阅读说明：本技术 一种基于核磁共振和分形维数的岩石渗透率预测方法 (Rock permeability prediction method based on nuclear magnetic resonance and fractal dimension ) 是由 吴丰 罗莹莹 代槿 石祥超 李玮 何江 于 2021-08-20 设计创作，主要内容包括：本发明公开一种基于核磁共振和分形维数的岩石渗透率预测方法,包括：获取待预测岩样的核磁共振数据以及少量刻度用岩样的核磁共振数据、气体渗透率；获取岩石样品的核磁共振孔隙度、束缚水饱和度、核磁共振T-(2)谱对数平均值和核磁共振T-(2)分布曲线；制作以logT-(2)为横坐标和logS-(v)为纵坐标的散点图；根据最优拟合法确定散点图中岩样的微孔隙分形维数和宏孔隙分形维数；确定岩石样品的渗流空间以微孔隙为主还是以宏孔隙为主,并建立该储层渗透率预测模型；计算储层渗透率预测模型的拟合系数；根据储层渗透率预测模型预测待预测岩样的渗透率。本发明仅需要少量的岩心气体渗透率进行标定,即可对测量了核磁共振的待预测岩样或地层进行渗透率预测。(The invention discloses a rock permeability prediction method based on nuclear magnetic resonance and fractal dimension, which comprises the following steps: acquiring nuclear magnetic resonance data of a rock sample to be predicted, nuclear magnetic resonance data of a small amount of scale rock samples and gas permeability; obtaining nuclear magnetic resonance porosity, irreducible water saturation and nuclear magnetic resonance T of rock sample 2 Spectral log mean and nuclear magnetic resonance T 2 A distribution curve; is made of logT 2 As abscissa and logS v A scatter plot of the ordinate; determining the micropore fractal dimension and the macro-pore fractal dimension of the rock sample in the scatter diagram according to an optimal fitting method; determining whether the seepage space of the rock sample is mainly microporous or macro-porous, and establishing a reservoir permeability prediction model; calculating a fitting coefficient of a reservoir permeability prediction model; and predicting the permeability of the rock sample to be predicted according to the reservoir permeability prediction model. The invention only needs a small amount of core gas permeability for calibration, namelyPermeability prediction can be performed on the rock sample or the stratum to be predicted, which is measured by nuclear magnetic resonance.)

1. A rock permeability prediction method based on nuclear magnetic resonance and fractal dimension is characterized by comprising the following steps:

s1, acquiring nuclear magnetic resonance data of the rock sample to be predicted, nuclear magnetic resonance data of a small amount of rock samples for calibration and gas permeability;

s2, obtaining the nuclear magnetic resonance porosity of the rock sample according to the nuclear magnetic resonance data of the rock sampleIrreducible water saturation S_wiNuclear magnetic resonance T₂Spectrum log mean value T_2lmAnd nuclear magnetic resonance T₂A distribution curve;

s3 according to nuclear magnetic resonance T₂Distribution curves were made with logT₂As abscissa and logS_vA scatter plot of the ordinate;

s4, selecting different logTs in the scatter diagram₂The value is used as a demarcation point, and the data point is divided into two sections; and respectively calculate the different logT₂Correlation coefficient R of linear fitting relation of left data point of boundary point with value as boundary point_micCorrelation coefficient R in linear fitting relation with data points on right side of boundary_macAnd the correlation coefficient R of the linear fitting relation of the data points on the left side of the boundary_micCorrelation coefficient R in linear fitting relation with data points on right side of boundary_macThe product of (a);

s5, according to different logT₂Correlation coefficient R of linear fitting relation of boundary left data point with value as boundary point_micCorrelation coefficient R in linear fitting relation with data points on right side of boundary_macDetermining the maximum value of the product, comparing the maximum value of the product with a preset threshold value, and if the maximum value of the product is greater than or equal to the threshold value, determining the micro-pore fractal dimension D of the rock sample in the scatter diagram by adopting a two-stage optimal fitting method_micAnd macro-pore fractal dimension D_mac(ii) a If the value is less than the threshold value, determining the micropore fractal dimension D of the rock sample in the scatter diagram by adopting a three-stage optimal fitting method_micAnd macro-pore fractal dimension D_mac；

S6, determining whether the seepage space of the rock sample is mainly micropore or macro-pore, and determining the fractal dimension D of the micropore_micAnd macro-pore fractal dimension D_macEstablishing a reservoir permeability prediction model;

s7, gas permeability and nuclear magnetic resonance porosity according to rock sampleIrreducible water saturation S_wiNuclear magnetic resonance T₂Spectrum log mean value T_2lmDetermining a fitting coefficient of a reservoir permeability prediction model;

and S8, predicting the permeability of the rock sample to be predicted according to the reservoir permeability prediction model with the fitting coefficient obtained in the step S7 and the nuclear magnetic resonance data of the rock sample to be predicted.

2. The method for predicting rock permeability based on nuclear magnetic resonance and fractal dimension as claimed in claim 1, wherein the specific steps of step S3 are as follows: according to nuclear magnetic resonance T₂Obtaining transverse relaxation time T of nuclear magnetic resonance of rock sample by distribution curve₂Cumulative pore volume ratio S of pore radius less than r in rock_vAre respectively aligned with T₂And S_vLogarithm is obtained to obtain logT₂And logS_v(ii) a Then using logT₂As abscissa, logS_vAs a ordinate, a scattergram was made.

3. The method of claim 1, wherein the calculation formula in step S4 is as follows:

in the formula: r is a correlation coefficient, and i is a data point serial number.

4. The method for predicting permeability of rock based on NMR and fractal dimension of claim 3, wherein the two-stage optimization method in step S5 determines the fractal dimension D of microporosity of rock sample in the scattergram_micAnd macro-pore fractal dimension D_macThe method comprises the following specific steps:

selecting logT corresponding to maximum product value₂The value is used as a demarcation point, and the data point in the scatter diagram is divided into two sections;

and respectively carrying out linear fitting on the two segments of data points to obtain the slope k of a linear fitting relation formula of the data points on the left side of the limit_micSlope k of a linear fit to the boundary right data points_mac；

Then linearly fitting the slope k of the relational expression according to the left data point of the limit_micSlope k of a linear fit to the boundary right data points_macCalculating the micropore fractal dimension D_micAnd macro-pore fractal dimension D_mac。

5. The method for predicting permeability of rock based on NMR and fractal dimension of claim 3, wherein the three-step optimization method in step S5 is used to determine the fractal dimension D of microporosity of rock sample in the scattergram_micAnd macro-pore fractal dimension D_macThe method comprises the following specific steps:

selecting two different logTs₂Dividing the data points in the scatter diagram into three sections by taking the values as demarcation points;

respectively calculating a plurality of groups of two different logTs₂Correlation coefficient R of linear fitting relation of left data point of boundary point with value as boundary point_micCorrelation coefficient R of linear fitting relation of intermediate data points_nonCorrelation coefficient R in linear fitting relation with data points on right side of boundary_macThen determining two logTs corresponding to the maximum value of the product₂A value;

calculating two logTs corresponding to the maximum value of the product₂Slope k of the linear fit relation for the left data points of the boundary at the time of the value_micSlope k of a linear fit to the boundary right data points_mac；

Then linearly fitting the slope k of the relational expression according to the left data point of the limit_micSlope k of a linear fit to the boundary right data points_macCalculating the micropore fractal dimension D_micAnd macro-pore fractal dimension D_mac。

6. Rock permeability prediction method based on NMR and fractal dimension according to claim 4 or 5, characterized in that the microporosity fractal dimension D_micAnd macro-pore fractal dimension D_macThe calculation formula of (2) is as follows:

D_mic＝3-k_mic

D_mac＝3-k_mac

in the formula: d_micFractal dimension for microporosities; d_macIs a macro-pore fractal dimension; k is a radical of_micLinearly fitting the slope of the relationship to the left data point of the boundary; k is a radical of_macThe slope of the relationship is linearly fit to the data points to the right of the limit.

7. The method of claim 1, wherein the reservoir permeability prediction model in step S6 comprises:

when the seepage space of the rock sample is mainly macro-pores, and the nuclear magnetic resonance of the completely saturated water and the bound water state is measured at the same time;

when the seepage space of the rock sample is mainly macro-pores, and only the nuclear magnetic resonance of the fully saturated water state is measured;

when the seepage space of the rock sample is mainly microporous, and the nuclear magnetic resonance of completely saturated water and bound water states is measured at the same time;

when the seepage space of the rock sample is mainly micropore, and only the nuclear magnetic resonance in a fully saturated water state is measured;

in the formula: k is the permeability;is nuclear magnetic resonance porosity; s_wiIrreducible water saturation; t is_2lmIs nuclear magnetic resonance T₂A spectrum log mean value; d_micFractal dimension for microporosities; d_macIs a macro-pore fractal dimension; a is₁、b₁、c₁、n₁、a₂、b₂、c₂、n₂、a₃、b₃、c₃、d₃、n₃、a₄、b₄、c₄、d₄、n₄Are all fitting coefficients.

8. The method of claim 1, wherein logarithm is obtained on both sides of the equation of the reservoir permeability prediction model in step S7, the gas permeability of the rock sample is used as a calibration, and the fitting coefficient in the reservoir permeability prediction model is obtained through multivariate fitting.

Technical Field

The invention relates to a rock permeability prediction method based on nuclear magnetic resonance and fractal dimension, and belongs to the technical field of petroleum and natural gas exploration and development.

Background

The permeability is one of basic physical property parameters of the rock, and the accurate calculation of the parameters is important for the accurate evaluation of the oil and gas reservoir. Nuclear magnetic resonance can reflect the pore information of rock and is also commonly used to make permeability measurements. However, due to the complex pore structure characteristics of most rocks, the prediction accuracy of the methods is generally low, and the research and development of a nuclear magnetic resonance permeability prediction method with higher accuracy is urgently needed.

The invention discloses a method for predicting permeability of rock, which comprises the steps of excavating nuclear magnetic resonance pore fractal information to represent the complex pore structure characteristics of the rock, and constructing a new permeability prediction model to improve the nuclear magnetic resonance permeability prediction precision.

Disclosure of Invention

In order to overcome the problems in the prior art, the invention provides a rock permeability prediction method based on nuclear magnetic resonance and fractal dimension; the invention utilizes nuclear magnetic resonance data, determines the micropore fractal dimension and the macro-pore fractal dimension of the rock by an optimal fitting method, calibrates through a small amount of core gas permeability, determines fitting parameters based on the fractal dimension and a nuclear magnetic resonance permeability model, and finally calculates the permeability of the rock, wherein the permeability calculation result is well matched with the gas logging permeability.

The technical scheme provided by the invention for solving the technical problems is as follows: a rock permeability prediction method based on nuclear magnetic resonance and fractal dimension comprises the following steps:

s1, acquiring nuclear magnetic resonance data of the rock sample to be predicted, nuclear magnetic resonance data of a small amount of rock samples for calibration and gas permeability;

s2, obtaining the nuclear magnetic resonance porosity of the rock sample according to the nuclear magnetic resonance data of the rock sampleIrreducible water saturation S_wiNuclear magnetic resonance T₂Spectrum log mean value T_2lmAnd nuclear magnetic resonance T₂A distribution curve;

s3 according to nuclear magnetic resonance T₂Distribution curves were made with logT₂As abscissa and logS_vA scatter plot of the ordinate;

s4, selecting different logTs in the scatter diagram₂The value is used as a demarcation point, and the data point is divided into two sections; and respectively calculate the different logT₂Correlation coefficient R of linear fitting relation of left data point of boundary point with value as boundary point_micCorrelation coefficient R in linear fitting relation with data points on right side of boundary_macAnd the correlation coefficient R of the linear fitting relation of the data points on the left side of the boundary_micCorrelation coefficient R in linear fitting relation with data points on right side of boundary_macThe product of (a);

s5, according to different logT₂Correlation coefficient R of linear fitting relation of boundary left data point with value as boundary point_micCorrelation coefficient R in linear fitting relation with data points on right side of boundary_macDetermining the maximum value of the product, comparing the maximum value of the product with a preset threshold value, and if the maximum value of the product is greater than or equal to the threshold value, determining the micro-pore fractal dimension D of the rock sample in the scatter diagram by adopting a two-stage optimal fitting method_micAnd macro-pore fractal dimension D_mac(ii) a If the value is less than the threshold value, determining the micropore fractal dimension D of the rock sample in the scatter diagram by adopting a three-stage optimal fitting method_micAnd macro-pore fractal dimension D_mac；

S6, determining whether the seepage space of the rock sample is mainly micropore or macro-pore, and determining the fractal dimension D of the micropore_micAnd macro-pore fractal dimension D_macEstablishing a reservoir permeability prediction model;

s7, gas permeability and nuclear magnetic resonance porosity according to rock sampleIrreducible water saturation S_wiNuclear magnetic resonance T₂Spectrum log mean value T_2lmDetermining fitting coefficients of a reservoir permeability prediction model;

and S8, predicting the permeability of the rock sample to be predicted according to the reservoir permeability prediction model with the fitting coefficient obtained in the step S7 and the nuclear magnetic resonance data of the rock sample to be predicted.

The further technical scheme is that the specific steps of step S3 are as follows: according to nuclear magnetic resonance T₂Obtaining transverse relaxation time T of nuclear magnetic resonance of rock sample by distribution curve₂Cumulative pore volume ratio S of pore radius less than r in rock_vAre respectively aligned with T₂And S_vLogarithm is obtained to obtain logT₂And logS_v(ii) a Then using logT₂As abscissa, logS_vAs a ordinate, a scattergram was made.

The further technical solution is that the calculation formula in step S4 is:

in the formula: r is a correlation coefficient, and i is a data point serial number.

The further technical scheme is that the two-stage optimal fitting method in the step S5 determines the micropore fractal dimension D of the rock sample in the scatter diagram_micAnd macro-pore fractal dimension D_macThe method comprises the following specific steps:

selecting logT corresponding to maximum product value₂The value is used as a demarcation point, and the data point in the scatter diagram is divided into two sections;

and respectively carrying out linear fitting on the two segments of data points to obtain the slope k of a linear fitting relation formula of the data points on the left side of the limit_micSlope k of a linear fit to the boundary right data points_mac；

Then linearly fitting the slope k of the relational expression according to the left data point of the limit_micSlope k of a linear fit to the boundary right data points_macComputingFractional dimension D of microporosity_micAnd macro-pore fractal dimension D_mac。

The further technical scheme is that the three-stage optimal fitting method in the step S5 determines the micropore fractal dimension D of the rock sample in the scatter diagram_micAnd macro-pore fractal dimension D_macThe method comprises the following specific steps:

selecting two different logTs₂Dividing the data points in the scatter diagram into three sections by taking the values as demarcation points;

respectively calculating a plurality of groups of two different logTs₂Correlation coefficient R of linear fitting relation of left data point of boundary point with value as boundary point_micCorrelation coefficient R of linear fitting relation of intermediate data points_nonCorrelation coefficient R in linear fitting relation with data points on right side of boundary_macThen determining two logTs corresponding to the maximum value of the product₂A value;

calculating two logTs corresponding to the maximum value of the product₂Slope k of the linear fit relation for the left data points of the boundary at the time of the value_micSlope k of a linear fit to the boundary right data points_mac；

Then linearly fitting the slope k of the relational expression according to the left data point of the limit_micSlope k of a linear fit to the boundary right data points_macCalculating the micropore fractal dimension D_micAnd macro-pore fractal dimension D_mac。

The further technical proposal is that the fractal dimension D of the microporosity_micAnd macro-pore fractal dimension D_macThe calculation formula of (2) is as follows:

D_mic＝3-k_mic

D_mac＝3-k_mac

in the formula: d_micFractal dimension for microporosities; d_macIs a macro-pore fractal dimension; k is a radical of_micLinearly fitting the slope of the relationship to the left data point of the boundary; k is a radical of_macThe slope of the relationship is linearly fit to the data points to the right of the limit.

The further technical scheme is that the reservoir permeability prediction model in the step S6 includes:

when the seepage space of the rock sample is mainly macro-pores, and the nuclear magnetic resonance of the completely saturated water and the bound water state is measured at the same time;

when the seepage space of the rock sample is mainly macro-pores, and only the nuclear magnetic resonance of the fully saturated water state is measured;

when the seepage space of the rock sample is mainly microporous, and the nuclear magnetic resonance of completely saturated water and bound water states is measured at the same time;

when the seepage space of the rock sample is mainly micropore, and only the nuclear magnetic resonance in a fully saturated water state is measured;

in the formula: k is the permeability;is nuclear magnetic resonance porosity; s_wiIrreducible water saturation; t is_2lmIs nuclear magnetic resonance T₂A spectrum log mean value; d_micFractal dimension for microporosities; d_macIs a macro-pore fractal dimension; a is₁、b₁、c₁、n₁、a₂、b₂、c₂、n₂、a₃、b₃、c₃、d₃、n₃、a₄、b₄、c₄、d₄、n₄Are all fitting coefficients.

The further technical scheme is that logarithm is solved on two sides of an equation of the reservoir permeability prediction model in the step S7, the gas permeability of the rock sample is used as calibration, and a fitting coefficient in the reservoir permeability prediction model is solved through multivariate fitting.

The invention has the following beneficial effects:

(1) determining micro-pore and macro-pore fractal boundaries by an optimal fitting method, and accurately calculating micro-pore fractal dimension and macro-pore fractal dimension;

(2) the fractal dimension is introduced into the nuclear magnetic resonance permeability model, so that the permeability prediction precision is obviously improved.

Drawings

FIG. 1 is a schematic diagram of NMR data preprocessing;

FIG. 2 is a schematic diagram of a two-stage optimal fitting method;

FIG. 3 is a diagram of two-stage optimal fitting method for determining micro-pore and macro-pore fractal limits;

FIG. 4 is a three-stage optimal fitting method for determining fractal boundary diagrams of micropores, fractal pores and macrocells;

FIG. 5 is a graph comparing predicted and gas permeability for the model of the present invention with a conventional model.

Detailed Description

The technical solutions of the present invention will be described clearly and completely with reference to the accompanying drawings, and it should be understood that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

The invention discloses a rock permeability prediction method based on nuclear magnetic resonance and fractal dimension, which comprises the following steps:

s1, acquiring nuclear magnetic resonance data of the rock sample to be predicted, nuclear magnetic resonance data of a small amount of calibration rock samples and gas permeability;

s2, obtaining the nuclear magnetic resonance porosity of the rock sample according to the nuclear magnetic resonance data of the rock sample with a small amount of scalesIrreducible water saturation S_wiNuclear magnetic resonance T₂Spectrum log mean value T_2lmAnd nuclear magnetic resonance T₂Profile (as shown in fig. 1A);

s3 according to nuclear magnetic resonance T₂Obtaining transverse relaxation time T of nuclear magnetic resonance of rock sample by distribution curve₂Cumulative pore volume ratio S of pore radius less than r in rock_vAre respectively aligned with T₂And S_vLogarithm is obtained to obtain logT₂And logS_v(ii) a Then using logT₂As abscissa, logS_vAs a ordinate, a scatter plot was made (as shown in fig. 1B);

s4, selecting different logTs in the scatter diagram₂The value is used as a demarcation point, and the data point is divided into two sections; and respectively calculate the different logT₂Correlation coefficient R of linear fitting relation of left data point of boundary point with value as boundary point_micCorrelation coefficient R in linear fitting relation with data points on right side of boundary_macAnd the correlation coefficient R of the linear fitting relation of the data points on the left side of the boundary_micCorrelation coefficient R in linear fitting relation with data points on right side of boundary_macThe product of (a);

s5, according to different logT₂Correlation coefficient R of linear fitting relation of boundary left data point with value as boundary point_micCorrelation coefficient R in linear fitting relation with data points on right side of boundary_macDetermining the maximum value of the product, comparing the maximum value of the product with a preset threshold value, and if the maximum value of the product is greater than or equal to the threshold value, determining the micro-pore fractal dimension D of the rock sample in the scatter diagram by adopting a two-stage optimal fitting method_micAnd macro-pore fractal dimension D_mac(ii) a If the value is less than the threshold value, determining the micropore fractal dimension D of the rock sample in the scatter diagram by adopting a three-stage optimal fitting method_micAnd macro-pore fractal dimension D_mac；

Two-stage optimal fitting method

According to the fractal theory, the following results are obtained:

in the formula: r is the rock pore radius, r_maxRadius of the largest pore in the rock, S_vThe cumulative pore volume ratio of the pore radius in the rock is smaller than r, and D is the pore fractal dimension of the rock.

In rocks, the general volume and diffusion relaxations are negligible and therefore the nuclear magnetic resonance transverse relaxation T is₂Positive correlation with pore radius r (equation 2):

T₂∝r (2)

from equations (1) and (2), we can derive:

in the formula: t is₂Transverse relaxation time, T, for nuclear magnetic resonance of rock samples_2maxIs the maximum transverse relaxation time of nuclear magnetic resonance of the rock sample.

Logarithm is obtained on both sides of equation (3) to obtain:

logS_V＝(3-D)logT₂-(3-D)logT_2max (4)

therefore, at logT₂Is abscissa and logS_vIn the ordinate scattergram, the data points with fractal characteristics will have a linear relationship, and the fractal dimension D of the pore can be obtained by fitting the slope (3-D) of the resulting linear relationship (equation 4).

In practical rocks, there is a certain difference in fractal characteristics of micropores and macrocells, and thus in logT₂Is abscissa and logS_vIn the ordinate scattergram, these data points can be fitted to at least 2-segment linear relationships, i.e., a linear relationship representing microporosities and a linear relationship representing macrocells.

At logT₂Is abscissa and logS_vSelecting different logTs in a scatter diagram for the ordinate₂As a demarcation point, the data points are divided into two segments (fig. 2). Calculating the limit left data point and the limit right data point respectively through the formula (5)Correlation coefficient R of linear fitting relation_micAnd R_macWhen R is_micAnd R_macWhen the product of (1) is the maximum (equation 6), the selected logT is used₂The optimal limits for the micropore fractal data and the macro-pore fractal data are shown in fig. 2E.

In the formula: r is a correlation coefficient, and i is a data point serial number.

max(R_micR_mac) (6)

In the formula: r_micAnd R_macAnd linear fitting correlation coefficients of the micropore fractal data points and the macro-pore fractal data points are respectively obtained.

② three-stage optimum fitting method

And when the maximum value of the product of the fitting correlation coefficients of the micropore data points and the macro pore data points is lower than a certain preset threshold value, indicating that a section of pore space without fractal characteristics exists in the rock. At this time, at logT₂Is abscissa and logS_vIn the scatter diagram of ordinate, two logTs need to be selected₂The values serve as demarcation points that divide the data points into three segments. Calculating the correlation coefficient R of the linear fitting relationship of the left data point, the middle data point and the right data point respectively through the formula (5)_mic、R_nonAnd R_macWhen R is_mic、R_nonAnd R_macWhen the product of (1) is the maximum (equation 7), two logTs are selected at this time₂The value is the optimal limit of micropore fractal data, fractal characteristic data and macro-pore fractal data.

max(R_micR_nonR_mac) (7)

In the formula: r_nonAnd linearly fitting correlation coefficients of the data points without fractal features.

Obtaining micro-pore fractal dimension and macro-pore fractal dimension

For the micropore data points, the micropore fractal dimension D can be obtained according to the slope of the linear fitting relation_mic。

D_mic＝3-k_mic (8)

In the formula: d_micFractal dimension for microporosities; k is a radical of_micThe slope of the relationship is linearly fit to the left data points of the boundary.

For the macro-pore data points, the macro-pore fractal dimension D can be obtained according to the slope of the linear fitting relation_mac。

D_mac＝3-k_mac (9)

In the formula: d_macIs a macro-pore fractal dimension; k is a radical of_macThe slope of the relationship is linearly fit to the data points to the right of the limit.

S6, determining whether the seepage space of the rock sample is mainly micropore or macro-pore, and determining the fractal dimension D of the micropore_micAnd macro-pore fractal dimension D_macEstablishing a reservoir permeability prediction model;

when the seepage space of the rock sample is mainly macro-pores, and the nuclear magnetic resonance of the completely saturated water and the bound water state is measured at the same time;

when the seepage space of the rock sample is mainly macro-pores, and only the nuclear magnetic resonance of the fully saturated water state is measured;

when the seepage space of the rock sample is mainly microporous, and the nuclear magnetic resonance of completely saturated water and bound water states is measured at the same time;

when the seepage space of the rock sample is mainly micro-macro pores, only the nuclear magnetic resonance of the fully saturated water state is measured;

in the formula: k is the permeability;is the magnetic resonance porosity; s_wiIrreducible water saturation; t is_2lmIs nuclear magnetic resonance T₂A spectrum log mean value; d_micFractal dimension for microporosities; d_macIs a macro-pore fractal dimension; a is₁、b₁、c₁、n₁、a₂、b₂、c₂、n₂、a₃、b₃、c₃、d₃、n₃、a₄、b₄、c₄、d₄、n₄Are all fitting coefficients;

s7 magnetic resonance porosity according to permeability of rock sampleIrreducible water saturation S_wiNuclear magnetic resonance T₂Spectrum log mean value T_2lmCalculating a fitting coefficient of the reservoir permeability prediction model established in the step S5;

logarithms are taken on both sides of the equations (10) to (13) to obtain equations (14) to (17), respectively. Selecting some rocks with nuclear magnetic resonance and gas permeability measured at the same time, and taking the gas permeability as calibration, namely obtaining permeability model parameters through multivariate fitting;

and S8, predicting the permeability of the rock sample to be predicted according to the reservoir permeability prediction model with the fitting coefficient obtained in the step S7 and the nuclear magnetic resonance data of the rock sample to be predicted.

For the rock measured nuclear magnetic resonance, the permeability can be calculated by formula (10) -formula (13) and the determined model parameters.

Examples are given.

The 30 sandstone samples of the XC-MJ gas field are taken as an example to illustrate the specific implementation mode of the invention.

(1) Nuclear magnetic resonance data preprocessing

Nuclear magnetic resonance measurements of fully saturated and bound water states were performed on 30 sandstone samples of the XC-MJ gas field, as well as gas permeability measurements (table 1, column 3). The sandstone samples with the numbers of #1 to #15 are used for calibrating the model and determining parameters in the permeability model; sandstone samples with numbers #16 to #30 are tested for the prediction effect of the model through the gas permeability and compared with a common prediction model. Nuclear magnetic resonance porosity, irreducible water saturation and nuclear magnetic resonance T obtained by nuclear magnetic resonance of the 30 sandstone samples₂The log mean values of the spectra are given in columns 4-6 of Table 1.

(2) Determination of microvoid and macrocellular boundaries by optimal fitting

The preset threshold value is 0.97, data points of a part of sandstone samples in 30 sandstone samples listed in table 1 can be fitted into a two-segment linear relationship, and two-segment optimal fitting methods are adopted to distinguish the fractal boundaries of micropores and macrocells, and fig. 3 shows the results of determining the fractal boundaries of the micropores and the macrocells by using the two-segment optimal fitting methods for sandstone samples #1, #6 and # 11; in addition, a pore interval without fractal characteristics exists in data points of a part of sandstone samples, boundaries of microporosities, non-fractal pores and fractal boundaries of macrocells are distinguished by adopting a two-stage optimal fitting method, and fig. 4 shows the results of determining the fractal boundaries of microporosities, non-fractal pores and macrocells by utilizing a three-stage optimal fitting method for sandstone samples #14, #17 and # 27.

TABLE 1 sandstone sample Properties and NMR parameters

(3) Obtaining micropore fractal dimension and macro-pore fractal dimension

For the micropore data points, a linear fitting relational expression is established, and the micropore fractal dimension D can be obtained according to the slope of the linear fitting relational expression_micAnd macro-pore fractal dimension D_mac(FIG. 3D, E, F and FIG. 4D, E, F are examples).

(4) Permeability model and parameter determination

In this embodiment, the seepage space of sandstone is mainly macro-pores, and the permeability of sandstone is mainly related to the fractal dimension of macro-pores, so that the permeability model can use the formula (10) and the formula (11).

Taking logk as the dependent variable,performing multivariate linear fitting to obtain a formula (18) by taking the independent variable as an independent variable; similarly, taking logk as a dependent variable,logT_2lmas independent variables, multivariate linear fitting was performed to obtain equation (19).

The formula (18) and the formula (19) are respectively corresponded to obtain a₁＝0.26934、b₁＝-11.81457、c₁＝4.19888、n₁＝1.23826、a₂＝0.001880627、b₂＝-14.72545、c₂＝4.7488、n₂0.78605. Therefore, the final permeability model equations (10) and (11) can be expressed as equations (20) and (21).

(5) Permeability calculation and verification

Permeability calculations were performed on the sandstone samples nos. #16 to #30 using the formula (10) and the formula (11), that is, the formula (20) and the formula (21), respectively. The comparison graph (figure 5) and the comparison table (table 2) of the predicted permeability and the gas permeability of the model (formula 10 and formula 11) and the common model (pore permeability model and SDR model) show that the prediction result of the permeability of the model of the invention has higher precision and smaller error than the prediction result of the permeability of the common model.

TABLE 2 sandstone sample Properties and NMR parameters

Note: the pore permeability model is a formula (22), the SDR model is a formula (23), the logarithmic error of the predicted permeability and the gas permeability is calculated by adopting a formula (24), and the average logarithmic error of the predicted permeability and the gas permeability of the sandstone samples with the numbers of #16 to #30 is calculated by adopting a formula (25).

LE＝logk^*-logk (24)

In the formula: k is a radical of^*To predict permeability, LE is the log error of the predicted permeability and the gas permeability.

In the formula: MALE is the average logarithmic error of predicted permeability and gas permeability of sandstone samples with numbers #16 to # 30.

Although the present invention has been described with reference to the above embodiments, it should be understood that the present invention is not limited to the above embodiments, and those skilled in the art can make various changes and modifications without departing from the scope of the present invention.