Unused reserve block classification method based on deep learning
By establishing nonlinear relationships among untapped reserve blocks using deep learning algorithms and classifying them using deep neural network models, the problems of long evaluation cycles and significant human influence in existing technologies are solved, enabling rapid and accurate evaluation of untapped reserve blocks.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-04
- Publication Date
- 2026-03-10
AI Technical Summary
Existing methods for classifying and evaluating untapped reserve blocks are complex, have long evaluation cycles, involve numerous parameters and manually assign weights, resulting in significant human influence on the evaluation results and hindering rapid decision-making.
Deep learning algorithms are used to determine the nonlinear relationship between reservoir parameters and development effects in developed areas. A deep learning training set is established, and untapped reserve blocks are classified and evaluated using a deep neural network model. Parameters with low correlation are removed, sample augmentation is performed, and the optimal model is constructed.
It enables rapid and accurate classification and evaluation of untapped reserve blocks, reduces the impact of human factors, and improves the objectivity and accuracy of the evaluation.
Smart Images
Figure CN121637138A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present disclosure relates to the field of oil and gas reservoir geology, and in particular to a method for grading and classifying unproduced reserve blocks using a deep neural network. BACKGROUND
[0002] The statements in this section merely provide background information related to the present disclosure and do not constitute prior art.
[0003] Reserve evaluation is a long-term work throughout the whole process of oil and gas exploration and development. With the changes in geological, engineering data and technical and economic conditions, it is necessary to evaluate the reserves in stages in a timely manner. As oilfield development enters the middle and later stages, the scale of high-quality reserves is continuously reduced, and the quality of remaining reserves is gradually deteriorated. In the reserve evaluation, it is very important to reasonably grade and classify the remaining unproduced reserve blocks and develop targeted countermeasures for unproduced reserve blocks of different grades and types according to their characteristics to improve the reserve producing degree.
[0004] The unproduced reserve blocks are controlled by multiple factors, and the reservoir geological conditions are complex, mainly reflected in the complex oil-water distribution, the absence of a unified oil-water interface, the unstable oil layer thickness distribution, the obvious difference in effective thickness between wells, and the large difference in productivity. At present, the evaluation of unproduced reserve blocks is based on the consideration of dynamic and static parameters, the classification evaluation method is adopted, and the classification standard of unproduced reserve blocks is established.
[0005] The conventional classification and evaluation method of unproduced reserve blocks first establishes an evaluation parameter set that reflects the good or bad of some characteristics of unproduced reserve blocks from different angles and aspects. These parameters must be combined with the characteristics of oilfield development to correctly evaluate the quality of the reserves. The classification parameters are required to have certain physical meaning and be able to reflect the characteristics of the reserves in a certain aspect. On the basis of the determination of the classification parameters, the classification parameters are optimized, the purpose is to make the classification parameters as comprehensive as possible and eliminate those parameters with low correlation. The optimized classification parameters should be able to comprehensively reflect the characteristics of the reserves in a certain aspect, and each parameter should reflect the good or bad of the reserves in a certain aspect, and multiple parameters should not repeatedly reflect the properties of the reserves in the same aspect. After the optimization of the classification parameters, different parameter weights are artificially assigned to the optimized classification parameters, the product of the evaluation parameters and the evaluation parameter weights of the unproduced reserve blocks is obtained, and the unproduced reserve block score value is obtained. Finally, the unproduced reserve blocks are graded and classified according to the unproduced reserve block score value, and the classification and evaluation of the unproduced reserve blocks are completed.
[0006] However, the classification and evaluation method of unproduced reserve blocks is complicated, requires many parameters, and has a long evaluation period, which cannot realize rapid evaluation and decision-making. In addition, the weights of the evaluation parameters are often artificially assigned, and the evaluation results are greatly affected by human factors.
[0007] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art. Summary of the Invention
[0008] In view of this, this disclosure provides a deep learning-based method for classifying unused reserve blocks, which solves the problems of long classification and evaluation cycles, inability to achieve rapid evaluation and decision-making, and the use of manual assignment of evaluation parameter weights, resulting in significant human influence on the evaluation results.
[0009] The technical concept of the deep learning-based classification method for untapped reserve blocks in this invention is as follows: based on deep learning algorithms, the nonlinear relationship between reservoir parameters and development effects in developed areas is determined, and then the learned complex nonlinear relationship is used to classify and evaluate untapped reserve blocks.
[0010] Based on the technical concept of this invention, and to achieve the above-mentioned objective, the deep learning-based method for classifying unused reserve blocks includes:
[0011] Using deep learning algorithms, the nonlinear relationship between reservoir parameters and development results in developed areas was determined.
[0012] The nonlinear relationship is used to classify and evaluate untapped reserve blocks.
[0013] In this disclosure and possible embodiments, the method for classifying and evaluating untapped reserve blocks using the nonlinear relationship includes:
[0014] Identify the reservoir parameter set for untapped reserve blocks;
[0015] Using the aforementioned nonlinear relationship and combining it with the reservoir parameter set of the untapped reserve blocks, the untapped reserve blocks are classified and evaluated.
[0016] In this disclosure and possible embodiments, a method for determining the nonlinear relationship between reservoir parameters and development effects in a developed area using a deep learning algorithm includes:
[0017] A deep learning training set for the developed area is established using reservoir parameters and development results. The classification rules of the deep learning training set for the developed area are learned using a deep learning algorithm, and the nonlinear relationship between reservoir parameters and development results in the developed area is obtained.
[0018] In this disclosure and possible embodiments, the method for establishing a deep learning training set for a developed area using reservoir parameters and development results includes:
[0019] Using the reservoir parameters of the developed area as deep learning training data and the development effect of the developed area as labels, a deep learning training set for the developed area is constructed.
[0020] In this disclosure and possible embodiments, the method for learning classification rules of the developed area deep learning training set using a deep learning algorithm includes:
[0021] Expand the deep learning training set of the developed area;
[0022] A deep neural network is established, and the deep neural network is trained and optimized using an expanded training dataset to construct a deep learning model. The nonlinear relationship between reservoir parameters and development effects in the developed area is determined through the deep learning model.
[0023] In this disclosure and possible embodiments, the method for expanding the deep learning training set of the developed area includes:
[0024] The deep learning training set data of the developed area is subjected to parameter optimization, sorting and sample augmentation processing.
[0025] The sample enhancement process involves performing principal component analysis on the optimized and sorted training set data, and sorting the samples from high to low based on the principal component analysis results.
[0026] Based on the sorted training set, a two-dimensional sample-parameter matrix is constructed. The principal and secondary range parameters of the two-dimensional sample-parameter matrix are analyzed using the geostatistical variogram analysis method. Based on the principal and secondary range parameters, the two-dimensional sample-parameter matrix is expanded by interpolation using Kriging interpolation.
[0027] In this disclosure and possible embodiments, the method for parameter optimization and sorting of the deep learning training set data of the developed area includes:
[0028] The parameters are optimized based on the Pearson correlation coefficient matrix between the deep learning training data in the developed areas, and then sorted in descending order of their correlation.
[0029] In this disclosure and possible embodiments, the method for establishing a reservoir parameter set for untapped reserve blocks includes:
[0030] We statistically analyze and calculate as many reservoir parameters as possible, including reservoir structure, dynamics, and fault block characteristics, of untapped reserve blocks, to form a parameter set for untapped reserve blocks.
[0031] In this disclosure and possible embodiments, the method for determining the reservoir parameters of the developed area includes:
[0032] Statistically analyze and calculate as many reservoir parameters as possible that reflect the reservoir, microstructure, dynamics, and fault block characteristics of the developed area.
[0033] In this disclosure and possible embodiments, the method for determining the development effect of the developed area includes:
[0034] Based on the development dynamic parameters of the developed reservoirs, the reservoir quality of the developed areas is evaluated and classified according to the evaluation results. The classification results represent the development effect of the developed reservoirs.
[0035] The beneficial effects of this invention are as follows:
[0036] The disclosed deep learning-based classification method for untapped reserve blocks first establishes a deep learning training set for developed areas using reservoir parameters and development results. Then, it uses a deep learning algorithm to learn the classification rules of the developed area training set. Finally, it performs rapid classification and evaluation of untapped reserve blocks. On the one hand, it utilizes the deep learning algorithm to fully learn the nonlinear relationship between reservoir parameters and development results in developed areas, achieving rapid classification and evaluation of untapped reserve blocks. On the other hand, it eliminates the need for manual assignment, ensuring that the evaluation results are not affected by human factors, and making the classification and evaluation results more accurate and reliable. Attached Figure Description
[0037] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the specification, serve to illustrate the technical solutions of this disclosure.
[0038] Figure 1 This is a flowchart of the deep learning-based method for classifying unused reserve blocks according to an embodiment of this disclosure;
[0039] Figure 2 This is an isomorphic map of the training set data after data augmentation according to an embodiment of this disclosure;
[0040] Figure 3 This is a diagram illustrating the optimal deep neural network model structure of this embodiment.
[0041] Figure 4 This is a diagram showing the training effect of a deep neural network according to an embodiment of this disclosure. Detailed Implementation
[0042] The present disclosure is described below based on embodiments; however, it is worth noting that the present disclosure is not limited to these embodiments. In the detailed description of the present disclosure below, certain specific details are described in detail. However, those skilled in the art will fully understand the present disclosure for the parts not described in detail.
[0043] Furthermore, unless the context explicitly requires it, the words "comprising," "including," and similar terms throughout the specification and claims should be interpreted as including rather than exclusive or exhaustive; that is, meaning "including but not limited to."
[0044] Current methods for classifying and evaluating untapped reserves are complex, requiring numerous parameters and lengthy evaluation cycles, hindering rapid evaluation and decision-making. Furthermore, the weights of evaluation parameters are often manually assigned, leading to significant human influence on the evaluation results. To address these issues, this invention utilizes deep learning algorithms to automatically learn the complex nonlinear relationships between reservoir parameters and development effectiveness in developed areas. These learned nonlinear relationships are then used to classify and evaluate untapped reserves, ultimately yielding more objective and accurate classification and evaluation results.
[0045] This disclosure discloses a deep learning-based method for classifying untapped reserve blocks, which can be implemented through the following technical solution. The specific steps of the solution are as follows: Step S10: Establish a deep learning training set using reservoir parameters and development effects of developed areas; Step S20: Perform parameter optimization, sorting, and enhancement processing on the deep learning training set data to obtain an expanded learning training set; Step S30: Establish a deep learning model; Step S40: Establish a reservoir parameter set for untapped reserve blocks; Step S50: Use the deep learning model, combined with the reservoir parameter set for untapped reserve blocks, to classify and evaluate the untapped reserve blocks.
[0046] Step S10: Using the reservoir parameters and development results of the developed area, establish a deep learning training set, as follows:
[0047] 1. Determine reservoir parameters that reflect the characteristics of the developed area:
[0048] Following conventional techniques in this field, select as many reservoir parameters as possible, regardless of their distribution range. These parameters can be continuous reservoir parameters, such as burial depth, or discrete reservoir parameters, such as sedimentary facies type. Strive to statistically analyze and calculate multiple reservoir parameters reflecting the reservoir, microstructure, dynamics, and fault-block characteristics of the developed area.
[0049] 2. Determine the reservoir quality classification results in the developed area:
[0050] The main purpose is to evaluate the quality of developed reservoirs based on dynamic development parameters such as initial production, production decline, and water cut, and to classify them according to the evaluation results. The classification results should represent the quality of the development of the developed reservoirs.
[0051] 3. Establish a deep learning training set:
[0052] Using multiple reservoir parameters reflecting the reservoir, microstructure, dynamics, and fault block characteristics of developed areas as deep learning training data, and using the reservoir quality classification results of developed areas as labels, a deep learning training set is constructed.
[0053] Step S20: Perform parameter optimization, sorting, and sample augmentation on the deep learning training set data to obtain an expanded training dataset, as detailed below:
[0054] 1. Optimize and sort parameters in the deep learning training data:
[0055] Parameters are optimized based on the correlation between data in the deep learning training set. The goal of this optimization is to make the classification as comprehensive as possible and to eliminate parameters that are closely correlated, thereby improving the applicability and stability of the model. The optimized classification parameters should comprehensively reflect the characteristics of the reserves in a specific aspect, with each parameter reflecting the quality of the reserves in that particular aspect, and multiple parameters should not repeatedly reflect the same property of the reserves.
[0056] Finally, the training set data is sorted from high to low based on the correlation of the parameters.
[0057] 2. Perform sample augmentation on the training set:
[0058] The optimized and sorted training set is augmented to increase the number of training set samples to meet the requirements of deep learning for the number of samples.
[0059] Before data augmentation, principal component analysis (after normalization) was performed on the training set data. Based on the results of the principal component analysis, the samples were sorted from high to low. Then, the sorted training set was augmented by constructing a two-dimensional sample-parameter matrix. Using geostatistical variogram analysis, the principal and secondary range parameters of the parameter matrix were first analyzed. Then, based on the principal and secondary range parameters, Kriging interpolation was used to interpolate and expand the two-dimensional sample-parameter matrix.
[0060] Step 30: Build a deep learning model, as follows:
[0061] 1. Establish a deep neural network:
[0062] According to the conventional techniques in this field, the main focus is on defining the hyperparameters of the deep neural network model (the number of hidden layers, the number of neurons in each layer, the activation function of each layer, etc.).
[0063] 2. Train and optimize the deep neural network to obtain the optimal deep neural network model:
[0064] By using expanded learning training set data to train a deep neural network, and by selecting appropriate epochs, validation percentages, optimization algorithms, and other parameters, the optimal model parameters (weights of each neuron) are obtained, and the optimal deep neural network model is established, which is the deep learning model.
[0065] Step S40: Establish the reservoir parameter set for the untapped reserve block, as follows:
[0066] Statistical analysis and calculation of reservoir parameters such as reservoir structure, microstructure, dynamics, and fault block characteristics of untapped reserve blocks are conducted to establish a parameter set for untapped reserve blocks.
[0067] Step S50: Using a deep learning model and combining the reservoir parameter set of untapped reserve blocks, classify and evaluate the untapped reserve blocks.
[0068] Using a deep learning model (optimal deep neural network model) and combining it with the reservoir parameter set of undeveloped reserve blocks, we evaluate and classify undeveloped reserve blocks in undeveloped areas and give the type of each undeveloped area.
[0069] Example
[0070] The following example, using a specific oil reservoir, demonstrates the implementation of the deep learning-based classification method for untapped reserve blocks disclosed in this invention, thereby proving the effectiveness of the classification method. Figure 1 As shown, the specific implementation steps are as follows:
[0071] Step 1: Select reservoir parameters for the developed area:
[0072] Choose as many reservoir parameters as possible, without considering the distribution range of the parameters. These can be continuous reservoir parameters, such as burial depth, or discrete reservoir parameters, such as sedimentary facies type.
[0073] Try to statistically analyze and calculate multiple reservoir parameters that reflect the reservoir, microstructure, dynamics, and fault block characteristics of the developed area.
[0074] In this example, 23 reservoir parameters were selected, namely, reservoir thickness, permeability, porosity, reservoir depth, reserve abundance, oil-bearing area, initial production, production ratio, cumulative oil production, production rate, recovery degree, source-reservoir connection, formation dip angle, fault density, oil column height, pore radius, displacement pressure, and median radius pores.
[0075] Step 2: Evaluate and classify the oil reservoirs in the developed areas:
[0076] The main purpose is to evaluate the quality of developed reservoirs based on dynamic development parameters such as initial production, production decline, and water cut, and to classify them according to the evaluation results. The classification results should represent the quality of the development of the developed reservoirs.
[0077] In this example, the 51 developed areas are divided into three types based on their development results: Type I, Type II, and Type III. Type I has the best development results, Type II has slightly worse results, and Type III has the worst results.
[0078] Step 3: Create a deep learning training set:
[0079] Using the 23 reservoir parameters selected in step 1 that reflect the reservoir, microstructure, dynamics, and fault block characteristics of the developed area as deep learning training data, and using the reservoir classification results of the developed area as labels, a deep learning training set is constructed.
[0080] Therefore, in this example, the constructed deep learning training set includes 23 reservoir parameters, and the number of training set samples is 51 (the number of developed areas).
[0081] Step 4: Optimize and sort the parameters of the deep learning training set:
[0082] Based on the correlation between the deep learning training set data in step 3, the training set parameters are optimized. The purpose of optimization is to make the classification as comprehensive as possible and to remove those parameters that are closely related, which can improve the applicability and stability of the model.
[0083] In this example, the parameters are optimized based on the Pearson correlation coefficient matrix between them.
[0084] After optimization, 15 parameters were retained, and 8 parameters with the worst correlation were removed. The correlation analysis of the 15 optimized parameters is shown in Table 1:
[0085] Table 1 shows the correlation of the 15 parameters after optimization.
[0086]
[0087] Then, the training set data are sorted from high to low according to the correlation of the parameters in Table 1.
[0088] In this example, among the 15 parameters retained, the most correlated are crude oil geological reserves and calculated area, with a Pearson correlation of 0.9.
[0089] Step 5: Perform sample augmentation on the training set:
[0090] The optimized and sorted training set is augmented to increase the number of training set samples, resulting in an expanded training dataset to meet the sample quantity requirements of deep learning.
[0091] Before expansion, the training set is normalized, then principal component analysis is performed, and the samples are sorted from high to low based on the results of the principal component analysis.
[0092] In this example, after performing principal component analysis on 51 samples, the three principal components can represent 96% of the data information. The scores of each sample are calculated using these three principal components. The 51 samples are then sorted from high to low according to the principal component calculation formula, and the sorted training set is then expanded.
[0093] The expansion method involves constructing a two-dimensional sample-parameter matrix, using geostatistical variogram analysis to first analyze the primary and secondary ranges of the parameter matrix, and then using Kriging interpolation to expand the two-dimensional sample-parameter matrix based on the primary and secondary ranges.
[0094] In this example, kriging interpolation data augmentation is performed on a 51*15 matrix, such as... Figure 2 As shown, the enhanced matrix is 765*15, and the training dataset has been expanded by 15 times.
[0095] Step 6: Build a deep learning model:
[0096] According to the conventional techniques in this field, the main focus is on defining the hyperparameters of the deep neural network model, such as the number of hidden layers, the number of neurons in each layer, and the activation function of each layer.
[0097] In this example, the deep neural network is defined with 15 neurons in the input layer, 1 neuron in the output layer, and 5 hidden layers. The number of neurons in each hidden layer is 64, 32, 32, 16, and 8, respectively. Figure 4 The output layer has 3 neurons (corresponding to three categories), such as... Figure 3 As shown in the figure (question marks in the figure indicate the number of input samples, which is determined by the number of input data samples); the 5 hidden layers use the ReLU activation function, and the output layer uses the softmax activation function; no regularization or dropout processing is used.
[0098] Step 7: Deep learning model training:
[0099] Using the expanded two-dimensional sample-parameter matrix as the training set, a deep neural network is trained. Appropriate parameters such as the number of epochs, validation percentage, and optimization algorithm are selected. Figure 3 As shown, the optimal model parameters (weights of each neuron) are obtained.
[0100] In this example, an expanded training dataset (the 765*15 matrix dataset augmented in step 5) is used for training, with 20 training epochs, a validation percentage of 10%, and the optimization algorithm is stochastic gradient descent (SGD), with the loss function being cross-entropy loss. Figure 4 As shown in the results, the accuracy of the training set (solid black line) can reach over 0.995, and the accuracy of the validation set (dashed black line) can also reach over 0.985 after 10 epochs, indicating that the deep neural network training effect is good.
[0101] Step 8: Calculation of reservoir parameters in undeveloped areas:
[0102] Statistical analysis and calculation of reservoir parameters such as reservoir structure, microstructure, dynamics, and fault block characteristics of undeveloped reserve blocks are conducted to establish a reservoir parameter set for undeveloped reserve blocks in undeveloped areas.
[0103] In this example, for the 25 undeveloped reservoir blocks that need to be classified, 15 reservoir parameters selected in step 4 are chosen to establish a reservoir parameter set for the untapped reserves blocks with 25 samples and 15 reservoir parameters.
[0104] Step 9: Evaluate and classify undeveloped and unused reserve blocks:
[0105] Using a trained deep learning model and combining it with the reservoir parameter set of undeveloped and untapped reserve blocks, the undeveloped and untapped reserve blocks are evaluated and classified, and the classification type of each untapped reserve block is given.
[0106] In this example, 25 untapped reserve blocks were classified. The calculation method used the reservoir parameter set of 25 samples and 15 parameters established in step 8 as input data, and the optimized deep neural network in step 7 was used to process the input data to obtain the classification results of the 25 blocks, as shown in Table 2:
[0107] Table 2. Classification Results of Reserves in Undeveloped Areas
[0108]
[0109] As shown in Table 2, there are 4 Class I blocks, accounting for 16% of the total. There are 13 Class II blocks, accounting for 52% of the total. There are 8 Class III blocks, accounting for 32% of the total. The untapped reserves are mainly Class II and Class III blocks with poor reservoir quality. Based on the classification results, 7 horizontal wells were deployed in the 4 Class I potential areas, producing 29 tons of oil per day, confirming the accuracy and reliability of the evaluation results.
[0110] The various embodiments of this disclosure have been described above. These descriptions are exemplary and not exhaustive, and are not limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical applications, or technical improvements to the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.
Claims
1. A deep learning-based unexploited reserve block classification method, characterized in that, include: Using deep learning algorithms, the nonlinear relationship between reservoir parameters and development results in developed areas was determined. The nonlinear relationship is used to classify and evaluate untapped reserve blocks.
2. The unreserved resources block classification method of claim 1, wherein, The method for classifying and evaluating untapped reserve blocks using the aforementioned nonlinear relationship includes: Identify the reservoir parameter set for untapped reserve blocks; Using the aforementioned nonlinear relationship and combining it with the reservoir parameter set of the untapped reserve blocks, the untapped reserve blocks are classified and evaluated.
3. The unreserved area block classification method of Claim 1 or 2, wherein, Methods for determining the nonlinear relationship between reservoir parameters and development effects in developed areas using deep learning algorithms include: A deep learning training set for the developed area is established using reservoir parameters and development results. The classification rules of the deep learning training set for the developed area are learned using a deep learning algorithm, and the nonlinear relationship between reservoir parameters and development results in the developed area is obtained.
4. The unreserved resources block classification method of claim 3, wherein, The method for establishing a deep learning training set for developed areas using reservoir parameters and development results includes: Using the reservoir parameters of the developed area as deep learning training data and the development effect of the developed area as a label, a deep learning training set for the developed area is constructed.
5. The unreserved resources block classifying method according to claim 3, wherein, The method for learning classification rules for the deep learning training set of the developed area using a deep learning algorithm includes: Expand the deep learning training set of the developed area; A deep neural network is established, and the deep neural network is trained and optimized using an expanded training dataset to construct a deep learning model. The nonlinear relationship between reservoir parameters and development effects in the developed area is determined through the deep learning model.
6. The unreserved area block classification method of claim 5, wherein, The method for expanding the deep learning training set of the developed area includes: The deep learning training set data of the developed area is subjected to parameter optimization, sorting and sample augmentation processing. The sample enhancement process involves performing principal component analysis on the optimized and sorted training set data, and sorting the samples from high to low based on the principal component analysis results. Based on the sorted training set, a two-dimensional sample-parameter matrix is constructed. The principal and secondary range parameters of the two-dimensional sample-parameter matrix are analyzed using the geostatistical variogram analysis method. Based on the principal and secondary range parameters, the two-dimensional sample-parameter matrix is expanded by interpolation using Kriging interpolation.
7. The unreserved resources block classifying method according to claim 6, characterized in that, The method for parameter optimization and sorting of the deep learning training set data in the developed area includes: The parameters are optimized based on the Pearson correlation coefficient matrix between the deep learning training data in the developed areas, and then sorted in descending order of correlation.
8. The unreserved resources block classifying method according to claim 2, wherein, The method for establishing a reservoir parameter set for untapped reserve blocks includes: We statistically analyze and calculate as many reservoir parameters as possible, including reservoir structure, dynamics, and fault block characteristics, of untapped reserve blocks, to form a parameter set for untapped reserve blocks.
9. The undeveloped reserve block classification method of Claim 1, wherein, The method for determining the reservoir parameters in the developed area includes: Statistically analyze and calculate as many reservoir parameters as possible that reflect the reservoir, microstructure, dynamics, and fault block characteristics of the developed area.
10. The uneconomic reserves block classification method of claim 1, wherein, The method for determining the development effect of the developed area includes: Based on the development dynamic parameters of the developed reservoirs, the reservoir quality of the developed areas is evaluated and classified according to the evaluation results. The classification results represent the development effect of the developed reservoirs.