A method for predicting high-quality tight sandstone reservoirs

Abstract

The present invention discloses a method for predicting high-quality reservoirs in tight sandstone, including obtaining seismic data, logging data and core data of a target work area; obtaining a plane prediction result map of underwater distributary channels in the target work area; obtaining a plane prediction result map of rock phases of fine sand and above in the target work area; obtaining a plane prediction result map of sand bodies of fine sand and above in underwater distributary channels in the target work area; obtaining a plane prediction result map of medium sand and above in strong dissolution phases in the target work area; and obtaining a plane prediction result map of high-quality reservoirs in the target work area. The present invention completes the prediction of favorable reservoirs in tight sandstone by combining core observation, thin section identification, pore structure analysis, and physical property analysis with logging data; conducting sedimentary microfacies analysis based on well-seismic combination, and selecting the dominant sedimentary microfacies type; conducting a dominant rock phase prediction based on sedimentary microfacies constraints; conducting a dominant diagenetic phase prediction based on sedimentary microfacies-rock phase constraints; and performing three-phase coupling analysis.

Description

Technical Field

[0001] The invention relates to a method for predicting high-quality dense sandstone reservoirs, and belongs to the technical field of petroleum exploration and development. Background Art

[0002] The development of tight sandstone reservoirs at home and abroad is progressing rapidly, and it occupies an important position in global oil and gas exploration. Tight sandstone reservoirs often have low porosity-ultra-low porosity, low permeability-ultra-low permeability, strong heterogeneity, unclear sand body plane distribution law and complex diagenesis geological characteristics, which seriously affect the development effect. Therefore, accurate identification and prediction of high-quality reservoirs is an important basis for achieving efficient development of tight sandstone oil and gas reservoirs.

[0003] There are various prediction methods for high-quality tight sandstone reservoirs. One method is guided by seismic sedimentology and performs reservoir prediction by restoring the ancient water system and ancient environment under the constraints of tectonic paleo-geomorphology and constructing a sedimentary evolution model. This method is not suitable for fine-scale reservoir prediction, and the accuracy of the prediction results is not ideal.

[0004] One type is guided by geophysics, based on geophysical data such as logging data and seismic data, and mathematical methods are used to process geophysical data from different angles to improve the response of geophysical data to favorable reservoir sand bodies, such as: correction of acoustic logging curves for well-seismic combined reservoir prediction method; multi-well logging parameter inversion prediction method under relative wave impedance constraints; phase-controlled random inversion prediction method of seismic frequency extension processing; relaxation factor prestack seismic inversion prediction method; well-seismic combined forward simulation, etc. This type of method can make full use of seismic wave field information and logging information to reflect the vertical and horizontal distribution of underground reservoir sand bodies, but the superposition of seismic and logging responses of sandstone and mudstone in tight reservoirs leads to different degrees of difficulty in distinguishing sandstone and mudstone in the above methods.

[0005] At present, a new type of reservoir prediction method is based on big data analysis and artificial intelligence machine learning theory, with relatively scientific calculation and analysis logic. However, this type of operation is relatively difficult and has a high threshold. It requires a certain number of successful samples in the target work area. For oil and gas exploration areas with few successful samples or few wells and high costs, such as areas without exploration breakthroughs, areas with low exploration levels, and offshore oil and gas fields that pursue efficient development, it is difficult to use this method. Moreover, this method operates within its own logic, which is to some extent separated from the guidance of geological and oil and gas theory, resulting in a lack of a sound theoretical basis for its prediction process and results. Summary of the invention

[0006] In order to overcome the defects in the prior art, the present invention aims to provide a method for predicting high-quality dense sandstone reservoirs.

[0007] The present invention provides a technical solution to solve the above technical problems: a method for predicting high-quality dense sandstone reservoirs, comprising the following steps:

[0008] Step S10, obtaining seismic data, well logging data and core data of the target work area;

[0009] Step S20, obtaining a sedimentary microfacies plane prediction result map of the target work area according to seismic data, well logging data and core data;

[0010] Step S30, obtaining a plane prediction result map of an underwater distributary channel in the target work area according to the plane prediction result map of the sedimentary microfacies in the target work area;

[0011] Step S40, obtaining a rock phase plane prediction result map of the target work area according to the core data;

[0012] Step S50, obtaining a rock phase plane prediction result map above fine sand in the target work area and a rock phase plane prediction result map above medium sand according to the rock phase plane prediction result map;

[0013] Step S60, obtaining a plane prediction result map of the underwater distributary channel above the fine sand in the target work area according to the plane prediction result map of the underwater distributary channel in the target work area and the plane prediction result map of the rock phase above the fine sand in the target work area;

[0014] Step S70, obtaining a plane prediction result map of the diagenetic phase of the target work area and a plane prediction result map of the sandstone in the strong dissolution phase and above according to the seismic data, the well logging data and the core data;

[0015] Step S80, obtaining a plane prediction result map of high-quality reservoirs in the target work area according to the plane prediction result map of fine sand and above in the underwater distributary channel of the target work area and the plane prediction result map of medium sand and above sandstone in the strong dissolution phase.

[0016] A further technical solution is that the core data includes rock thin sections, pore structure, porosity, permeability, and particle size analysis test data.

[0017] A further technical solution is that the specific process of step S20 is:

[0018] Step S21, determining the sedimentary microfacies type based on the core-logging-seismic calibration results;

[0019] Step S22, based on the logging and seismic response characteristics of the sedimentary microfacies, the sedimentary microfacies are analyzed to determine the evolution and superposition relationship between different sedimentary microfacies in the direction of cutting or following the provenance;

[0020] Step S23: With the sedimentary microfacies profile analysis results as constraints, conduct a quantitative analysis of seismic attributes of well-seismic combination, optimize and extract planar seismic attributes, and then, under the guidance of the sedimentary microfacies planar distribution pattern, use the plane-profile multi-dimensional interactive method to conduct well-seismic-pattern fitting of the sedimentary microfacies, realize the planar characterization of the sedimentary microfacies, obtain the morphology, scale and planar combination of each sedimentary microfacies, and the longitudinal evolution relationship, and determine the sedimentary microfacies plane prediction result map of the target work area.

[0021] A further technical solution is that the sedimentary microfacies types in step S21 include underwater distributary channels, underwater distributary channel flanks, estuary sand bars, sheet sands, and distributary channel spaces.

[0022] A further technical solution is that the specific process of step S40 is:

[0023] Step S41, determining the rock phase type according to the core observation result and the particle size analysis result, wherein the rock phase type includes medium sandstone, fine sandstone, coarse sandstone, mudstone and conglomerate;

[0024] Step S42, constructing a rock phase logging interpretation model based on machine learning;

[0025] Step S43, performing intelligent interpretation of rock phase single well logging according to the rock phase logging interpretation model;

[0026] Step S44: obtaining a rock phase plane prediction result map of the target work area according to the rock phase single well logging intelligent interpretation result.

[0027] A further technical solution is that the specific process of step S50 is:

[0028] Step S51, constructing a dominant rock phase seismic interpretation model based on multi-seismic attribute fusion;

[0029] Step S52: According to the logging lithology interpretation results, the sand-to-ground ratio of fine sand and above sand bodies in the target layer of each well is calculated, and a variable database of the sand-to-ground ratio of fine sand and above sand bodies in the reservoir and seismic attributes is established by using SPSS software. Based on the Pearson correlation analysis of Bayesian statistical theory, the correlation coefficient between each seismic attribute of the target layer and the sand-to-ground ratio is obtained;

[0030] Step S53, according to the correlation coefficient, select seismic attributes to establish a univariate or multivariate linear regression fitting model of seismic attributes, and after error calculation, correlation optimization, and linear correction, determine the final fine sand and above sand body sand-to-ground ratio seismic attribute linear regression fitting model;

[0031] Step S54, obtaining a rock phase plane prediction result map above the fine sand in the target work area according to the linear regression fitting model of the seismic attribute of the fine sand and above sand body sand-to-ground ratio.

[0032] A further technical solution is that the specific process of step S70 is:

[0033] Step S71, determining the rock phase type according to the core data, wherein the rock phase type includes a strong dissolution phase, a strong calcareous cementation phase, a strong mud filling phase, a fracture phase, and a compaction phase;

[0034] Step S72, constructing a rock phase logging interpretation model based on machine learning;

[0035] Step S73, conducting intelligent interpretation of diagenetic phase single well logging according to the rock phase logging interpretation model;

[0036] Step S74, obtaining a diagenetic facies plane prediction result map of the target work area according to the diagenetic facies single well logging intelligent interpretation result;

[0037] Step S75, constructing a dominant diagenetic facies seismic interpretation model based on multi-seismic attribute fusion;

[0038] Step S76, using SPSS software to perform Pearson correlation analysis on sand-to-ground ratio and seismic attributes, and selecting appropriate seismic attributes to establish seismic attribute prediction models for sand bodies above strongly eroded medium sand;

[0039] Step S77, obtaining a plane prediction result map of the strongly eroded medium sand and above in the target work area according to the seismic attribute prediction model of the sand body above the strongly eroded medium sand.

[0040] The present invention has the following beneficial effects: the present invention combines well logging data with core observation, thin section identification, pore structure analysis, and physical property analysis; conducts sedimentary microfacies analysis based on well-seismic combination to optimize dominant sedimentary microfacies types; conducts dominant rock phase prediction based on sedimentary microfacies constraints; conducts dominant diagenetic phase prediction based on sedimentary microfacies-rock phase constraints; and performs three-phase coupling analysis to complete the prediction of favorable tight sandstone reservoirs. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 is a flow chart of the present invention;

[0042] Figure 2 This is a flow chart of well-seismic combined sand body multi-seismic attribute analysis based on PEARSON correlation analysis;

[0043] Figure 3 is the Pearson correlation coefficient graph;

[0044] Figure 4 It is the intersection relationship diagram of the multivariate model fitting results of fine sand and sandstone sand-to-ground ratio and the well logging interpretation results;

[0045] Figure 5This is the prediction result diagram of the dominant rock phase plane in the target work area;

[0046] Figure 6 This is the cross-correlation diagram of the multivariate model fitting results of the sand-to-ground ratio of the sandstone in the strong dissolution phase and above and the well logging interpretation results;

[0047] Figure 7 This is the seismic prediction result map of the plane distribution of high-quality reservoirs in the target work area. Implementation

[0048] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] like Figure 1 As shown, a method for predicting high-quality tight sandstone reservoirs of the present invention comprises the following steps:

[0050] Step S10, obtaining seismic data, well logging data and core data (rock thin sections, pore structure, porosity, permeability, particle size and other analytical data) of the target work area;

[0051] Step S20, obtaining a sedimentary microfacies plane prediction result map of the target work area according to seismic data, well logging data and core data;

[0052] Step S21, based on the core-logging-seismic calibration results, determining the sedimentary microfacies type, wherein the sedimentary microfacies type includes underwater distributary channel, underwater distributary channel flank, estuary sand bar, sheet sand, and distributary channel inter-channel;

[0053] Step S22, based on the logging and seismic response characteristics of the sedimentary microfacies, the sedimentary microfacies are analyzed to determine the evolution and superposition relationship between different sedimentary microfacies in the direction of cutting or following the provenance;

[0054] Step S23, using the sedimentary microfacies profile analysis results as constraints, conducting a quantitative analysis of seismic attributes of well-seismic combination, optimizing and extracting planar seismic attributes, and then under the guidance of the sedimentary microfacies plane distribution pattern, using the plane-profile multi-dimensional interactive method, conducting well-seismic-pattern fitting of the sedimentary microfacies, realizing the sedimentary microfacies plane characterization, obtaining the morphology, scale and plane combination of each sedimentary microfacies, and the longitudinal evolution relationship, and determining the sedimentary microfacies plane prediction result map of the target work area;

[0055] Step S30, obtaining a plane prediction result map of an underwater distributary channel in the target work area according to the plane prediction result map of the sedimentary microfacies in the target work area;

[0056] Step S40, obtaining a rock phase plane prediction result map of the target work area according to the core data;

[0057] Step S41, determining the rock phase type according to the core observation result and the particle size analysis result, wherein the rock phase type includes medium sandstone, fine sandstone, coarse sandstone, mudstone and conglomerate;

[0058] Step S42, constructing a rock phase logging interpretation model based on machine learning;

[0059] Step S43, performing intelligent interpretation of rock phase single well logging according to the rock phase logging interpretation model;

[0060] The logging parameters GR (natural gamma), AC (acoustic time difference), CNCF (neutron), and DEN (density) were selected, and a double-factor chart was established in pairs to make a preliminary identification and judgment of the lithology. It was found that GR and AC had good identification effects on fine sandstone, very fine sandstone, coarse siltstone, and mudstone, but had poor identification effects on fine-medium sandstone and medium-fine sandstone.

[0061] As the fine-medium sandstone and medium-fine sandstone cannot be effectively identified by GR-AC intersection, the stepwise discrimination method is adopted, and the logging parameters GR, AC, CNCF, and DEN are selected to construct a multivariate stepwise discriminant function.

[0062] According to SPSS discriminant analysis, the discriminant functions of fine-medium sandstone and medium-fine sandstone are as follows:

[0063] Medium-fine sandstone discriminant function:

[0064] F=0.094GR-0.151CNCF+0.223AC+28.087DEN-88.165>0.5

[0065] The discriminant function returned a judgment accuracy of 78.5%.

[0066] Fine-medium sandstone discriminant function:

[0067] F=0.094GR-0.151CNCF+0.223AC+28.087DEN-88.165<0.5

[0068] The discriminant function returned a judgment accuracy of 84.1%.

[0069] Step S44, obtaining a rock phase plane prediction result map of the target work area according to the rock phase single well logging intelligent interpretation result;

[0070] Step S50, obtaining a rock phase plane prediction result map of fine sand and above in the target work area according to the rock phase plane prediction result map;

[0071] Step S51, constructing a dominant rock phase seismic interpretation model based on multi-seismic attribute fusion;

[0072] Different seismic attributes extracted from seismic data can reflect the characteristics of sandstone from different aspects. We believe that there is a certain correlation between seismic attributes and reservoir sand-to-formation ratio. By statistically analyzing the attributes and sand-to-formation ratio and performing mathematical analysis, we finally established a model of seismic attribute fitting of reservoir sand-to-formation ratio to characterize the planar distribution of sandstone in the main layer.

[0073] Step S52: According to the logging lithology interpretation results, the sand-to-ground ratio of the target layer above fine sand in each well is calculated (Table 3), and a variable database of the sand-to-ground ratio of the reservoir above fine sand and seismic attributes is established using SPSS software. Based on the Pearson correlation analysis of Bayesian statistical theory, the correlation coefficient between each seismic attribute of the target layer and the sand-to-ground ratio is obtained (Table 1);

[0074] Table 1 Correlation coefficient between seismic attributes of target work area and sand-to-ground ratio above fine sand

[0075] minimum vibration Maximum vibration Half Energy Zero crossing number Average negative vibration Average instantaneous frequency r 0.825 -0.667 -0.604 -0.601 -0.708 -0.662

[0076] The correlation coefficient (r) ranges from -1 to 1: If |r| = 0, it means that the two variables have no linear correlation. If |r| = 1, it means that the two variables are completely linearly correlated. The direction of the linear correlation is indicated by the sign of the correlation coefficient, with a "+" sign indicating a positive correlation and a "-" sign indicating a negative correlation.

[0077] When 0.8<|r|≤1.0, it means that there is an extremely strong correlation between the seismic attributes and the sand-to-ground ratio; when 0.6<|r|≤0.8, it means that there is a strong correlation between the seismic attributes and the sand-to-ground ratio; when 0.4<|r|≤0.6, it means that there is a moderate correlation between the seismic attributes and the sand-to-ground ratio; when 0.2<|r|≤0.4, it means that there is a weak correlation between the seismic attributes and the sand-to-ground ratio; when |r|≤0.2, it means that there is an extremely weak correlation or no correlation between the seismic attributes and the sand-to-ground ratio.

[0078] Step S53, according to the level of correlation coefficient, seismic attributes are selected to establish a univariate or multivariate linear regression fitting model of seismic attributes, and after error calculation, correlation optimization, and linear correction, the final linear regression fitting model of the seismic attributes of the sand-to-ground ratio of fine sand and above sandstone and the linear regression fitting model of the seismic attributes of the sand-to-ground ratio of medium sand and above are determined;

[0079] Y = -2.232-1.61*average negative amplitude+5.27*minimum amplitude (R 2 =0.88)

[0080] The cross-comparison results of seismic attribute prediction and logging prediction are shown in Figure 4 .

[0081] Step S54, obtaining a rock phase plane prediction result diagram of fine sand and above sandstone in the target work area according to the linear regression fitting model of the seismic attribute of fine sand and above sandstone sand-to-ground ratio;

[0082] Step S60, obtaining a plane prediction result map of fine sand and above sand bodies in the underwater distributary channel of the target work area according to the plane prediction result map of the underwater distributary channel of the target work area and the plane prediction result map of the rock phase of fine sand and above in the target work area;

[0083] Step S70, obtaining a plane prediction result map of the diagenetic phase of the target work area and a plane prediction result map of the sandstone in the strong dissolution phase and above according to the seismic data, the well logging data and the core data;

[0084] Step S71, determining the rock phase type according to the core data, wherein the rock phase type includes a strong dissolution phase, a strong calcareous cementation phase, a strong mud filling phase, a fracture phase, and a compaction phase;

[0085] Step S72, constructing a rock phase logging interpretation model based on machine learning;

[0086] The random forest machine learning method is used in this lithofacies logging interpretation. Its principle is to use the existing core analysis and logging data as the input data of the classification model, analyze its data characteristics and potential through model training and learning, classify the diagenetic phases, and evaluate the accuracy of the classification results. Subsequently, based on the trained model, the logging data (without lithofacies interpretation) is used as input data for prediction and classification.

[0087] Random forest is a classifier that uses multiple trees to train and predict samples. M training sets are generated by random sampling with replacement. For each training set, a decision tree is constructed, and a part of the features is randomly extracted from the features to find the optimal solution for the node, split it, and finally vote through multiple decision trees to get the optimal classification.

[0088] According to the random forest machine learning method, by setting and adjusting the model parameters, the parameters ntree = 800, mtry = 2 were selected to establish the recognition model of diagenetic phase. Through cross-validation of sample data, it was found that this method had a good recognition effect, with an accuracy rate of more than 80.0%, and finally the diagenetic phase recognition model was established.

[0089] Step S73, conducting intelligent interpretation of diagenetic phase single well logging according to the rock phase logging interpretation model;

[0090] Step S74, obtaining a diagenetic facies plane prediction result map of the target work area according to the diagenetic facies single well logging intelligent interpretation result;

[0091] Step S75, constructing a dominant diagenetic facies seismic interpretation model based on multi-seismic attribute fusion;

[0092] Firstly, according to the interpretation results of lithofacies and diagenetic facies of single wells, the sand-to-ground ratio above fine sand in the strong dissolution phase and the sand-to-ground ratio above medium sand in the strong dissolution phase of the target layer were statistically analyzed (Table 2).

[0093] Table 2 The ratio of fine sand and medium sand in the strong dissolution phase in the target work area

[0094] Strongly eroded fine sand and above sand-to-ground ratio Strongly eroded medium sand and above sand-to-ground ratio Well A① 0.15 0.03 Well B① 0.10 0.03 Well A② 0.02 0 Well B② 0.09 0.04

[0095] Then, SPSS software was used to perform Pearson correlation analysis between sand-to-ground ratio and seismic attributes, and appropriate seismic attributes were selected to establish seismic attribute prediction models for sand bodies above strongly eroded fine sand and sand bodies above strongly eroded medium sand (Table 3).

[0096] Table 3 Correlation between the ratio of fine sand and sand to ground in the target area with strong dissolution phase and the ratio of medium sand and sand to ground and seismic attributes

[0097]

[0098] Seismic attribute prediction model for sand bodies above strongly dissolved fine sand:

[0099] Y = 1.729-1.506*sum of negative amplitudes-11.041*energy sum (R2 = 1)

[0100] Cross-comparison results of seismic attribute prediction results and well logging prediction results Figure 5 .

[0101] Seismic attribute prediction model for sand bodies above strongly dissolved medium sand:

[0102] Y=0.477-1.786*maximum amplitude+0.941*energy sum(R 2 =0.91)

[0103] Cross-comparison results of seismic attribute prediction results and well logging prediction results Figure 6 ;

[0104] Step S76, using SPSS software to perform Pearson correlation analysis on sand-to-ground ratio and seismic attributes, and selecting appropriate seismic attributes to establish seismic attribute prediction models for sand bodies above strongly eroded medium sand;

[0105] Step S77, obtaining a plane prediction result map of the target work area above the medium sand of the strongly eroded phase according to the seismic attribute prediction model of the sand body above the medium sand of the strongly eroded phase;

[0106] Step S80: Obtain a plane prediction result map of high-quality reservoirs in the target work area based on the plane prediction result map of fine sand in the underwater distributary channel and above and the plane prediction result map of medium sand and above sandstone in the strong dissolution phase (e.g. Figure 7 shown).

[0107] The above description is not intended to impose any form of limitation on the present invention. Although the present invention has been disclosed through the above embodiments, it is not intended to limit the present invention. Any technician familiar with the profession can make some changes or modifications to equivalent embodiments of equivalent changes using the technical contents disclosed above without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the scope of the technical solution of the present invention.

Claims

1. A method for predicting high-quality tight sandstone reservoirs, characterized in that: The following steps are involved: Step S10, obtaining seismic data, well logging data and core data of the target work area; Step S20, obtaining a sedimentary microfacies plane prediction result map of the target work area according to seismic data, well logging data and core data; Step S30, obtaining a plane prediction result map of an underwater distributary channel in the target work area according to a plane prediction result map of sedimentary microfacies in the target work area; Step S40, obtaining a rock phase plane prediction result map of the target work area according to the core data; Step S50, obtaining a rock phase plane prediction result map of fine sand and above in the target work area according to the rock phase plane prediction result map; Step S60, obtaining a plane prediction result map of fine sand and above sand bodies in the underwater distributary channel of the target work area according to the plane prediction result map of the underwater distributary channel of the target work area and the plane prediction result map of the rock phase of fine sand and above in the target work area; Step S70, obtaining a plane prediction result map of the diagenetic phase of the target work area and a plane prediction result map of the sandstone in the strong dissolution phase and above according to the seismic data, the well logging data and the core data; Step S80, obtaining a plane prediction result map of high-quality reservoirs in the target work area according to the plane prediction result map of fine sand and above in the underwater distributary channel of the target work area and the plane prediction result map of medium sand and above sandstone in the strong dissolution phase.

2. A method for predicting high-quality tight sandstone reservoirs according to claim 1, characterized in that: The core data include rock thin sections, pore structure, porosity, permeability, and particle size analysis data.

3. A method for predicting high-quality tight sandstone reservoirs according to claim 1, characterized in that: The specific process of step S20 is as follows: Step S21, determining the sedimentary microfacies type based on the core-logging-seismic calibration results; Step S22, based on the logging and seismic response characteristics of the sedimentary microfacies, the sedimentary microfacies are analyzed to determine the evolution and superposition relationship between different sedimentary microfacies in the direction of cutting or following the provenance; Step S23: With the sedimentary microfacies profile analysis results as constraints, conduct a quantitative analysis of seismic attributes of well-seismic combination, optimize and extract planar seismic attributes, and then, under the guidance of the sedimentary microfacies planar distribution pattern, use the plane-profile multi-dimensional interactive method to conduct well-seismic-pattern fitting of the sedimentary microfacies, realize the planar characterization of the sedimentary microfacies, obtain the morphology, scale and planar combination of each sedimentary microfacies, and the longitudinal evolution relationship, and determine the sedimentary microfacies plane prediction result map of the target work area.

4. A method for predicting high-quality tight sandstone reservoirs according to claim 3, characterized in that: The sedimentary microfacies types in step S21 include underwater distributary channels, underwater distributary channel flanks, estuary sand bars, sheet sands, and distributary channel spaces.

5. A method for predicting high-quality tight sandstone reservoirs according to claim 1, characterized in that: The specific process of step S40 is as follows: Step S41, determining the rock phase type according to the core observation result and the particle size analysis result, wherein the rock phase type includes medium sandstone, fine sandstone, coarse sandstone, mudstone and conglomerate; Step S42, constructing a rock phase logging interpretation model based on machine learning; Step S43, performing intelligent interpretation of rock phase single well logging according to the rock phase logging interpretation model; Step S44: obtaining a rock phase plane prediction result map of the target work area according to the rock phase single well logging intelligent interpretation result.

6. A method for predicting high-quality tight sandstone reservoirs according to claim 5, characterized in that: The specific process of step S50 is as follows: Step S51, constructing a dominant rock phase seismic interpretation model based on multi-seismic attribute fusion; Step S52: According to the logging lithology interpretation results, the sand-to-ground ratio of fine sand and above sand bodies in the target layer of each well is calculated, and a variable database of the sand-to-ground ratio of fine sand and above sand bodies in the reservoir and seismic attributes is established by using SPSS software. Based on the Pearson correlation analysis of Bayesian statistical theory, the correlation coefficient between each seismic attribute of the target layer and the sand-to-ground ratio is obtained; Step S53, according to the correlation coefficient, select seismic attributes to establish a univariate or multivariate linear regression fitting model of seismic attributes, and after error calculation, correlation optimization, and linear correction, determine the final fine sand and above sand body sand-to-ground ratio seismic attribute linear regression fitting model; Step S54, obtaining a rock phase plane prediction result map above the fine sand in the target work area according to the linear regression fitting model of the seismic attribute of the fine sand and above sand body sand-to-ground ratio.

7. A method for predicting high-quality tight sandstone reservoirs according to claim 1, characterized in that: The specific process of step S70 is as follows: Step S71, determining the rock phase type according to the core data, wherein the rock phase type includes a strong dissolution phase, a strong calcareous cementation phase, a strong mud filling phase, a fracture phase, and a compaction phase; Step S72, constructing a rock phase logging interpretation model based on machine learning; Step S73, conducting intelligent interpretation of diagenetic phase single well logging according to the rock phase logging interpretation model; Step S74, obtaining a diagenetic facies plane prediction result map of the target work area according to the diagenetic facies single well logging intelligent interpretation result; Step S75, constructing a dominant diagenetic facies seismic interpretation model based on multi-seismic attribute fusion; Step S76, using SPSS software to perform Pearson correlation analysis on sand-to-ground ratio and seismic attributes, and selecting appropriate seismic attributes to establish seismic attribute prediction models for sand bodies above strongly eroded medium sand; Step S77, obtaining a plane prediction result map of the strongly eroded medium sand and above in the target work area according to the seismic attribute prediction model of the sand body above the strongly eroded medium sand.

Images