[0025] In order to make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following describes the embodiments of the present invention in further detail with reference to the accompanying drawings. Here, the exemplary embodiments of the present invention and the description thereof are used to explain the present invention, but not as a limitation to the present invention.
[0026] In order to solve the above-mentioned problems, the embodiment of the present invention provides a comprehensive analysis method for the depletion production characteristics of a fractured-vuggy carbonate reservoir. figure 1 It is a flowchart of a comprehensive analysis method for the depletion exploitation characteristics of fracture-cavity carbonate reservoirs in the embodiment of the present invention, such as figure 1 As shown, the method can include:
[0027] Step 101: Select the characteristic attribute parameters of the fracture-cavity carbonate reservoir to construct an evaluation sample set of the depletion exploitation characteristics of the fracture-cavity carbonate reservoir;
[0028] Step 102: By performing discrete particle swarm global optimization processing on the evaluation sample set, adaptively solve the initial cluster center and the optimal number of clustering groups of the evaluation sample set;
[0029] Step 103: Perform a fuzzy cluster analysis on the depleted production characteristics of the fractured-vuggy carbonate reservoir according to the initial cluster center and the optimal cluster group number.
[0030] In specific implementation, based on the analysis of production dynamics, the independent characteristic attribute parameters can be screened to construct an evaluation sample set, and then combined with the discrete particle swarm global optimization technology to adaptively solve the initial cluster centers and determine the optimal number of clusters Finally, the fuzzy cluster analysis algorithm is used to quantitatively evaluate the dynamic response characteristics of different types of depleted oil wells. The fuzzy clustering analysis algorithm can be, for example, the classic fuzzy C-means clustering algorithm, or other fuzzy clustering analysis algorithms can be used according to actual needs.
[0031] In the embodiment, when selecting the characteristic attribute parameters of the fracture-cavity carbonate reservoir, it can be selected according to actual needs. For example, the production well depletion time and the initial daily oil production of the fracture-cavity carbonate reservoir can be selected. Volume, peak daily oil production, peak water cut, cumulative oil production, cumulative water production, predicted recoverable reserves, initial decline rate, and final decline rate are the characteristic attribute parameters of fracture-cavity carbonate reservoirs. After selecting the characteristic attribute parameters of the fracture-cavity carbonate reservoir, these characteristic attribute parameters are used to construct the evaluation sample set of the depletion production characteristics of the fracture-cavity carbonate reservoir.
[0032] In specific implementation, after the evaluation sample set is constructed, the initial cluster center and the optimal number of clustering groups of the evaluation sample set can be adaptively solved by performing the discrete particle swarm global optimization process on the evaluation sample set. In order to improve the accuracy of the solution results, you can first standardize the evaluation sample set, which includes eliminating the systematic errors caused by the inconsistency of different characteristic attribute parameters; and then perform the discrete particle swarm globalization on the standardized evaluation sample set Optimization processing. Various methods can be used for the standardization processing, for example, methods such as translation standard deviation can be used.
[0033] When performing the discrete particle swarm global optimization processing on the evaluation sample set, the control parameters of the discrete particle swarm global optimization algorithm can be initialized; then samples are randomly selected from the evaluation sample set to form the initial clustering center set, and the initial clustering center is determined according to the control parameters The ensemble performs a global random search to obtain the initial clustering center and the optimal number of clustering groups; the following process is continuously repeated until the convergence of the discrete particle swarm optimization algorithm stops:
[0034] A new sample is randomly selected from the evaluation sample set, combined with the initial cluster center and the best cluster group number obtained in the last global random search, the initial cluster center set is updated, and the global random search is performed again on the new initial cluster center set , Get the new initial cluster center and the optimal number of cluster groups.
[0035] Among them, when initializing the control parameters of the discrete particle swarm global optimization algorithm, you can initialize the particle swarm size, cognitive learning factor, group learning factor, inertia weight, fuzzy weighting index, maximum number of iterations, and discrete particle swarm global optimization according to actual needs. Control parameters such as the iterative convergence threshold of the algorithm.
[0036] In an embodiment, the foregoing initialization process may further include: initializing the global optimal values of the position, velocity, and position of the particle group in the initial cluster center set, and the local optimal values of the position, velocity, and position of a single discrete particle. In this way, when performing a global random search on the initial clustering center set according to the control parameters to obtain the initial clustering center and the optimal number of clustering groups, the following global random optimization solution process for the particle position can be repeated until the discrete particle swarm global optimization The algorithm reaches the maximum number of iterations or the fitness error is less than the iteration convergence threshold:
[0037] Calculate the initial partition matrix according to the position of each discrete particle; calculate the fitness of each discrete particle; update the global optimal value of the position of the particle group and the local optimal value of the position of a single discrete particle; update the position and velocity of each discrete particle .
[0038] As mentioned earlier, some fuzzy cluster analysis algorithms can be used to perform fuzzy cluster analysis on the depletion mining characteristics of fractured-vuggy carbonate reservoirs. Here, the fuzzy C-means clustering algorithm is taken as an example, and the fuzzy cluster analysis is specifically carried out According to the initial clustering center and the optimal number of clustering groups, the fuzzy C-means clustering algorithm can be used to iteratively solve the membership degree matrix and the clustering center matrix until the least square objective function reaches the minimum, and the fracture-cavity carbonate is obtained. Fuzzy cluster analysis results of depletion mining characteristics of rock reservoirs. In the embodiment, before implementing the fuzzy C-means clustering algorithm, it includes initializing the control parameters of the fuzzy C-means clustering algorithm, for example, initializing the clustering iteration convergence threshold of the algorithm.
[0039] In the embodiment, after obtaining the fuzzy cluster analysis result of the depleted production characteristics of the fractured-vuggy carbonate reservoir, the fuzzy cluster analysis result can be further applied in the field, for example, the fuzzy cluster analysis result is determined Dynamic response characteristics of different types of depleted oil wells.
[0040] The following is a specific implementation to illustrate the specific implementation process of the comprehensive analysis method for the depletion exploitation characteristics of the fracture-cavity carbonate reservoir. In this example, the fuzzy clustering analysis using fuzzy C-means clustering algorithm is taken as an example. The comprehensive analysis process is as follows: figure 2 As shown, it can include:
[0041] Step 201. Optimizing the 9 dynamic indicators of production well depletion production time, initial daily oil production, peak daily oil production, peak water cut, cumulative oil production, cumulative water production, predicted recoverable reserves, initial decline rate and final decline rate As a characteristic attribute parameter, an evaluation sample set X is constructed.
[0042] Step 202: Perform standardization processing on the evaluation sample set X based on the translation standard deviation method to eliminate the systematic errors caused by the inconsistent dimension of different feature attribute parameters.
[0043] Step 203: Initialize the control parameters of the discrete particle swarm optimization (PSO) global optimization algorithm and the fuzzy C-mean (FCM) clustering algorithm, including: particle swarm size P, cognitive learning factor c 1 , Group learning factor c 2 , Inertia weight w, fuzzy weighted index m, maximum number of iterations T max , PSO iteration convergence threshold ε 1 , FCM clustering iteration convergence threshold ε 2 Wait.
[0044] Step 204: Randomly select P samples from the evaluation sample set X to form an initial cluster center set M, and use a 0-1 binary variable to randomly initialize the positions of the P discrete particles, and the speed of the particles is used to characterize the changes in the positions of the particles. Among them, the value of each dimension of each particle indicates whether the sample in the corresponding M set is selected as the initial cluster center, and the total number of samples selected as the initial cluster center is the cluster group number C.
[0045] Step 205: Initialize the position and velocity of a single discrete particle, the local optimal value pbest of the individual position, and the global optimal value gbest of the particle group position.
[0046] Step 206. (1) Obtain the optimal initial clustering center and determine the optimal number of clustering groups through the global random optimization of the particle position, including the following steps: ①. Calculate the initial division matrix according to the position of each particle ②. Calculate the fitness of each particle; ③. Update the global optimal value gbest of the group position and the local optimal value pbest of the individual position; ④ Update the position and velocity vector of each particle at the same time; ⑤ If the PSO global optimization process is discrete The maximum number of iterations is reached or the fitness error is less than the convergence threshold ε 1 , The calculation stops, otherwise, go to step ① to continue the solution.
[0047] (2) Randomly select a new sample from the evaluation sample set X, combine the initial cluster center results obtained by the process of (1) optimization, update the initial cluster center set M, and repeat the global randomness of (1) for the new set M The search process stops until the algorithm converges.
[0048] Step 207: Based on the adaptive solution result of the initial clustering center and the optimal number of groups, the classical fuzzy C-means clustering algorithm is used to iteratively solve the membership matrix U and the clustering center matrix V, so that the least square objective function J m (U,V) reaches the minimum, and finally realizes the cluster analysis of the evaluation sample set, including the following steps: ①, calculate the cluster center; ②, calculate the Euclidean distance between each sample and the cluster center; ③, modify the membership matrix U, calculate the objective function J m ④For a given threshold ε 2 , If the objective function satisfies Then the algorithm is terminated; otherwise, go to step ① for the next iteration calculation.
[0049] Step 208: Apply the results of the comprehensive clustering evaluation to actual on-site conditions to guide the formulation of differentiated comprehensive management of different types of depleted oil wells and the formulation of measures to increase recovery efficiency.
[0050] by figure 2 The shown process shows that in this example, the fuzzy C-means clustering analysis algorithm improved based on the discrete particle swarm global optimization technology is used to comprehensively evaluate the depletion characteristics of the production well in the fractured-vuggy carbonate reservoir. First, based on the production dynamic analysis, the independent characteristic attribute parameters are screened to construct the evaluation sample set, and then combined with the discrete particle swarm global optimization technology to adaptively solve the initial clustering center and determine the optimal number of clustering groups, and finally adopt the classic The fuzzy C-means cluster analysis algorithm quantitatively evaluates the dynamic response characteristics of different types of depleted oil wells.
[0051] An application example of the comprehensive analysis method for the depletion production characteristics of a fractured-cavity carbonate reservoir in an embodiment of the present invention is given below.
[0052] In this example, the main effective storage space of the M fractured-vuggy carbonate reservoir includes large caves, secondary dissolution pores and fractures, etc. The reservoir is an unsaturated oil reservoir with a large pressure difference (35-66MPa). , An average of 50.5MPa, the formation crude oil is mainly low-viscosity, low-sulfur, medium-containing colloidal asphaltene, high-wax light crude oil, density 0.82~1.10g/cm 3 , The volume coefficient is 1.04-1.66, the average formation pressure is 73.3MPa, and the average formation temperature is 154.4℃, which is a normal temperature-pressure system. According to the production performance data of 62 production wells in the M reservoir, and based on the comprehensive analysis method of the embodiment of the present invention, all oil wells are divided into four major categories and 8 subcategories according to their dynamic response characteristics. Table 1 shows the characteristic attribute parameters of the cluster centers of different oil well types obtained by the improved fuzzy C-means clustering analysis of the discrete PSO global search optimization.
[0053] Table 1 Characteristic attribute parameters of cluster centers belonging to different oil well types
[0054]
[0055] It can be seen from Table 1 that as the classification level of depleted oil wells increases, the three production performance indicators of depleted production time, cumulative oil production and predicted recoverable reserves show a clear increasing trend, and the initial decline rate shows a clear decline. Trend, which indicates that the more natural energy of the fractured-vuggy reservoir controlled by a single well, the larger the reservoir size.
[0056] image 3 In this example, a comparison chart of predicted recoverable reserves of different types of depleted oil wells with depleted production time; Figure 4 In this example, the comparison chart of the initial decline rate of different depletion production well types with the depletion production time; Figure 5 In this example, the cumulative oil production comparison chart of different depletion production well types with depletion production time; Image 6 In this example, the comparison chart of the terminal decline rate of different depletion production well types with depletion production time. in Figure 3 to Figure 6 Based on the comprehensive clustering evaluation results of depleted production characteristics shown, the dynamic response characteristics of four major categories and 8 sub-categories of depleted wells are described quantitatively as follows:
[0057] For Type I depleted oil wells, the depleted production time is generally greater than 800 days, and the predicted recoverable reserves are higher than 10.0×10 4 m 3 , The cumulative oil production is higher than 6.0×10 4 m 3 , The initial decline rate D i And the terminal decline rate D e Generally between 0% and 10%, it is believed that the depletion mining stage meets the characteristics of high production + stable production, and the overall performance is the production characteristics of "large fracture-cavity aggregates".
[0058] For Type II depleted oil wells, the depleted production time is generally less than 800 days, and the predicted recoverable reserves are less than 10.0×10 4 m 3 , The cumulative oil production is less than 4.0×10 4 m 3 , The initial decline rate D i Generally between 0% and 30%, the terminal decline rate D e Between 0% and 10%, it is considered that the depletion production stage meets the characteristics of slow decline. Combined with the Arps decline analysis, the Type II depletion production well can be subdivided into three subcategories II-1, II-2 and II-3. Corresponding to the low-yield + stable-yield type (D i <10%), hyperbolic slowly decreasing type (10% i <30% and D i D e ) And slowly decreasing index type (10% i <30% and D i =D e ), the overall performance is the production characteristics of "cavern-type" reservoirs.
[0059] For Type III depleted oil wells, the depleted production time is generally less than 400 days, and the predicted recoverable reserves are less than 5.0×10 4 m 3 , The cumulative oil production is less than 1.0×10 4 m 3 , The initial decline rate D i Generally between 30% and 70%, the terminal decline rate D e Between 0% and 40%, it is believed that the depletion production stage meets the characteristics of express decline. Combined with the Arps decline analysis, the Type III depletion production well can be subdivided into two sub-categories III-1 and III-2, which in turn correspond to double Quickly decreasing type (30% i <70% and D i D e ) And exponential rapidly decreasing type (30% i <70% and D i =D e ), the overall performance is the production characteristics of "fracture-cavity type" reservoirs.
[0060] For Type IV depleted oil wells, the depleted production time is generally less than 100 days, and the predicted recoverable reserves are less than 1.0×10 4 m 3 , The cumulative oil production is less than 0.3×10 4 m 3 , The initial decline rate D i Generally above 70%, it is considered that the depletion production stage conforms to the characteristics of violent decline. Combined with the analysis of the decline trend of Arps, the Type IV depleted oil wells can be subdivided into two subcategories IV-1 and IV-2, corresponding to double Curve violent decreasing type (D i 70% and D i D e ) And exponentially decreasing type (D i 70% and D i =D e ), the overall performance is the production characteristics of "fractured" reservoirs.
[0061] Figure 7 Summarizes the dynamic attribute classification standards of various well types in the M fracture-vuggy reservoir in this example, where D i Is the initial decline rate, D e Is the terminal decline rate, Q o Is the output and t is the time. For depleted oil wells with a short production time, based on the initial production data, understand the dynamic response mode of depleted production and determine the type of oil well to which they belong, so as to propose targeted technical measures to improve oil recovery and greatly improve reservoir development effect.
[0062] Based on the same inventive concept, the embodiments of the present invention also provide a comprehensive analysis device for the depletion production characteristics of fractured-vuggy carbonate reservoirs, as described in the following embodiments. Since the problem-solving principle of this device is similar to the comprehensive analysis method for the depletion exploitation characteristics of fractured-vuggy carbonate reservoirs, the implementation of this device can refer to the comprehensive analysis method for the depletion exploitation characteristics of fractured-vuggy carbonate reservoirs. Implementation, the repetition will not be repeated.
[0063] Figure 8 It is a schematic diagram of a comprehensive analysis device for depletion production characteristics of fracture-cavity carbonate reservoirs in an embodiment of the present invention. Such as Figure 8 As shown, the device may include:
[0064] The sample set construction module 801 is used to select the characteristic attribute parameters of the fracture-cavity carbonate reservoir to construct the evaluation sample set of the depletion exploitation characteristics of the fracture-cavity carbonate reservoir;
[0065] The global optimization module 802 is used to adaptively solve the initial cluster center and the optimal number of clustering groups of the evaluation sample set by performing the discrete particle swarm global optimization processing on the evaluation sample set;
[0066] The fuzzy clustering module 803 is used to perform fuzzy clustering analysis on the depletion mining characteristics of the fractured-vuggy carbonate reservoir according to the initial clustering center and the optimal clustering group number.
[0067] In specific implementation, the sample set building module 801 can be specifically used to: select the production well depletion time of the fracture-cavity carbonate reservoir, the initial daily oil production, the peak daily oil production, the peak water cut, the cumulative oil production, Cumulative water production, predicted recoverable reserves, initial decline rate and final decline rate are used as characteristic attribute parameters of fracture-vuggy carbonate reservoirs.
[0068] In specific implementation, the global optimization module 802 can also be used to standardize the evaluation sample set before performing the discrete particle swarm global optimization processing on the evaluation sample set. The standardization processing includes eliminating system errors caused by inconsistent dimensions of different feature attribute parameters. .
[0069] Such as Picture 9 As shown, during specific implementation, the global optimization module 802 may include:
[0070] The initialization unit 901 is used to initialize the control parameters of the discrete particle swarm optimization algorithm;
[0071] The global random search unit 902 is configured to randomly select samples from the evaluation sample set to form an initial cluster center set, perform a global random search on the initial cluster center set according to the control parameters, and obtain the initial cluster center and the optimal number of cluster groups; The process is as follows until the convergence of the discrete particle swarm global optimization algorithm stops: a new sample is randomly selected from the evaluation sample set, combined with the initial cluster center and the optimal cluster group number obtained by the last global random search, and the initial cluster center set is updated. The new initial cluster center set re-does a global random search to obtain the new initial cluster center and the optimal number of cluster groups.
[0072] During specific implementation, the initialization unit 901 may be specifically used to initialize the particle population size, cognitive learning factor, population learning factor, inertia weight, fuzzy weighting index, maximum number of iterations, and iterative convergence threshold of the discrete particle swarm optimization algorithm.
[0073] During specific implementation, the initialization unit 901 may also be used to:
[0074] Initialize the global optimal value of the position, velocity and position of the particle group in the initial cluster center set, as well as the local optimal value of the position, velocity and position of a single discrete particle;
[0075] The global random search unit 902 may be specifically configured to: repeat the following global random optimization solution process for particle positions until the discrete particle swarm global optimization algorithm reaches the maximum number of iterations or the fitness error is less than the iteration convergence threshold:
[0076] Calculate the initial partition matrix according to the position of each discrete particle;
[0077] Calculate the fitness of each discrete particle;
[0078] Update the global optimal value of the position of the particle group and the local optimal value of the position of a single discrete particle;
[0079] Update the position and velocity of each discrete particle.
[0080] During specific implementation, the fuzzy clustering module 803 can be specifically used for:
[0081] According to the initial clustering center and the optimal number of clustering groups, the fuzzy C-means clustering algorithm is used to perform fuzzy clustering analysis on the depletion mining characteristics of fractured-vuggy carbonate reservoirs.
[0082] During specific implementation, the fuzzy clustering module 803 can be specifically used for:
[0083] According to the initial clustering center and the optimal number of clustering groups, the fuzzy C-means clustering algorithm is used to iteratively solve the membership matrix and the clustering center matrix until the least square objective function reaches the minimum, and the fracture-cavity carbonate reservoir is obtained The results of fuzzy cluster analysis of the depletion mining characteristics.
[0084] Such as Picture 10 As shown in the implementation, Figure 8 The shown comprehensive analysis device for the depletion exploitation characteristics of fracture-cavity carbonate reservoirs may also include:
[0085] The result application module 1001 is used to determine the dynamic response characteristics of different types of depleted oil wells according to the fuzzy cluster analysis results of the depleted production characteristics of fractured-vuggy carbonate reservoirs. In the embodiment, the result application module 1001 can be located in Picture 9 Shown in the comprehensive analysis device for the depletion exploitation characteristics of fracture-cavity carbonate reservoirs.
[0086] In summary, the embodiments of the present invention overcome the inherent shortcomings of traditional fuzzy clustering algorithms, such as being sensitive to initial data, relying on artificially setting the initial clustering center and number of clustering groups, and easily falling into local optimal solutions. Discrete particle swarm optimization is performed on the set, and the initial clustering center and optimal clustering group number of the evaluation sample set are adaptively solved. Based on this, the fuzzy clustering analysis of the depletion mining characteristics of the fracture-cavity carbonate reservoir is realized. , Improve the reliability of the comprehensive analysis results, can provide a theoretical basis for the differential comprehensive management of different types of depleted oil wells and the formulation of measures to increase the recovery rate, thereby improving the development effect of fracture-cavity carbonate reservoirs.
[0087] The embodiment of the present invention can comprehensively and quantitatively evaluate the production dynamic response characteristics of different types of depleted oil wells, and finally determine the typical fractured-vuggy development mode in combination with the fine reservoir description results. On the basis of understanding the remaining oil distribution law and the main contradiction between development, To implement strategies for tapping the potential of remaining oil, such as deploying new or old wells sidetracking, turning machines, reservoir reconstruction, and water injection and gas injection development, in order and in a targeted manner.
[0088] Those skilled in the art should understand that the embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may be in the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.
[0089] The present invention is described with reference to flowcharts and/or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and/or block in the flowchart and/or block diagram, and the combination of processes and/or blocks in the flowchart and/or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing equipment to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing equipment can be generated In the process Figure one Process or multiple processes and/or boxes Figure one A device with functions specified in a block or multiple blocks.
[0090] These computer program instructions can also be stored in a computer-readable memory that can guide a computer or other programmable data processing equipment to work in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture including the instruction device. The device is implemented in the process Figure one Process or multiple processes and/or boxes Figure one Functions specified in a box or multiple boxes.
[0091] These computer program instructions can also be loaded on a computer or other programmable data processing equipment, so that a series of operation steps are executed on the computer or other programmable equipment to produce computer-implemented processing, so as to execute on the computer or other programmable equipment. Instructions are provided to implement the process Figure one Process or multiple processes and/or boxes Figure one Steps of functions specified in a box or multiple boxes.
[0092] The specific embodiments described above further describe the purpose, technical solutions and beneficial effects of the present invention in further detail. It should be understood that the above are only specific embodiments of the present invention and are not intended to limit the scope of the present invention. The protection scope, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
