of the sandstone reservoir in the actual sample in step (I) comprises the following steps: (i) Drilling samples Select a location on the actual sample where the gravel particles have uniform size to drill a standard columnar rock sample; (ii) Image Recognition The standard columnar rock sample is separated from the center along the axial direction to obtain a rock sample section, a high-definition camera is used to take a picture of the rock sample section, and a neural network image recognition algorithm is used to analyze the picture to obtain gravel distribution data of the sandy conglomerate reservoir, wherein the gravel distribution data of the sandy conglomerate reservoir includes the size of gravel particles, the distribution of gravel particle sizes, and the distribution of gravel positions; (iii) Manual Identification Manually identify and color the gravel particles in the rock sample section, outline the size of the gravel particles, and obtain the size of the gravel particles, the distribution of gravel particle size, and the distribution of gravel locations; (iv) Compare the gravel distribution data of the conglomerate reservoir obtained by image recognition in step (ii) with the gravel distribution data of the conglomerate reservoir obtained by manual identification in step (iii). If the error between the two is less than 10%, the image recognition result is considered to be credible, and the result of the image recognition in step (ii) is used as the basis for calculating the particle size distribution curve in step (ii). If the error is greater than 10%, adjust the training parameters of the neural network image recognition algorithm until the error is less than 10%.

3. The method for characterizing the distribution of conglomerate in a conglomerate tight reservoir according to claim 2, characterized in that: The actual sample is a full-diameter core sample with a core diameter of 12 cm; the standard columnar rock sample has a diameter of 2.5 cm and a length of 3 cm to 5 cm; the drilling of the standard columnar rock sample is carried out in accordance with the standard GB / T29172-2012.

4. The method for characterizing the distribution of conglomerate in a conglomerate tight reservoir according to claim 1, characterized in that: The particle size gradation curve includes a normal distribution gradation; the normal distribution gradation is calculated using the following formula: Where, μ is the diameter of the gravel particles, in mm; σ is the deviation factor, without unit; x is the gravel distribution data of the sandy conglomerate reservoir; f(x) is the distribution function of gravel.

5. The method for characterizing the distribution of conglomerate in a conglomerate tight reservoir according to claim 1, characterized in that: The aggregate types of the gravel particles include two-dimensional regular polygonal aggregate, random polygonal aggregate, two-dimensional circular aggregate, two-dimensional elliptical aggregate and three-dimensional polyhedral aggregate.

6. The method for characterizing the distribution of conglomerate in a conglomerate tight reservoir according to claim 5, characterized in that: When the aggregate type of the particles is two-dimensional polygonal aggregate or three-dimensional polyhedral aggregate, the minimum number of edges is ≥3, the maximum number of edges is ≥the minimum number of edges; the radius random coefficient is between 0 and 1; the angle random coefficient is between 0 and 1; and the shape coefficient is ≥1.

2.

7. The method for characterizing the distribution of conglomerate in a conglomerate tight reservoir according to claim 4, characterized in that: The particle size distribution of the particles in step (II) is developed using any one of Fuller grading, Fuller grading + Walaven or Taybo grading.

8. The method for characterizing the distribution of conglomerate in a conglomerate tight reservoir according to claim 7, characterized in that: When D max / d min When ≤5, the particle size distribution of the particles in step (II) is expanded using Fuller gradation, and the Fuller gradation expansion formula is formula (2): When 5<D max / d min When ≤10, the particle size distribution of the particles in step (II) is expanded by Fuller gradation + Walawen gradation, and the expansion formula of Fuller gradation + Walawen gradation is formula (3), When D max / d min >10, the particle size distribution of the particles in step (II) is expanded using the Taybo grading formula, and the Taybo grading formula is formula (4): P(D <d i )=P k (d i / D max ) n ,n∈(0.3~0.7) Formula (4) In formulas (2) to (4), D is the particle diameter in mm; d min D is the minimum particle diameter, in mm; max is the maximum particle diameter, in mm; P k For a particle size d i The passing percentage of the skeleton, in %; d i For the skeleton i The diameter of the pipe is in mm; n is the exponential coefficient, 0.3<n<0.

7.

9. The method for characterizing the distribution of conglomerate in a conglomerate tight reservoir according to claim 2, characterized in that: If the error between the particle size gradation parameters generated in step (II) and the gravel particle size and gravel content obtained in step (I) is less than 10%, the simulation accuracy is considered to meet the requirements, otherwise the gradation parameters are further adjusted until they meet the requirements.

Images