Hydrocarbon reservoir data processing method, apparatus and device

Abstract

The embodiment of the present application provides a kind of oil and gas reservoir data processing method, device and equipment.The method comprises: determining the target parameter related to the oil and gas reservoir mechanism;Potential evaluation matrix is constructed based on the target parameter, and the weight of target parameter is determined according to potential evaluation matrix;The parameter value of the target parameter of each reservoir is obtained, and the potential coefficient of each reservoir is determined based on the parameter value and weight;Determine the potential layer of oil and gas reservoir according to the potential coefficient of each reservoir of multiple reservoirs.The method is used to achieve the effect of accurately identifying the potential layer of oil and gas reservoir.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence, and in particular to a method, apparatus and equipment for processing oil and gas reservoir data. Background Technology

[0002] As exploration of major oil and gas basins deepens, the main targets for oil and gas field exploration and development are low-porosity and low-permeability clastic rocks, fractured-vuggy carbonate rocks, low-saturation and unconventional (tight oil and gas, shale gas, shale oil, etc.) complex oil and gas reservoirs. These complex oil and gas reservoirs have complex lithology, complex pore structure, and complex oil-water relationships. In order to identify more valuable reservoirs for development from these complex oil and gas reservoirs, it is necessary to conduct a comprehensive evaluation of multiple reservoirs within the reservoirs.

[0003] Existing technologies use two-dimensional fluid identification charts to evaluate the reservoir fluid properties of oil and gas reservoirs. Typically, for low-saturation oil and gas reservoirs, two-dimensional fluid identification charts are drawn using the acoustic transit time and resistivity information of the reservoir to identify oil layers, oil-water co-layers, oil-water-bearing layers, water layers, dry layers, etc.

[0004] The technical approach of using two-dimensional fluid identification charts for reservoir logging interpretation uses reservoir data with fewer dimensions, making it impossible to determine the effective oil-water boundary. This results in low accuracy in fluid property identification and inaccurate acquisition of potential layers. Summary of the Invention

[0005] The oil and gas reservoir data processing method, apparatus, and equipment provided in this application are used to improve the accuracy of reservoir potential layer prediction.

[0006] In a first aspect, embodiments of this application provide a method for processing oil and gas reservoir data, wherein the oil and gas reservoir includes multiple reservoirs, and the method includes:

[0007] Determine the target parameters related to the genesis mechanism of hydrocarbon reservoirs;

[0008] A potential evaluation matrix is ​​constructed based on the target parameters, and the weights of the target parameters are determined based on the potential evaluation matrix.

[0009] Obtain the parameter values ​​of the target parameters for each reservoir, and determine the potential coefficient of each reservoir based on the parameter values ​​and the weights;

[0010] The potential oil and gas reservoirs are determined based on the potential coefficients of each of the multiple reservoirs.

[0011] In one possible implementation, the target parameters include one or more of the following: mud content, oil volume, water content, and production capacity.

[0012] In one possible implementation, the method further includes:

[0013] Based on a pre-trained fluid category recognition model, the parameter values ​​of the target parameters of each of the multiple reservoirs are processed to obtain the fluid category of each reservoir.

[0014] In one possible implementation, the training operation of the fluid category recognition model includes:

[0015] Construct a training sample set. Each training sample in the training sample set includes an input-output pair. The input includes the sample parameter values ​​of the target parameters of the sample reservoir, and the output includes the fluid category corresponding to the sample reservoir.

[0016] The Long Short-Term Memory Network was trained using the training sample set to obtain the first training result.

[0017] The convolutional neural network model is trained using the training sample set to obtain the second training result.

[0018] By combining the first training result and the second training result, a combined training result is obtained.

[0019] In one possible implementation, the training operation of the fluid category recognition model further includes:

[0020] For each training sample in the training set, the model parameters are adjusted based on the loss value obtained from the cross-entropy loss function to minimize the loss value between the predicted value and the true value output by the adjusted model.

[0021] In one possible implementation, constructing a potential evaluation matrix based on the target parameters and determining the weights of the target parameters according to the potential evaluation matrix includes:

[0022] To determine the importance of each target parameter relative to that target parameter and relative to any other target parameter when determining the potential coefficient of each reservoir;

[0023] Based on the importance of each of the multiple target parameters, the potential evaluation matrix is ​​constructed;

[0024] The weights of the mud content, oil volume, water content, and production capacity are determined using the analytic hierarchy process (AHP) based on the potential evaluation matrix.

[0025] In one possible implementation, obtaining the parameter value of the target parameter for each reservoir includes:

[0026] Determine the computational model for obtaining the target parameter values;

[0027] For each reservoir, the parameter value of the target parameter is determined using the calculation model based on the well logging data of that reservoir.

[0028] In one possible implementation, determining the potential oil and gas reservoir based on the potential coefficients of each of the plurality of reservoirs includes:

[0029] The reservoirs are grouped according to their respective fluid categories, resulting in two or more grouping results; each grouping result includes at least one reservoir included in at least one fluid category;

[0030] For each grouping result, at least one reservoir in the grouping result is sorted according to its potential coefficient to obtain the sorting result corresponding to the grouping result;

[0031] Based on the sorting results corresponding to the two or more grouping results, the potential oil and gas reservoir layers are determined.

[0032] Secondly, embodiments of this application provide an oil and gas reservoir data processing apparatus, wherein the oil and gas reservoir includes multiple reservoirs, and the apparatus includes:

[0033] The parameter determination unit is used to determine target parameters related to the genesis mechanism of hydrocarbon reservoirs;

[0034] A data processing unit is used to construct a potential evaluation matrix based on the target parameters and determine the weights of the target parameters according to the potential evaluation matrix.

[0035] Obtain the parameter values ​​of the target parameters for each reservoir, and determine the potential coefficient of each reservoir based on the parameter values ​​and the weights;

[0036] A potential layer determination unit is used to determine the potential layers of oil and gas reservoirs based on the potential coefficients of the plurality of reservoirs.

[0037] Thirdly, embodiments of this application provide an electronic device, including: a memory and a processor;

[0038] The memory stores computer-executed instructions;

[0039] The processor executes computer execution instructions stored in the memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.

[0040] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible implementations of the first aspect.

[0041] The oil and gas reservoir data processing method, apparatus, and equipment provided in this application analyze the genetic mechanism of oil and gas reservoirs to determine target parameters for evaluation; a potential evaluation matrix is ​​constructed using the target parameters, and the weights of the target parameters are derived from the potential evaluation matrix; then, based on the target parameter values ​​and weights of each reservoir, the potential coefficients of each reservoir are determined, and the potential layers of the oil and gas reservoir are identified through the potential coefficients of each reservoir. Because the target parameters used are related to the genetic mechanism of complex oil and gas reservoirs, the determined potential coefficients fully characterize the potential value of the reservoirs. Therefore, the method provided in this application provides a more accurate interpretation of the reservoir data and obtains more accurate oil and gas potential layers. Attached Figure Description

[0042] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0043] Figure 1 A flowchart illustrating the oil and gas reservoir data processing method provided in this application;

[0044] Figure 2 This is a partial statistical table of experimental data for fluid category recognition on the validation dataset;

[0045] Figure 3 This is a partial statistical table of experimental data for fluid category recognition on the test dataset;

[0046] Figure 4 A potential evaluation matrix was constructed using clay content, oil volume, water content, and production capacity.

[0047] Figure 5 This is a table showing the partial sorting results of potential sorting for multiple reservoirs on a validation dataset.

[0048] Figure 6 A schematic diagram of the oil and gas reservoir data processing device provided in this application;

[0049] Figure 7 A schematic diagram of the structure of the electronic device provided in this application.

[0050] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0051] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0052] The technical solution provided in this application is mainly used to interpret and evaluate the reservoir data of low-saturation oil and gas reservoirs, so as to identify high-value oil and gas reservoirs from low-saturation oil and gas reservoirs, which will facilitate the specific implementation of subsequent oil and gas exploration work.

[0053] Low-saturation oil and gas reservoirs are complex, and current technology uses two-dimensional fluid identification charts to interpret the reservoir data. Specifically, fluid identification charts are created using two logging data points: reservoir resistivity and sonic transit time. Analysis of these charts helps identify oil and water layers within the reservoir. However, due to the complex structure of low-saturation oil and gas reservoirs, resistivity and sonic transit time cannot accurately reflect the structural characteristics of each reservoir. This results in oil and water layers appearing mixed together on the two-dimensional fluid identification chart, and the unclear oil-water boundary makes it impossible to visually analyze the oil and gas layers within the reservoir.

[0054] Based on the above scenarios, it can be seen that the existing technology for interpreting reservoir data of low-saturation oil and gas reservoirs using two-dimensional fluid identification charts suffers from low interpretation accuracy, which in turn affects subsequent oil and gas exploration work.

[0055] The oil and gas reservoir data processing method provided in this application determines the target parameters for describing the reservoir by analyzing the formation mechanism of low-saturation oil and gas reservoirs; and improves the logging interpretation accuracy of low-saturation oil and gas reservoirs by using machine learning and analytic hierarchy process (AHP) to analyze the parameter values ​​of the target parameters, thereby solving the technical problem of inaccurate identification of potential layers in low-saturation oil and gas reservoirs in existing technologies.

[0056] The entity executing the oil and gas reservoir data processing method of this application can be a server with data processing capabilities, or any other device with data processing capabilities; this application does not limit this.

[0057] The technical solution of this application and how it solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. For ease of description, the oil and gas reservoirs mentioned in these specific embodiments all refer to low-saturation oil and gas reservoirs. The embodiments of this application will be described below with reference to the accompanying drawings.

[0058] Figure 1 This is a flowchart illustrating the oil and gas reservoir data processing method provided in this application. The oil and gas reservoir includes multiple reservoirs, such as... Figure 1 As shown, the method includes:

[0059] S101. Determine the target parameters related to the genesis mechanism of hydrocarbon reservoirs.

[0060] S102. Construct a potential evaluation matrix based on the target parameters, and determine the weights of the target parameters based on the potential evaluation matrix.

[0061] In this step, the weights of the target parameters determined based on the potential evaluation matrix are used to represent the importance of the target parameters in evaluating the reservoir's potential value. The potential value of a reservoir includes its exploitation value and market value. Reservoir fluid categories for oil and gas reservoirs include oil-bearing reservoirs, poor-oil reservoirs, oil-water co-containing reservoirs, oil-water-bearing reservoirs, water-bearing reservoirs, and dry reservoirs. Generally speaking, the potential value of oil-bearing reservoirs is higher than that of water-bearing and dry reservoirs.

[0062] S103. Obtain the parameter values ​​of the target parameters for each reservoir, and determine the potential coefficient of each reservoir based on the parameter values ​​and weights.

[0063] In this step, the determined reservoir potential coefficient can be used to characterize the reservoir's potential value.

[0064] S104. Determine the potential oil and gas reservoir layers based on the potential coefficients of each of the multiple reservoir layers.

[0065] In this step, since the potential coefficient can be used to characterize the potential value of a reservoir, the reservoir with higher potential value can be identified as the potential oil and gas reservoir by comparing the potential coefficients of multiple reservoirs.

[0066] The oil and gas reservoir data processing method provided in this embodiment determines target parameters for evaluation by analyzing the formation mechanism of oil and gas reservoirs; it constructs a potential evaluation matrix using the target parameters and derives the weights of the target parameters from the matrix; then, based on the target parameter values ​​and weights of each reservoir, it determines the potential coefficients of each reservoir, and identifies the potential layers of the oil and gas reservoir using these potential coefficients. Because the target parameters used are related to the formation mechanism of complex oil and gas reservoirs, the determined potential coefficients fully characterize the potential value of the reservoirs. Therefore, the method provided in this application provides a more accurate interpretation of the reservoir data and obtains more precise oil and gas potential layers.

[0067] To better understand the oil and gas reservoir data processing method provided in this application, an embodiment is provided to illustrate the detailed operation steps of the method provided in this application.

[0068] Through analysis of extensive research data on the genesis mechanisms of oil and gas reservoirs, the inventors discovered that well logging data includes target parameters strongly correlated with the assessment results of potential reservoirs. These target parameters include one or more of the following: clay content, oil-bearing volume, water cut, and production capacity. These target parameters can characterize the reservoir structure of oil and gas reservoirs.

[0069] Shale content, oil-bearing volume, water cut, and productivity can be obtained by analyzing reservoir logging data. Logging data includes various measurement curves and parameters. Manually analyzing logging data to obtain target parameters is laborious and prone to errors. Therefore, computational models can be used to automatically extract the parameter values ​​of multiple target parameters required by the method from the logging data. Specifically, the parameter values ​​of the target parameters for each reservoir are obtained, including:

[0070] First, determine the calculation model for obtaining the parameter values ​​of the target parameters;

[0071] Second, for each reservoir, the parameter values ​​of the target parameters are determined using a computational model based on the well logging data of that reservoir.

[0072] The following provides a calculation model for obtaining the parameter values ​​of reservoir clay content, oil-bearing volume, water cut, and productivity. It should be noted that the calculation model provided below can obtain the parameter values ​​of each target parameter of the reservoir. Other calculation models or methods that can obtain the parameter values ​​of each target parameter can also be used in the method provided in this application, and the method provided in this application is not limited in this respect.

[0073] The parameter values ​​for extracting reservoir clay content can be obtained using the calculation model of the natural gamma method. The formula (1) used in the calculation model is as follows:

[0074]

[0075] Among them, V shThe value represents the clay content; GCUR is an empirical coefficient related to the stratigraphic age of the reservoir, with a value of 2 for older strata and 3.7 for younger strata; I sh It can be calculated using the following formula (2):

[0076]

[0077] Among them, GR, GR max GR min These represent the natural gamma value, maximum natural gamma value, and minimum natural gamma value of the well logging data, respectively.

[0078] The calculation model used to extract the oil-bearing volume of the reservoir is determined by the effective porosity and oil saturation of the reservoir. The formula (3) used in the calculation model is as follows:

[0079] Vo = PRO × So (3);

[0080] Where Vo is the oil-bearing volume, PRO is the effective porosity, and So is the oil saturation.

[0081] The calculation model for extracting reservoir water cut can be obtained from the relationship between water cut and water saturation established by relative permeability experiments. The relationship between water cut and water saturation is generally an exponential function relationship. Different calculation models exist for different types of oil and gas reservoirs. The calculation model using the following formula (4) can be used:

[0082]

[0083] Where F is the moisture content and S is the water saturation.

[0084] The production capacity calculation models for different blocks in different oil fields are generally different. The following calculation model (5) can be used:

[0085]

[0086] Among them, E s The capacity coefficient can be calculated using the following formula (6):

[0087] E s =H×PRO×RI (6);

[0088] Where Q represents production capacity, H represents reservoir thickness, and RI represents the rate of increase in apparent resistivity.

[0089] Through the above process of selecting target parameters and obtaining target parameter values, parameter values ​​that fully characterize the structure of each reservoir in an oil and gas reservoir can be obtained, so as to conduct effective data analysis on the parameter values ​​when determining the potential layers of an oil and gas reservoir.

[0090] Because using two-dimensional fluid property identification charts for reservoir fluid category identification can easily confuse reservoirs with similar physical properties, leading to errors in reservoir fluid category identification. In some embodiments of this example, the method further includes: processing the parameter values ​​of the target parameters of multiple reservoirs based on a pre-trained fluid category identification model to obtain the fluid category of each reservoir.

[0091] The training operations for the fluid category recognition model include:

[0092] First, a training sample set is constructed. Each training sample in the training sample set includes an input-output pair. The input includes the sample parameter values ​​of the target parameters of the sample reservoir, and the output includes the fluid category corresponding to the sample reservoir.

[0093] Second, the Long Short-Term Memory (LSTM) network was trained using the training sample set to obtain the first training result.

[0094] Third, the Convolutional Neural Network (CNN) model is trained using the training sample set to obtain the second training result.

[0095] Fourth, the first training result and the second training result are combined to obtain the combined training result.

[0096] Generally, water-bearing and dry layers constitute a larger proportion of reservoirs in oil and gas reservoirs, while oil-bearing and oil-containing layers (including oil-bearing layers, poor-oil-bearing layers, oil-water co-existing layers, and oil-water-bearing layers) constitute a smaller proportion, resulting in an uneven distribution of reservoirs with different fluid types. However, in oil and oil-bearing layers, the potential value of oil-bearing and oil-containing layers is higher than that of dry and water-bearing layers. If the same data processing method is used for reservoirs with different fluid types during model training, the accuracy of the trained model in fluid category identification remains low. Therefore, the training operations for the fluid category identification model also include:

[0097] For each training sample in the training set, the model parameters are adjusted based on the loss value obtained from the cross-entropy loss function to minimize the loss value between the predicted value and the true value output by the adjusted model.

[0098] The traditional formula for calculating the cross-entropy loss function is shown in formula (7):

[0099]

[0100] Among them, L ori The cross-entropy loss value is given by y, where M and K represent the total number of samples in the training set and the number of reservoir fluid categories, respectively. m The true value representing the reservoir fluid type, y ′ mThis represents the model prediction value indicating the reservoir fluid type.

[0101] The optimized cross-entropy loss function provided in this application is calculated as shown in formula (8):

[0102]

[0103] Where, α m α represents the weight of reservoir fluid category m. m It can be set as the reciprocal of the ratio of the number of reservoir samples of fluid category m to the total number of samples.

[0104] In these implementations, the model training uses training sample sets to train both LSTM and CNN models. LSTM model training enables the trained fluid category recognition model to accurately identify the temporal features of sample target parameters, while CNN model training enables the fluid category recognition model to accurately identify the interrelationships between sample target parameters, thereby improving the accuracy of the method in reservoir fluid category recognition. Simultaneously, by assigning higher weights to oil-bearing layers using the cross-entropy loss function, the model focuses more on learning from oil-bearing layer sample data, further improving the fluid category recognition model's performance in identifying oil-bearing layers.

[0105] To select high-potential reservoirs from multiple reservoirs in an oil and gas reservoir, it is necessary to analyze the potential value of multiple reservoirs. The method provided in this application uses a reservoir potential coefficient to characterize the potential value of a reservoir. Step S102, which involves constructing a potential evaluation matrix based on target parameters and determining the weights of the target parameters according to the potential evaluation matrix, includes:

[0106] First, obtain the importance of each target parameter relative to that target parameter and relative to any other target parameter when determining the potential coefficient of each reservoir.

[0107] The numerical value of importance can be obtained by analyzing research data from geological experts on the relative importance of multiple reservoir target parameters.

[0108] Second, a potential evaluation matrix is ​​constructed based on the importance of each of the multiple target parameters.

[0109] The constructed potential matrix is ​​shown in formula (9):

[0110]

[0111] Where A is the potential evaluation matrix, n is the order of the potential evaluation matrix, which is equal to the total number of target parameters, and a ij This indicates the importance of target parameter i relative to target parameter j. A higher value indicates greater relative importance. The importance of the same target parameter relative to itself is 1.ij With a ji They are reciprocals of each other.

[0112] Third, the weights of mud content, oil volume, water content, and production capacity are determined using the Analytic Hierarchy Process (AHP) based on the potential evaluation matrix.

[0113] The calculation process for the target parameter weights of the potential evaluation matrix constructed above includes:

[0114] First, to avoid errors in setting the importance level, it is determined whether the potential evaluation matrix A passes the consistency check. The determination process is as follows:

[0115] The maximum eigenvalue of the potential evaluation matrix A is calculated using the formula (10):

[0116]

[0117] Where, λ max Let W be the largest eigenvalue of the potential evaluation matrix A, and AW be the standardized weights of the potential evaluation matrix. i The weighting coefficients for the target parameter i are calculated using formula (11):

[0118]

[0119] The consistency index CI is calculated using formula (12):

[0120]

[0121] The consistency ratio (CR) is calculated using the formula shown in formula (13):

[0122]

[0123] Here, RI is the average random consistency index, which can be obtained by looking up the consistency test index table. The consistency test index table records the statistical results of RI, which are obtained by AHP statistics. If CR is less than 0.1, the potential evaluation matrix is ​​judged to have passed the consistency test.

[0124] Secondly, after the potential evaluation matrix passes the consistency test, the weights of each objective parameter can be determined using the sum-product method or the square root method. The formula for calculating the weight of objective parameter i using the sum-product method is shown in formula (14):

[0125]

[0126] The formula for calculating the weight of the objective parameter i using the square root method is shown in formula (15):

[0127]

[0128] in, The calculation formula is as shown in formula (16):

[0129]

[0130] This yields the weights of each target parameter. It should be noted that after obtaining the weights of each target parameter, the reservoir's potential coefficient is obtained by weighted summing of the normalized parameter values ​​and weight values ​​of the multiple target parameters.

[0131] In these implementations, by constructing a potential evaluation matrix of multiple target parameters and incorporating geological experts' research data on the reservoir, the method provided in this application can more accurately interpret reservoir data, thereby obtaining more accurate potential coefficients.

[0132] To visually identify the potential layers of an oil and gas reservoir, step S104 determines the potential layers based on the potential coefficients of multiple reservoirs, including:

[0133] First, the reservoirs are grouped according to their respective fluid categories to obtain two or more grouping results; each grouping result includes at least one reservoir included in at least one fluid category.

[0134] Second, for each grouping result, at least one reservoir in the grouping result is sorted according to its potential coefficient to obtain the sorting result corresponding to the grouping result.

[0135] Third, determine the potential oil and gas reservoir layers based on the sorting results corresponding to the two or more grouping results.

[0136] Specifically, all oil-bearing layers can be grouped into one group, and all non-oil-bearing layers (including dry and water-bearing layers) into another group. All oil-bearing layers are ranked in descending order of their potential coefficients, and all non-oil-bearing layers are ranked in descending order of their potential coefficients. Since the potential value of oil-bearing layers is higher than that of non-oil-bearing layers, in the ranking results of oil-bearing layers, the reservoir with the higher the potential coefficient (i.e., the higher the ranking position), the greater the potential value. The top few reservoirs in the ranking results of oil-bearing layers can be considered as potential oil and gas reservoirs. It should be noted that the ranking method illustrated in this application is the preferred ranking method provided in this application; other ranking methods are also applicable to this application, and the method provided in this application is not limited in this regard.

[0137] To further illustrate that the technical solution used in the method provided in this application can solve the technical problems existing in the prior art, the following example is given using the experimental results of an experiment on a specific low-saturation oil and gas reservoir block.

[0138] Multiple wells with oil testing data were selected within a low-saturation oil and gas reservoir block. Oil testing data, including testing conclusions and production rates, were extracted for each reservoir. Simultaneously, a computational model was used to extract parameters such as clay content, oil-bearing volume, water cut, and productivity from the logging data collected from each well across multiple reservoirs. Data from three wells (Well I, Well II, and Well III) were selected as the test dataset. The remaining wells were divided into two parts at a 3:1 ratio: three-quarters of the well data served as the training dataset, and one-quarter as the validation dataset. For any experimental data set in the test, training, or validation datasets, the experimental data included oil testing data for one reservoir, along with the parameters for clay content, oil-bearing volume, water cut, and productivity for that reservoir.

[0139] First, it should be noted that the fluid category identification method provided in this application can effectively improve the accuracy of reservoir fluid category identification. Figure 2 This table presents partial experimental data statistics for fluid category recognition on the validation dataset. Experiments were conducted using three methods: a reservoir fluid recognition method based on the LSTM algorithm, a reservoir fluid recognition method based on a combination of LSTM and CNN, and a reservoir fluid recognition method based on the LSTM and CNN combination provided in this application, optimized using the cross-entropy loss function. The recall rate for reservoir fluid category recognition was statistically analyzed for each method.

[0140] The reservoir fluid identification method based on the combined LSTM and CNN algorithm concatenates the results of processing experimental data using the LSTM algorithm with the results of processing experimental data using the CNN algorithm to obtain a concatenated result. This concatenated result is then used to identify the reservoir fluid category.

[0141] like Figure 2As shown, the LSTM-based reservoir fluid identification method exhibits relatively high recall rates and good identification performance for water-bearing and dry reservoirs. However, it suffers from relatively low recall rates and poor identification performance for oil-bearing, poor-quality, oil-water co-layers, and oil-bearing water-bearing reservoirs. This is because water-bearing and dry reservoirs constitute a large proportion of the reservoir fluid categories, causing the model to neglect the characteristics of oil-bearing reservoirs during training, resulting in poor identification performance. After introducing CNN into the algorithm, the reservoir fluid identification method based on the combination of LSTM and CNN shows improved recall rates for oil-bearing, oil-water co-layers, and poor-quality oil-bearing reservoirs compared to the LSTM-based method. This demonstrates that the reservoir fluid identification method based on the combination of LSTM and CNN can effectively characterize the temporal features of target parameters and the interrelationships between multiple target parameters. After optimizing the model by introducing the cross-entropy loss function, the fluid identification recall rate for water-bearing and dry reservoirs decreases relatively, but the recall rate for high-value oil-bearing, poor-quality, oil-water co-layers, and oil-bearing water-bearing reservoirs improves. Compared with reservoir fluid identification methods based on LSTM and CNN combined algorithms, the fluid category identification method provided in this application improves the recall rate by 29% for oil layer identification, 26% for poor oil layer identification, 34% for oil-water co-containment layer identification, and 22% for oil-water layer identification.

[0142] To further illustrate that the fluid category identification method provided in this application can effectively improve the accuracy of reservoir fluid category identification, experiments were conducted on the test dataset using the reservoir fluid category identification method based on the algorithm combining LSTM and CNN, and the reservoir fluid identification method provided in this application based on the algorithm combining LSTM and CNN and using the cross-entropy loss function for model optimization. The corresponding oil test results of the experimental data were used for comparison and illustration.

[0143] Figure 3 This is a partial statistical table of experimental data for fluid category recognition on the test dataset, such as... Figure 3 As shown, the reservoir fluid category identification method based on the LSTM and CNN combination algorithm performs poorly in identifying oil-bearing reservoirs, misidentifying most oil-bearing reservoirs as water layers. The reservoir fluid identification method provided in this application, based on the LSTM and CNN combination algorithm and optimized using the cross-entropy loss function, performs better in identifying oil-bearing reservoirs than the LSTM and CNN combination algorithm.

[0144] The following explains how the oil and gas reservoir data processing method provided in this application can effectively improve the accuracy of reservoir data interpretation. In the stage of constructing the potential evaluation matrix, a potential evaluation matrix is ​​constructed based on research data from geological experts on the relative importance of every two of the four target parameters: clay content, oil-bearing volume, water cut, and production capacity. Figure 4 A potential evaluation matrix was constructed using clay content, oil volume, water content, and production capacity, such as... Figure 4 As shown, the importance of oil volume relative to mud content is 3, the importance of water content relative to mud content is 5, the importance of water content relative to oil volume is 3, the importance of production capacity relative to mud content is 7, the importance of production capacity relative to oil volume is 5, the importance of production capacity relative to water content is 3, the importance of each target parameter relative to itself is 1, and the importance of mud content relative to oil volume is the reciprocal of the importance of oil volume relative to mud content, and so on, to obtain the potential evaluation matrix.

[0145] The analytic hierarchy process (AHP) was used to test the consistency of the constructed potential evaluation matrix. The CR value of the potential evaluation matrix was 0.039, which is less than 0.1; therefore, the constructed potential evaluation matrix passed the consistency test. After passing the consistency test, the weights corresponding to each objective parameter in the potential evaluation matrix were calculated using the sum-product method. The calculated weights were: clay content 5.646%, oil content 12.182%, water content 26.31%, and production capacity 55.861%.

[0146] For each experimental data point in the validation dataset, a potential coefficient is calculated using the weights of the calculated mud content, oil-bearing volume, water content, and productivity, along with the parameter values ​​of each target parameter in the experimental data. Simultaneously, the fluid category identification method provided in this application is used to analyze the parameter values ​​of each target parameter in the experimental data to obtain the corresponding fluid category.

[0147] After obtaining the fluid category and potential coefficient corresponding to each of the experimental data, the experimental data were first sorted according to the fluid category, resulting in the first sorting result. Based on the first sorting result, the experimental data were then sorted a second time according to the potential coefficient, resulting in the second sorting result. Finally, according to the standard definition of industrial oil production, reservoirs with production exceeding the lower limit of industrial oil production were classified as Class I, reservoirs with production below the industrial oil production standard were classified as Class II, and other reservoirs were classified as Class III. Figure 5 This is a table showing the partial sorting results of potential sorting for multiple reservoirs on a validation dataset. For example... Figure 5 As shown, Class I reservoirs have the highest potential value, followed by Class II, and Class III have the lowest potential value. Depending on the business scenario, Class I or Class II reservoirs can be considered as potential oil and gas reservoirs.

[0148] Figure 6This is a schematic diagram of the structure of the oil and gas reservoir data processing device provided in this application. The oil and gas reservoir includes multiple reservoirs, such as... Figure 6 As shown, the oil and gas reservoir data processing device 60 provided in this embodiment includes:

[0149] The parameter determination unit 601 is used to determine target parameters related to the formation mechanism of oil and gas reservoirs.

[0150] The data processing unit 602 is used to construct a potential evaluation matrix based on the target parameters, determine the weights of the target parameters according to the potential evaluation matrix, obtain the parameter values ​​of the target parameters for each reservoir, and determine the potential coefficients of each reservoir based on the parameter values ​​and weights.

[0151] Potential layer determination unit 603 is used to determine the potential layers of oil and gas reservoirs based on the potential coefficients of multiple reservoirs.

[0152] In one possible implementation, the target parameters determined by the parameter determination unit 601 include one or more of the following: mud content, oil-bearing volume, water content, and production capacity.

[0153] In one possible implementation, the parameter determination unit 601 is also used to process the parameter values ​​of the target parameters of multiple reservoirs based on a pre-trained fluid category recognition model to obtain the fluid category of each reservoir.

[0154] In one possible implementation, the data processing unit 602 is further configured to construct a training sample set, wherein each training sample in the training sample set includes an input-output pair, the input including sample parameter values ​​of the target parameters of the sample reservoir, and the output including the fluid category corresponding to the sample reservoir; the training sample set is used to train a long short-term memory network to obtain a first training result; the training sample set is used to train a convolutional neural network to obtain a second training result; and the first training result and the second training result are concatenated to obtain a concatenated training result.

[0155] In one possible implementation, the data processing unit 602 is further configured to adjust the model parameters for each training sample in the training set based on the loss value obtained from the cross-entropy loss function, so as to minimize the loss value between the predicted value and the true value of the adjusted model output.

[0156] In one possible implementation, the data processing unit 602 is further configured to obtain the importance of each target parameter relative to the target parameter and relative to any other target parameter when determining the potential coefficient of each reservoir; construct a potential evaluation matrix based on the importance of each of the multiple target parameters; and determine the weights of the clay content, oil-bearing volume, water cut, and production capacity using the analytic hierarchy process based on the potential evaluation matrix.

[0157] In one possible implementation, the parameter determination unit 601 is further used to determine a calculation model for obtaining the parameter values ​​of the target parameters; for each reservoir, the parameter values ​​of the target parameters are determined using the calculation model based on the logging data of that reservoir.

[0158] In one possible implementation, the potential layer determination unit 603 is further configured to group multiple reservoirs according to their respective fluid categories to obtain two or more grouping results; each grouping result includes at least one reservoir included in at least one fluid category; for each grouping result, at least one reservoir in the grouping result is sorted according to its respective potential coefficient to obtain the sorting result corresponding to the grouping result; and the potential layer of the oil and gas reservoir is determined according to the sorting results corresponding to the two or more grouping results.

[0159] The oil and gas reservoir data processing device provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0160] Figure 7 A schematic diagram of the structure of the electronic device provided in this application. Figure 7 As shown, the electronic device 70 provided in this embodiment includes at least one processor 301 and a memory 702. Optionally, the device 70 further includes a communication component 703. The processor 701, memory 702, and communication component 703 are connected via a bus 704.

[0161] In a specific implementation, at least one processor 701 executes computer execution instructions stored in memory 702, causing at least one processor 701 to perform the above-described method.

[0162] The specific implementation process of processor 701 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0163] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0164] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0165] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0166] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0167] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0168] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0169] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0170] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0171] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0172] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0173] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0174] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A method for processing oil and gas reservoir data, characterized in that, The oil and gas reservoir includes multiple reservoirs, and the method includes: Determine the target parameters related to the genesis mechanism of hydrocarbon reservoirs; A potential evaluation matrix is ​​constructed based on the target parameters, and the weights of the target parameters are determined based on the potential evaluation matrix. Obtain the parameter values ​​of the target parameters for each reservoir, and determine the potential coefficient of each reservoir based on the parameter values ​​and the weights; The potential oil and gas reservoirs are determined based on the potential coefficients of each of the multiple reservoirs.

2. The method according to claim 1, characterized in that, The target parameters include one or more of the following: mud content, oil volume, water content, and production capacity.

3. The method according to claim 2, characterized in that, The method further includes: Based on a pre-trained fluid category recognition model, the parameter values ​​of the target parameters of each of the multiple reservoirs are processed to obtain the fluid category of each reservoir.

4. The method according to claim 3, characterized in that, The training operations for the fluid category recognition model include: Construct a training sample set. Each training sample in the training sample set includes an input-output pair. The input includes the sample parameter values ​​of the target parameters of the sample reservoir, and the output includes the fluid category corresponding to the sample reservoir. The Long Short-Term Memory Network was trained using the training sample set to obtain the first training result. The convolutional neural network model is trained using the training sample set to obtain the second training result. By combining the first training result and the second training result, a combined training result is obtained.

5. The method according to claim 4, characterized in that, The training operation of the fluid category recognition model further includes: For each training sample in the training set, the model parameters are adjusted based on the loss value obtained from the cross-entropy loss function to minimize the loss value between the predicted value and the true value output by the adjusted model.

6. The method according to any one of claims 2-5, characterized in that, The step of constructing a potential evaluation matrix based on the target parameters and determining the weights of the target parameters based on the potential evaluation matrix includes: To determine the importance of each target parameter relative to that target parameter and relative to any other target parameter when determining the potential coefficient of each reservoir; Based on the importance of each of the multiple target parameters, the potential evaluation matrix is ​​constructed; The weights of the mud content, oil volume, water content, and production capacity are determined using the analytic hierarchy process (AHP) based on the potential evaluation matrix.

7. The method according to any one of claims 2-5, characterized in that, The process of obtaining the target parameter values ​​for each reservoir includes: Determine the computational model for obtaining the target parameter values; For each reservoir, the parameter value of the target parameter is determined using the calculation model based on the well logging data of that reservoir.

8. The method according to claim 3, characterized in that, The step of determining the potential oil and gas reservoir layers based on the potential coefficients of each of the plurality of reservoir layers includes: The reservoirs are grouped according to their respective fluid categories, resulting in two or more grouping results; each grouping result includes at least one reservoir included in at least one fluid category; For each grouping result, at least one reservoir in the grouping result is sorted according to its potential coefficient to obtain the sorting result corresponding to the grouping result; Based on the sorting results corresponding to the two or more grouping results, the potential oil and gas reservoir layers are determined.

9. A data processing device for oil and gas reservoirs, characterized in that, The oil and gas reservoir includes multiple reservoirs, and the apparatus includes: The parameter determination unit is used to determine target parameters related to the genesis mechanism of hydrocarbon reservoirs; A data processing unit is used to construct a potential evaluation matrix based on the target parameters and determine the weights of the target parameters according to the potential evaluation matrix. Obtain the parameter values ​​of the target parameters for each reservoir, and determine the potential coefficient of each reservoir based on the parameter values ​​and the weights; A potential layer determination unit is used to determine the potential layers of oil and gas reservoirs based on the potential coefficients of the plurality of reservoirs.

10. An electronic device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-8.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-8.

Images