Method and device for predicting continental facies mud shale diagenesis, storage medium and equipment

Abstract

The invention relates to a method and device for predicting the diagenesis of continental facies mud shale, a storage medium and equipment, and the method comprises the steps: 1) extracting reservoir features of the continental facies mud shale, obtaining Archie parameters in the continental facies mud shale, determining a cementation coefficient m of the continental facies mud shale through three-dimensional regression, and calculating a diagenesis coefficient MD; 2) determining most sensitive parameter data for diagenesis, and dividing the most sensitive parameter data into a training set, a verification set and a test set; 3) establishing an artificial neural network model according to the Archie parameter and the cementation coefficient, and training the artificial neural network model by using the training set; 4) continuously testing errors of the model in the training process of the artificial neural network model by using the verification set and the test set; (5) evaluating the performance of the trained artificial neural network model, if the performance acceptability is good, executing the step (6), otherwise, returning to the step (3); and 6) predicting the diagenesis of continental shale by using the artificial neural network model with good performance acceptance.

Description

Technical Field

[0001] The present invention relates to the field of geophysical technology, and in particular to a method, a device, a storage medium and equipment for predicting diagenesis of continental mudstone. Background Art

[0002] Compared with sandstone and carbonate rock, mud shale has small grain size and is difficult to observe. At the same time, it is limited by microscopic experimental conditions. Its diagenesis is a weak field of research in the sedimentology community and even the geological community. The diagenesis of mud shale is restricted by the evolution of physical, chemical and biological factors, and the driving mechanism of diagenesis is complex and changeable due to the superposition of actual geological conditions and complex factors. Moreover, microorganisms are involved in almost all diagenetic processes, so diagenesis is a typical biogeochemical process. The characteristics inherited from the early diagenesis process will affect the progress of the middle and late diagenesis, thereby affecting the physical and mechanical properties of the mud shale strata. Controlled by complex actual geological conditions, especially the limitations of the study of the material transport process and its transport mechanism in the middle and late diagenesis stages, the driving mechanism and evolution path of diagenesis are relatively weak, and the control mechanism of diagenesis is extremely complex. In addition, during the diagenesis process, deformation and crack morphology are usually related to stress distribution and geometry.

[0003] Diagenesis is the action of physical, chemical, and / or biological processes on the structure and mineral composition of rocks during the process of sedimentation and diagenesis. Diagenesis continues as long as the sediments are in contact with sufficient amounts of chemically active fluids, either atmospheric or marine (brackish, normal salinity, or hypersaline). Diagenesis is usually restricted to low temperature and low pressure conditions, explicitly excluding processes and products associated with metamorphism. Diagenetic processes are usually related to the diagenetic environment, which provides the general framework for interpreting the geometry of cementation and porous units. Several factors influence diagenesis, including the prevailing chemical conditions (redox conditions, pH, and CO2), diffusion and ion exchange rates, the ratio of low-Mg calcite to aragonite, erosion rates, sediments preceding the diagenetic history, grain size, organic matter availability, rate and composition of interstitial hydrocarbon fluids, bacterial abundance, clay percentage, and physical conditions. Diagenesis typically reduces total porosity through various forms of cementation and mineral growth, but may also increase porosity by forming secondary pore space through leaching of the particle matrix (dissolution) or by creating / redistributing pores through dolomitization.

[0004] However, it is difficult to accurately predict the diagenesis of continental shale using existing technologies. Summary of the invention

[0005] In view of the above problems, the object of the present invention is to provide a method, device, storage medium and equipment for predicting the diagenesis of continental mud shale, which can accurately predict the diagenesis of continental mud shale.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] The method for predicting continental shale diagenesis of the present invention comprises the following steps:

[0008] 1) Extract the reservoir characteristics of continental shale, obtain the Archie parameters in continental shale, determine the cementation coefficient m of continental shale through three-dimensional regression, and calculate the diagenesis coefficient MD;

[0009] 2) Based on the diagenesis coefficient, determine the most sensitive parameter data to diagenesis, and divide the most sensitive parameter data into a training set, a validation set, and a test set;

[0010] 3) Establishing an artificial neural network model according to Archie parameters and cementation coefficients, and using the training set to train the artificial neural network model;

[0011] 4) Use the validation set and test set to continuously test the model error during the training process of the artificial neural network model;

[0012] 5) Evaluate the performance of the trained artificial neural network model. If the performance acceptance is good, proceed to step 6), otherwise return to step 3);

[0013] 6) Use an artificial neural network model with good performance acceptance to predict the diagenesis of continental mudstone.

[0014] In the method, preferably, in the above step 1), the Resform software is used to extract the characteristics of the continental shale reservoir, and the Pickett Plot technology is used to obtain the Archie parameters in the continental shale.

[0015] In the method, preferably, in the above step 2), the most sensitive parameter data to diagenesis are determined using the relative importance algorithm of Microsoft Excel.

[0016] In the method described above, preferably, in step 1), the method for determining the cementation coefficient m of the continental shale by three-dimensional regression is:

[0017] In rock physics, the formation factor F and effective porosity Represents the fundamental parameter that controls electrical properties and is physically measured using resistivity and porosity logs or conventional core analysis. Formation coefficient is mathematically related to resistivity as:

[0018]

[0019] Where, F is the formation factor; R o is the true resistivity of the water-saturated formation; R w is the formation water resistivity; is the effective porosity; a is the proportional coefficient related to the rock, which is 0.6 to 1.55; m is the cementation coefficient of continental shale;

[0020] The formation water saturation S is quantitatively calculated using the formation coefficient F in formula (1). w :

[0021]

[0022] In the formula, S w is the formation water saturation; n is the saturation index; R t is the true resistivity of the formation rock;

[0023] Substituting formula (1) into formula (2) yields the saturation S of the rock fabric characteristics: wa and the saturation S of the pore fabric characteristics wr :

[0024]

[0025] The Picket Plot technique can be mathematically expressed as the well logging value S by formula (5): w , R w / R t and porosity The function is:

[0026]

[0027] Furthermore, formula (5) can be decomposed into three new equations, which can be regarded as the plane equations in the three-dimensional space of x, y and z coordinates:

[0028] Therefore, formula (5) can be rewritten as the formula for each physical measurement point i:

[0029] Z i =-A+mX i +nY i (6)

[0030] Among them, the parameters X, Y, and Z are related to the well logging curve, A = loga, a, m, and n are obtained by constructing three different equations at different physical measurement points i through drilling records. Formula (7) is obtained by summing formula (6) for N measurement points, and formulas (8) and (9) are respectively obtained by multiplying formula (7) by ∑X i and ∑Y i The specific three equations are:

[0031] ∑Z i =-N*A+m*∑X i +n*∑Y i (7)

[0032]

[0033] ∑Y i Z i =-N*A*∑Y i +m*∑X i Y i +n*∑Y i 2 (9)

[0034] By using MATLAB to simultaneously solve the equations composed of formulas (7)-(9) to obtain the values ​​of A, m and n, the results are then compared with the Picket Plot technology results, and a reasonable cementation coefficient m of continental shale is obtained.

[0035] In the method described above, preferably, the diagenesis coefficient MD can be expressed by formula (10):

[0036] MD=m 2 +m*(S wr -S wa ) (10)

[0037] Substituting formula (3) and (4) into formula (10), we can obtain:

[0038]

[0039] In the formula, R lls is the shallow resistivity; R lld is the deep resistivity; R mf is the resistivity of mud filtrate.

[0040] The device for predicting continental shale diagenesis of the present invention comprises:

[0041] The first processing unit is used to extract the characteristics of the continental shale reservoir, obtain the Archie parameters in the continental shale, determine the cementation coefficient m of the continental shale through three-dimensional regression, and calculate the diagenesis coefficient MD;

[0042] A second processing unit is used to determine the most sensitive parameter data to the diagenesis based on the diagenesis coefficient, and divide the most sensitive parameter data into a training set, a validation set and a test set;

[0043] A third processing unit is used to establish an artificial neural network model according to the Archie parameter and the cementation coefficient, and train the artificial neural network model using the training set;

[0044] A fourth processing unit, used to continuously test the error of the model during the training process of the artificial neural network model using the validation set and the test set;

[0045] a fifth processing unit, used for evaluating the performance of the trained artificial neural network model;

[0046] The sixth processing unit is used to predict the diagenesis of continental shale by using an artificial neural network model with good performance acceptance.

[0047] The computer storage medium of the present invention stores a computer program thereon, and when the computer program is executed by a processor, the method steps for predicting the diagenesis of continental shale are implemented.

[0048] The computer device described in the present invention includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method steps for predicting the diagenesis of continental mudstone are implemented.

[0049] The present invention adopts the above technical solution, which has the following advantages:

[0050] The method for predicting continental shale diagenesis of the present invention can accurately predict the continental shale diagenesis, and the artificial neural network is highly efficient in quantifying the diagenesis using well logging data, thereby better characterizing the complex shale system. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Throughout the accompanying drawings, the same reference numerals are used to represent the same components. In the accompanying drawings:

[0052] Figure 1 It is a flow chart of a method for predicting continental shale diagenesis provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0053] The exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0054] The present invention provides a method for predicting diagenesis of continental shale, using artificial neural network technology with feedforward back propagation to directly predict shale diagenesis from well logs of different reservoirs with different characteristics. The well log and core data of these wells are analyzed, the diagenesis is quantified (diagenesis coefficients or MD coefficients are calculated), and these diagenesis values ​​are related to known diagenesis types and pore types. The final artificial neural network output model develops a numerical scale between 0 and 10 for shale diagenesis, and is compared with the main diagenesis in the corresponding core samples to determine the relationship between the numerical scale and diagenesis.

[0055] The method for predicting the diagenesis of continental shale provided by the present invention comprises the following steps:

[0056] 1) The Resform software was used to extract the reservoir characteristics of continental shale; the Pickett Plot technique was used to obtain the Archie parameters in the continental shale, the cementation coefficient m of the continental shale was determined by three-dimensional regression, and the diagenesis coefficient MD was calculated;

[0057] 2) Based on the diagenesis coefficient MD, the most sensitive parameter data to diagenesis are determined using the relative importance algorithm of Microsoft Excel, and the most sensitive parameter data are divided into a training set, a validation set, and a test set;

[0058] 3) Establishing an artificial neural network model according to Archie parameters and cementation coefficients, and using the training set to train the artificial neural network model;

[0059] 4) Use the validation set and test set to continuously test the model error during the training process of the artificial neural network model;

[0060] 5) Evaluate the performance of the trained artificial neural network model. If the performance acceptance is good, proceed to step 6), otherwise return to step 3);

[0061] 6) Use an artificial neural network model with good performance acceptance to predict the diagenesis of continental mudstone.

[0062] In the above embodiment, preferably, the cementation coefficient m of the continental shale determined by three-dimensional regression in step 1) is specifically:

[0063] In rock physics, the formation factor F and effective porosity Represents the fundamental parameter that controls electrical properties and is physically measured using resistivity and porosity logging or conventional core analysis. Formation coefficient is mathematically related to resistivity as:

[0064]

[0065] Where, F is the formation factor; R o is the true resistivity of the water-saturated formation; R w is the formation water resistivity; is the effective porosity; a is the proportional coefficient related to the rock, which is generally 0.6 to 1.55; m is the cementation coefficient of continental shale.

[0066] The formation water saturation S is quantitatively calculated using the formation coefficient F in formula (1). w :

[0067]

[0068] In the formula, S w is the formation water saturation; n is the saturation index; R t is the true resistivity of the formation rock;

[0069] Substituting formula (1) into formula (2) yields the saturation S of the rock fabric characteristics: wa and the saturation S of the pore fabric characteristics wr :

[0070]

[0071] The Picket Plot technique can be mathematically expressed as the well logging value S by formula (5): w , R w / R t and porosity The function is:

[0072]

[0073] Furthermore, formula (5) can be decomposed into three new equations, which can be regarded as the plane equations in the three-dimensional space of x, y and z coordinates:

[0074] Therefore, formula (5) can be rewritten as the formula for each physical measurement point i:

[0075] Z i =-A+mX i +nY i (6)

[0076] Among them, the parameters X, Y, and Z are related to the well logging curve, A = loga, a, m, and n are obtained by constructing three different equations at different physical measurement points i through drilling records. Formula (7) is obtained by summing formula (6) for N measurement points, and formulas (8) and (9) are respectively obtained by multiplying formula (7) by ∑X i and ∑Y i The specific three equations are:

[0077] ∑Z i =-N*A+m*∑X i +n*∑Y i (7)

[0078]

[0079] ∑Y i Z i =-N*A*∑Y i +m*∑X i Y i +n*∑Y i 2 (9)

[0080] By using MATLAB to simultaneously solve the equations composed of formulas (7)-(9) to obtain the values ​​of A, m and n, the results are then compared with the results of the Picket Plot technique, and the cementation coefficient m of a reasonable continental shale is obtained.

[0081] In the above embodiment, preferably, the diagenesis coefficient MD can be expressed by formula (10):

[0082] MD=m 2 +m*(S wr -S wa ) (10) Substituting formula (3) and (4) into formula (10), we can obtain:

[0083]

[0084] In the formula, R lls is the shallow resistivity; R lld is the deep resistivity; R mf is the resistivity of mud filtrate.

[0085] The present invention also provides a device for predicting continental shale diagenesis, comprising:

[0086] The first processing unit is used to extract the characteristics of the continental shale reservoir, obtain the Archie parameters in the continental shale, determine the cementation coefficient m in the continental shale through three-dimensional regression, and calculate the diagenesis coefficient MD;

[0087] The second processing unit is used to determine the most sensitive parameter data to the diagenesis based on the diagenesis coefficient; and divide the most sensitive parameter data into a training set, a validation set and a test set;

[0088] A third processing unit is used to establish an artificial neural network model according to the Archie parameter and the cementation coefficient, and train the artificial neural network model using the training set;

[0089] A fourth processing unit, used to continuously test the error of the model during the training process of the artificial neural network model using the validation set and the test set;

[0090] a fifth processing unit, used for evaluating the performance of the trained artificial neural network model;

[0091] The sixth processing unit is used to predict the diagenesis of continental shale by using an artificial neural network model with good performance acceptance.

[0092] The present invention also provides a computer storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the method steps for predicting the diagenesis of continental shale are implemented.

[0093] The present invention also provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned method steps for predicting the diagenesis of continental shale when executing the computer program.

[0094] In addition, it should be noted that in order to facilitate the calculation of the diagenesis coefficient (MD), a feed-forward back-propagation artificial neural network technique was developed to link the diagenesis coefficient MD to direct well logging measurements. To achieve this, formula (6) uses the relative importance algorithm of Microsoft Excel to determine the most sensitive parameters to obtain the MD coefficient, and then to obtain the correlation between the MD coefficient and the well logging measurement. wr -S wa ), NPHI, PHIE, S w and gamma rays have a high correlation with the calculated MD coefficient. LLS and LLD logging are due to the (S wr -S wa) were ignored because of their good representation in the ANN. A total of 3077 measurements were used to build the ANN, of which 70% (2156) were randomly selected for training and validation, and the remaining 30% (921) were used for ANN testing. ANN training involved testing various hidden layer configurations, different numbers of neutrons per layer, and multiple types of activation functions. In addition, several learning techniques that modified the weights and biases of individual input variables were tested to help establish a stable computational system. The optimal ANN structure was determined based on the best correlation between the ANN prediction coefficients and the diagenesis coefficient "MD" calculated according to equation (11). Finally, the developed ANN structure was tested with the remaining 30% of the total data, and the performance of the ANN prediction was ensured by comparing the ANN output with the actual target "MD" coefficient. The performance and development of the ANN was evaluated using the techniques described by Abdulaziz et al. All well log calculations and interpretations were done using Resform software, and ANN construction and prediction management and 3D regression were done using MATLAB 2016b.

[0095] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting diagenesis of continental shale, characterized in that: The steps include: 1) Extract the reservoir characteristics of continental shale, obtain the Archie parameters in continental shale, determine the cementation coefficient m of continental shale through three-dimensional regression, and calculate the diagenesis coefficient MD; 2) Based on the diagenesis coefficient, determine the most sensitive parameter data to diagenesis, and divide the most sensitive parameter data into a training set, a validation set, and a test set; 3) Establishing an artificial neural network model according to Archie parameters and cementation coefficients, and using the training set to train the artificial neural network model; 4) Use the validation set and test set to continuously test the model error during the training process of the artificial neural network model; 5) Evaluate the performance of the trained artificial neural network model. If the performance acceptance is good, proceed to step 6), otherwise return to step 3); 6) Use an artificial neural network model with good performance acceptance to predict the diagenesis of continental mudstone.

2. The method according to claim 1, characterized in that: In the above step 1), the Resform software is used to extract the characteristics of the continental shale reservoir, and the Pickett Plot technology is used to obtain the Archie parameters in the continental shale.

3. The method according to claim 1, characterized in that: In the above step 2), the most sensitive parameter data to diagenesis are determined using the relative importance algorithm of Microsoft Excel.

4. The method according to claim 1, characterized in that: The method for determining the cementation coefficient m of continental shale by three-dimensional regression in step 1) is: In rock physics, the formation factor F and effective porosity Represents the fundamental parameter that controls electrical properties and is physically measured using resistivity and porosity logs or conventional core analysis. Formation coefficient is mathematically related to resistivity as: Where, F is the formation factor; R o is the true resistivity of the water-saturated formation; R w is the formation water resistivity; is the effective porosity; a is the proportional coefficient related to the rock, which is 0.6 to 1.55; m is the cementation coefficient of continental shale; The formation water saturation S is quantitatively calculated using the formation coefficient F in formula (1). w : In the formula, S w is the formation water saturation; n is the saturation index; R t is the true resistivity of the formation rock; Substituting formula (1) into formula (2) yields the saturation S of the rock fabric characteristics: wa and the saturation S of the pore fabric characteristics wr : The Picket Plot technique can be mathematically expressed as the well logging value S by formula (5): w , R w / R t and porosity The function is: Furthermore, formula (5) can be decomposed into three new equations, which can be regarded as the plane equations in the three-dimensional space of x, y and z coordinates: [Y = logS w ], Therefore, formula (5) can be rewritten as the formula for each physical measurement point i: Z i =-A+mX i +nY i (6) Among them, the parameters X, Y, and Z are related to the well logging curve, A = loga, a, m, and n are obtained by constructing three different equations at different physical measurement points i through drilling records. Formula (7) is obtained by summing formula (6) for N measurement points, and formulas (8) and (9) are respectively obtained by multiplying formula (7) by ∑X i and ∑Y i The specific three equations are: ∑Z i =-N*A+m*∑X i +n*∑Y i (7) ∑Y i Z i =-N*A*∑Y i +m*∑X i Y i +n*∑Y i 2 (9) By using MATLAB to simultaneously solve the equations composed of formulas (7)-(9) to obtain the values ​​of A, m and n, the results are then compared with the Picket Plot technology results, and a reasonable cementation coefficient m of continental shale is obtained.

5. The method according to claim 4, characterized in that The diagenesis coefficient MD can be expressed by formula (10): MD=m 2 +m*(S wr -S wa ) (10) Substituting formula (3) and (4) into formula (10), we can obtain: In the formula, R lls is the shallow resistivity; R lld is the deep resistivity; R mf is the resistivity of mud filtrate.

6. A device for predicting diagenesis of continental shale, characterized in that: include: The first processing unit is used to extract the characteristics of the continental shale reservoir, obtain the Archie parameters in the continental shale, determine the cementation coefficient m of the continental shale through three-dimensional regression, and calculate the diagenesis coefficient MD; A second processing unit is used to determine the most sensitive parameter data to the diagenesis based on the diagenesis coefficient, and divide the most sensitive parameter data into a training set, a validation set and a test set; A third processing unit is used to establish an artificial neural network model according to the Archie parameter and the cementation coefficient, and train the artificial neural network model using the training set; A fourth processing unit, used to continuously test the error of the model during the training process of the artificial neural network model using the validation set and the test set; a fifth processing unit, used for evaluating the performance of the trained artificial neural network model; The sixth processing unit is used to predict the diagenesis of continental shale by using an artificial neural network model with good performance acceptance.

7. A computer storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method steps for predicting continental shale diagenesis according to any one of claims 1 to 5 are implemented.

8. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method steps for predicting the diagenesis of continental shale according to any one of claims 1 to 5 are implemented.

Images