Well spacing optimization method and device based on injection-production function modeling, equipment and medium

Abstract

The application discloses a well spacing optimization method and device based on injection-production function diagram modeling, equipment and medium, belongs to the intelligent optimization technical field of oil and gas field development parameters, and the method comprises the steps that corresponding injection-production function diagrams are established for a group of well spacing schemes; the injection-production function diagrams are input into a response prediction model to obtain a prediction value of a whole well group development response index; a target function value is calculated according to the prediction value; another group of well spacing schemes is selected, and the injection-production function diagram establishing step is returned for the next cycle until the target function values corresponding to all well spacing schemes are obtained; all the target function values are sorted to determine the optimization result. The application takes the injection-production relationship in the development process of an oil and gas reservoir as the core, depicts the inter-well coupling structure by constructing the injection-production function diagram, predicts the development effect under different well spacing configurations by using the response prediction model on the basis of the diagram structure, further combines the target function definition and the optimization strategy, and realizes the intelligent recommendation of the optimal well spacing parameter combination.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of intelligent optimization of oil and gas field development parameters, and particularly relates to a well spacing optimization method and device based on injection-production function graph modeling, equipment and medium. BACKGROUND

[0002] In the process of oil and gas reservoir development, injection-production well spacing is a key parameter for controlling oil displacement efficiency, injection-production connectivity and recovery. Reasonable well spacing configuration helps to optimize the drive system, reduce injection energy consumption and improve overall development benefit. The traditional well spacing design method mainly relies on analogy experience, geological zoning or numerical simulation means, which is difficult to fully represent the injection-production coupling structure between wells and lacks efficient multi-scheme evaluation mechanism. Especially under the conditions of heterogeneity, multi-well group or insufficient historical data, well spacing optimization faces the problems of low evaluation efficiency and weak decision basis, which restricts the precision and intelligent level of well pattern deployment.

[0003] Although the current numerical simulation method can be used for response evaluation, it highly depends on model accuracy, initial boundary conditions and personnel experience, and the simulation process is time-consuming, which is difficult to meet the needs of multi-region and multi-scheme comparison and selection. At the same time, the existing methods generally do not establish a modeling framework of systematic injection-production function relationship, which cannot effectively express the intervention intensity and topological structure between wells, so that the well spacing configuration decision cannot realize the combination of structure driving and intelligent optimization. SUMMARY

[0004] The application proposes a well spacing optimization method, device, equipment and medium based on injection-production function graph modeling. The application takes the injection-production relationship in the process of oil and gas reservoir development as the core, describes the coupling structure between wells by constructing an injection-production function graph, and on the basis of the graph structure, uses a pre-constructed response prediction model to predict the development effect under different well spacing configurations, further combines the target function definition and optimization strategy, and realizes the intelligent recommendation of the optimal well spacing parameter combination.

[0005] The application is implemented by the following technical solutions:

[0006] A well spacing optimization method based on injection-production function graph modeling, comprising:

[0007] establishing a corresponding injection-production function graph for a group of well spacing schemes;

[0008] inputting the injection-production function graph into a pre-established response prediction model to obtain a predicted value of a well group overall development response index;

[0009] calculating a target function value corresponding to the well spacing scheme according to the predicted value of the well group overall development response index;

[0010] selecting another set of well spacing schemes from the pre-set multiple sets of optional well spacing schemes, and returning to the injection-production effect map establishment step for the next cycle until the target function values corresponding to all well spacing schemes in the pre-set multiple sets of optional well spacing schemes are obtained;

[0011] sorting the target function values corresponding to all well spacing schemes to determine the optimal well spacing scheme.

[0012] In some embodiments, the injection-production effect map establishment process comprises:

[0013] collecting well group structure and injection-production response related data and preprocessing to construct structured time series data;

[0014] constructing each well in the well group as a node, taking the inter-well response relationship as an edge, and calculating the inter-well edge weight according to the historical production and injection data or geological relationship data to form an injection-production effect map.

[0015] In some embodiments, the response prediction model establishment process comprises:

[0016] taking the injection-production effect map as input to construct a graph neural network model; the input features of each node in the graph neural network model include well type, historical data statistics or geological parameters, the edge weight is used as the weight of graph convolution or attention calculation to model the inter-well intervention strength, and the model training target is to predict the overall development response index under different well group structures; the loss function is the mean square error of the predicted index and the actual observed value.

[0017] In some embodiments, the formula for calculating the target function value is:

[0018]

[0019] wherein, is the target function value, is the total production predicted by the model, is the energy consumption required for unit production predicted by the model; is the average water cut or unbalanced intervention index, is the task weight coefficient.

[0020] In some embodiments, the determination process of the optimal well spacing scheme comprises:

[0021] selecting the well spacing scheme corresponding to the maximum value in the target function value as the optimal well spacing scheme.

[0022] In some embodiments, the method further comprises:

[0023] outputting the optimal well spacing scheme and its corresponding injection-production effect map structure to other systems or platforms.

[0024] In a second aspect, the application provides an injection-production effect graph modeling-based well spacing optimization device, comprising:

[0025] a graph modeling unit configured to establish corresponding injection-production effect graphs for a group of well spacing schemes;

[0026] a prediction unit configured to input the injection-production effect graphs into a pre-established response prediction model to obtain well group overall development response index prediction values;

[0027] an evaluation unit configured to calculate target function values corresponding to the well spacing schemes according to the well group overall development response index prediction values;

[0028] a loop unit configured to determine whether the target function value calculation of all well spacing schemes in a pre-set group of selectable well spacing schemes is completed, and if yes, drive an optimization unit, and if not, select another group of well spacing schemes from the pre-set group of selectable well spacing schemes and drive the graph modeling unit to perform the next loop;

[0029] and an optimization unit configured to sort the target function values corresponding to all well spacing schemes to determine an optimal well spacing scheme.

[0030] In some embodiments, the device further comprises:

[0031] an output unit configured to output the optimal well spacing scheme and its corresponding injection-production effect graph structure to other systems or platforms.

[0032] In a third aspect, the application provides an electronic device comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the above-mentioned injection-production effect graph modeling-based well spacing optimization method when executing the computer program.

[0033] In a fourth aspect, the application provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program is executed by a processor to implement any of the above-mentioned injection-production effect graph modeling-based well spacing optimization methods.

[0034] The application provides an injection-production effect graph modeling-based well spacing optimization method, which takes the injection-production relationship in the process of oil and gas reservoir development as the core, constructs an injection-production effect graph, comprehensively describes the spatial structure and intervention mode among multiple well groups, and constructs a response prediction model based on the graph structure to predict the development effect under different well spacing configurations (i.e., well group structures), thereby breaking through the expression ability and efficiency bottleneck of traditional experience methods and numerical simulation methods in complex well group scenarios, further combining a target function and an optimization strategy to realize intelligent recommendation of an optimal well spacing scheme, and providing technical support for assisting well pattern design.

[0035] Correspondingly, the well spacing optimization device, the electronic device and the computer readable storage medium based on injection-production function diagram modeling provided by the present application also have the above technical effects. BRIEF DESCRIPTION OF DRAWINGS

[0036] The drawings described herein are used to provide further understanding of the embodiments of the present application, form a part of the present application, and do not constitute a limitation of the embodiments of the present application. In the drawings:

[0037] Figure 1 The well spacing optimization method flowchart provided for the embodiments of the present application;

[0038] Figure 2 The well spacing optimization device principle block diagram provided for the embodiments of the present application;

[0039] Figure 3 The well spacing optimization system architecture schematic diagram provided for the embodiments of the present application;

[0040] Figure 4 The electronic device schematic diagram provided for the embodiments of the present application;

[0041] Figure 5 The computer readable storage medium schematic diagram provided for the embodiments of the present application;

[0042] Figure 6 Injection-production function diagram after well spacing structure optimization for old wells;

[0043] Figure 7 New area well pattern injection-production topology structure diagram;

[0044] Reference signs and corresponding component names:

[0045] 200-well spacing optimization device, 201-diagram modeling unit, 202-prediction unit, 203-evaluation unit, 204-cyclic unit, 205-optimization unit, 206-output unit, 300-well spacing optimization system, 301-input device, 302-output device, 303-processor A, 304-memory A, 400-electronic device, 410-memory B, 420-processor B, 411-computer program A, 500-computer readable storage medium, 511-computer program B. DETAILED DESCRIPTION

[0046] Hereinafter, the term "include" or "may include" used in various embodiments of the present application indicates existence of the inventive function, operation, or element, and does not limit one or more additional functions, operations, or elements. Also, the terms "include", "have", and their conjugates as used in various embodiments of the present application, are merely intended to denote specific features, numbers, steps, operations, elements, components, or combinations thereof, and are not intended to, in and of themselves, exclude the existence of one or more other features, numbers, steps, operations, elements, components, or combinations thereof, unless clearly so defined in the various embodiments of the present application.

[0047] In various embodiments of the present application, the expression "or" or "at least one of A or / and B" includes any and all combinations of the listed terms. For example, the expression "A or B" or "at least one of A or / and B" can include A, can include B, or can include both A and B.

[0048] The expressions, such as "first", "second", and the like, used in various embodiments of the present application can modify various components in various embodiments, but can not limit the corresponding components. For example, the above expressions do not limit the order and / or importance of the elements. The above expressions are used merely for the purpose of distinguishing a component from other components. For example, a first user device and a second user device indicate different user devices, although both are user devices. For example, a first element can be termed a second element, and likewise, a second element can be termed a first element, without departing from the scope of various embodiments of the present application.

[0049] It should be noted that if a description describes one component "connected to" another component, a first component can be directly connected to a second component, and "connected" to the second component via a third component. Conversely, if a description describes one component "directly connected to" another component, it should be understood that there are no third components between the first component and the second component.

[0050] The terms used in various embodiments of the present application are used only to describe specific embodiments and are not intended to limit various embodiments of the present application. As used herein, the singular forms are intended to include the plural forms as well, unless the context clearly indicates otherwise. Unless otherwise defined, all terms used herein (including technical terms and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which various embodiments of the present application belong. The terms, such as those defined in a generally used dictionary, are to be interpreted as having a meaning that is the same as, or similar to, the contextual meaning of the relevant technical field, and are not to be interpreted as having an idealized or overly formal meaning, unless expressly so defined in various embodiments of the present application.

[0051] For the purpose, technical solutions and advantages of the present application to be clearer, the present application is further described in detail below with reference to the embodiments and drawings. The illustrative embodiments and their descriptions are only used to explain the present application and do not limit the present application.

[0052] Embodiment 1:

[0053] The embodiment of the present application proposes a well spacing optimization method based on injection-production function diagram modeling, aiming at the injection-production response prediction and well spacing parameter optimization tasks in well pattern structure design, and constructing an efficient and structured decision support scheme.

[0054] As shown in Figure 1 The well spacing optimization method proposed by the embodiment of the present application includes the following steps:

[0055] Step 110, a corresponding injection-production function diagram is established for a group of well spacing schemes;

[0056] Step 120, the injection-production function diagram is input into a pre-established response prediction model to obtain a predicted value of a well group overall development response index;

[0057] Step 130, a target function value corresponding to the well spacing scheme is calculated according to the predicted value of the well group overall development response index;

[0058] Step 140, another group of well spacing schemes is selected from a plurality of pre-set optional well spacing schemes, and the next cycle is returned to step 110 until the target function values corresponding to all well spacing schemes in the plurality of pre-set optional well spacing schemes are obtained;

[0059] Step 150, the target function values corresponding to all well spacing schemes are sorted to determine the optimal well spacing scheme.

[0060] Further, in step 120 of the embodiment of the present application, the establishment process of the injection-production function diagram includes:

[0061] Collect and preprocess the well group structure and injection-production response related data, construct structured time series data, and the input basis of injection-production effect graph. Specifically, for different optimization tasks, the collected related data is different. For example, in the old area well pattern optimization task, the injection-production well production data including injection rate, production rate, wellhead pressure, casing pressure, water cut and other key parameters are extracted from the historical database; in the new area well pattern design task, combined with the geological modeling results, spatial (geological relationship) parameters such as reservoir thickness, permeability, structural boundary, fault position and candidate well location coordinates are extracted. The historical data is cleaned and standardized, mainly including missing value filling (linear interpolation method can be used), outlier removal (Z-Score outlier detection method can be used), uniform sampling period (daily average resampling method can be used), etc. The output is a standard structured data set for subsequent graph model construction.

[0062] Construct an injection-production effect graph. Specifically, the well is taken as a node, and the response relationship between wells is taken as an edge to construct a graph model structure that can express the injection-production coupling strength. Each well is constructed as a node , and the edge weight between wells is calculated according to the historical production and injection data or the geological relationship data . The edge weight definition is as follows:

[0063] For the well pair with historical data , the response strength between the injection well and the production well is calculated:

[0064]

[0065] Wherein, is the delay correlation coefficient between injection and production, are the time series data of the injection well and the production well , respectively, is the response time lag decay function, is the sampling interval.

[0066] For new candidate wells, the edge weight is defined as a geological distance weighted function or a structure connectivity score.

[0067] Finally, the injection-production effect graph is formed and represented as , wherein, is the node set, is the edge weight matrix.

[0068] Further, in step 130 of the embodiment of the present application, the establishment process of the response prediction model includes:

[0069] On the basis of the injection-production effect graph, a graph neural network model is constructed, and a response prediction model is trained. Specifically, the injection-production effect graph is taken as the input, and a graph neural network model is constructed, which can use GCN (Graph Convolution Networks), GAT (Graph Attention Networks), etc. The input features of each node include well type (injection / production), historical data statistics (such as daily average, coefficient of variation, etc.) or geological parameters. The edge weight is used as the weight of graph convolution or attention calculation, and the intervention intensity between wells is modeled. The training target of the model is to predict the overall development response index under different well group structures, such as unit block production, energy consumption, water cut, etc. The loss function is the mean square error of the predicted index and the actual observed value, which is expressed as:

[0070]

[0071] wherein, is the loss function value, is the predicted production of the model on the th day, is the actual measured value of the production on the th day, is the predicted energy consumption of the model on the th day, is the actual measured value of the energy consumption on the th day, is the predicted water cut or unbalanced intervention index of the model on the th day, is the actual measured value of the water cut on the th day or the actual measured value of the unbalanced intervention index, is the prediction length (in days), all are weight parameters.

[0072] Further, in step 140 of the embodiments of the present application, the well spacing scheme is a parameter combination , that is, the well spacing configuration between each group of injection-production wells, wherein, is the well spacing of any two wells in the well pattern. For each well spacing scheme, the objective function value calculation formula is:

[0073]

[0074] wherein, is the objective function value, is the total production predicted by the model, is the energy consumption required for unit production predicted by the model; is the average water cut or unbalanced intervention index, is the task weight coefficient, reflecting the priority of different objectives. For example, according to the objective requirement, if the yield optimization is mainly targeted, then max, is smaller or set to 0, that is, the objective function supports single-objective extremum, weighted combination and other optimization modes, and has good expansibility and scene adaptability.

[0075] It should be noted that the optional well spacing scheme generation process needs to meet the following constraints: That is, the well spacing between any two wells must meet the constraints of minimum progress and maximum well spacing, and also meet the total number of injection wells ratio requirement.

[0076] Further, in step 160 of the embodiment of the present application, the objective function values corresponding to all well spacing schemes are sorted, and the well spacing scheme corresponding to the maximum value is preferably output as the optimal well spacing scheme.

[0077] Further, the well spacing optimization method proposed by the embodiment of the present application further includes:

[0078] The optimal well spacing scheme and its corresponding injection-production action graph are output in a structured manner. Optionally, the objective function value and the predicted index corresponding to each group of well spacing schemes are also output. Optionally, well pattern layout sketches and node topology graphs are also output.

[0079] Further, the well spacing optimization method proposed by the embodiment of the present application further includes:

[0080] The output data in the optional export format such as JSON / XML structure or platform interface protocol supports docking with well pattern design systems, digital well site platforms or graph databases to form visual optimization results and engineering parameter support data.

[0081] The well spacing optimization method proposed by the embodiment of the present application takes the injection-production relationship in the process of oil and gas reservoir development as the core, describes the inter-well coupling relationship by constructing an injection-production action graph, and integrates graph modeling technology and data-driven models to break through the expression ability and efficiency bottleneck of traditional experience methods and numerical simulation methods in complex well group scenarios. Among them, the injection-production action graph takes a single well as a node, the injection-production coupling relationship between wells as an edge, and the edge weight reflects the actual intervention strength or historical production-injection synergy characteristics. The spatial structure and intervention mode of multiple well groups can be described comprehensively. On the basis of this graph structure, a response prediction model is constructed to predict the development effect under different well spacing configurations. Further, combined with the objective function and the optimization strategy, the intelligent recommendation of the optimal well spacing parameter combination is realized.

[0082] Based on the same technical concept described above, the embodiment of the present application also proposes a well spacing optimization device based on injection-production action graph modeling, as shown in Figure 2 The well spacing optimization device 200 includes:

[0083] The graph modeling unit 201 is configured to establish a corresponding injection-production action graph for each well spacing scheme. The establishment of the injection-production action graph is as described in the step 110 above, and thus will not be repeated here.

[0084] The prediction unit 202 is configured to input the injection-production action graph into a pre-established response prediction model to obtain a predicted value of a well group overall development response index. The establishment of the response prediction model is as described in the step 120 above, and thus will not be repeated here.

[0085] The evaluation unit 203 is configured to calculate a target function value corresponding to the well spacing scheme according to the predicted value of the well group overall development response index. The calculation method of the target function value is as described in the step 130 above, and thus will not be repeated here.

[0086] The cycle unit 204 is configured to determine whether the calculation of the target function value of all well spacing schemes in the pre-set multiple groups of selectable well spacing schemes is completed, and if yes, drive the optimization unit, and if not, select another group of well spacing schemes from the pre-set multiple groups of selectable well spacing schemes and drive the graph modeling unit to perform the next cycle.

[0087] In addition, the optimization unit 205 is configured to sort the target function values corresponding to all well spacing schemes to determine an optimal well spacing scheme. The specific optimization process is as described in the step 150 above, and thus will not be repeated here.

[0088] Further, the well spacing optimization apparatus 200 proposed in the embodiments of the present application further comprises:

[0089] The output unit 206 is configured to structurally output the optimal well spacing scheme and the injection-production action graph corresponding thereto. Optionally, the output unit 206 is further configured to output the target function value, the predicted index, the output well pattern sketch, the node topology graph and the like corresponding to each group of well spacing schemes. Optionally, the output unit 206 exports the data to be output in a selectable format such as a JSON / XML structure or a platform interface protocol. The output structure supports the connection with a well pattern design system, a digital well site platform or a graph database to form visual optimization results and engineering parameter support data.

[0090] Based on the same technical concept, the embodiments of the present application further propose a well spacing optimization system based on injection-production action graph modeling, as shown in Figure 3 The well spacing optimization system 300 proposed in the embodiments of the present application comprises:

[0091] An input device 301, an output device 302, a processor A 303 and a memory A 304; wherein the number of the processor A 303 and the memory A 304 can be one or more, Figure 3The following description uses a processor A303 and a memory A304 as an example. The input device 301, output device 302, processor A303, and memory A304 can be connected via a bus or other means. Figure 3 Taking the example of a connection between China and Israel via a bus.

[0092] Specifically, by calling the operation instructions stored in memory A304, processor A303 executes the following steps:

[0093] Establish corresponding injection-production action diagrams for a set of well spacing schemes;

[0094] The injection-production action diagram is input into a pre-established response prediction model to obtain the predicted values ​​of the overall development response index of the well group;

[0095] Based on the predicted values ​​of the overall development response indicators of the well group, the objective function value corresponding to the well spacing scheme is calculated.

[0096] Select another well spacing scheme from the pre-set multiple sets of optional well spacing schemes, and return to the injection-production action diagram establishment step to perform the next loop until the objective function values ​​corresponding to all well spacing schemes in the pre-set multiple sets of optional well spacing schemes are obtained;

[0097] The objective function values ​​corresponding to all well spacing schemes are sorted to determine the optimal well spacing scheme.

[0098] Optionally, by calling the operation instructions stored in memory A304, processor A303 is also used to execute any of the embodiments in the corresponding examples of the above-described well spacing optimization method.

[0099] Based on the same technical concept described above, this application also proposes an electronic device, such as... Figure 4 As shown, the electronic device 400 includes: a memory B410, a processor B420, and a computer program A411 stored in the memory B410 and executable on the processor B420. When the processor B420 executes the computer program A411, it performs the following steps:

[0100] Establish corresponding injection-production action diagrams for a set of well spacing schemes;

[0101] The injection-production action diagram is input into a pre-established response prediction model to obtain the predicted values ​​of the overall development response index of the well group;

[0102] Based on the predicted values ​​of the overall development response indicators of the well group, the objective function value corresponding to the well spacing scheme is calculated.

[0103] select another set of well spacing scheme from the pre-set multiple sets of optional well spacing schemes, and return to the injection-production effect map establishing step for the next cycle until the target function values corresponding to all well spacing schemes in the pre-set multiple sets of optional well spacing schemes are obtained;

[0104] The target function values corresponding to all well spacing schemes are sorted to determine the optimal well spacing scheme.

[0105] Optionally, when the processor B 420 executes the computer program A 411, any of the implementation manners in the corresponding embodiments of the well spacing optimization method described above can be implemented.

[0106] It should be noted that the electronic device proposed in the embodiments of the present application is a device used to implement the well spacing optimization method described above, and therefore based on the well spacing optimization method described above proposed in the embodiments of the present application, those skilled in the art can understand the specific implementation manners of the electronic device of the embodiments of the present application and various forms of changes thereof, and therefore the specific implementation manners of the electronic device for implementing the well spacing optimization method described above will not be introduced in detail here, as long as the electronic device used to implement the well spacing optimization method described above is implemented by those skilled in the art, it belongs to the scope of protection intended by the present application.

[0107] Based on the same technical concept described above, the embodiments of the present application further propose a computer readable storage medium, as shown in the Figure 5 The computer readable storage medium 500 stores a computer program B 511, and the computer program B 511 is executed by a processor to implement the following steps:

[0108] establishing a corresponding injection-production effect map for a set of well spacing schemes;

[0109] inputting the injection-production effect map into a pre-established response prediction model to obtain a predicted value of a well group overall development response index;

[0110] calculating a target function value corresponding to the well spacing scheme according to the predicted value of the well group overall development response index;

[0111] selecting another set of well spacing scheme from the pre-set multiple sets of optional well spacing schemes, and returning to the injection-production effect map establishing step for the next cycle until the target function values corresponding to all well spacing schemes in the pre-set multiple sets of optional well spacing schemes are obtained;

[0112] The target function values corresponding to all well spacing schemes are sorted to determine the optimal well spacing scheme.

[0113] Optionally, the computer program B 511 can implement any of the implementation manners in the corresponding embodiments of the well spacing optimization method described above when executed by the processor.

[0114] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and the parts not described in detail in a certain embodiment can be referred to the related description of other embodiments.

[0115] Embodiment 2:

[0116] The well spacing optimization method proposed in Embodiment 1 is applied to the well spacing structure optimization of a typical old well group. In the typical well network area of the well group, the well group connection relationship is complex, and the production of some production wells continues to decline. In order to identify the structural well spacing problem and optimize the injection-production configuration structure, and improve the overall productivity and injection utilization rate. The specific implementation process is as follows:

[0117] First, the production data of the past 180 days is obtained from the production database of the area where the well group is located, including injection rate, production, wellhead pressure, water cut and other key indicators, covering 12 injection wells and 18 production wells. After data sliding average noise reduction, missing value interpolation and unit normalization processing, standard structured time series data is formed, and the sampling interval is 1 day.

[0118] Then, the injection-production action graph is constructed. 30 wells are taken as graph nodes, and the graph edges are constructed according to the injection-production correlation, and the edge weight is defined as the 5-day lag correlation coefficient between injection wells and production wells multiplied by the time delay attenuation factor. Finally, an injection-production action graph with about 140 edges is obtained, and the average connectivity is 4.7.

[0119] Then, the response prediction model is used to predict the future 7-day production and unit energy consumption of the well group. The response prediction model is based on the graph convolution network structure, and the node features include well type, daily production mean and coefficient of variation, and the edge weight is defined by historical injection-production response. The training target is to predict the future 7-day production and unit energy consumption of the well group. The loss function is:

[0120]

[0121] Then, according to the model prediction result, the target function value is calculated by the following formula:

[0122] The engineering constraints are set as: , keep the injection well ratio between 30%-40%.

[0123] Repeat the above process to get the target function values of 40 candidate well spacing schemes, and sort them, and finally select an optimal well spacing scheme (with the highest target function value).

[0124] The optimal well spacing scheme improves the predicted production by about 7.1% compared with the current scheme, and reduces the unit energy consumption by 3.9%.

[0125] The optimization results are output in the form of well spacing matrix and injection-production action graph structure, and the well spacing matrix and injection-production action graph structure are generatedFigure 6 The structure topology diagram is shown. The injection wells are represented in red and the production wells are represented in blue in the topology diagram, and the edge width represents the injection-production effect intensity, which intuitively shows the coupling relationship of the well group after optimization. The diagram is submitted to the well pattern design system as an optimization suggestion to assist in forming a structure adjustment proposal on site.

[0126] Example 3:

[0127] In the planning stage of a new block, a group of injection-production well layout schemes are planned to be deployed to improve the late-stage connectivity efficiency and productivity potential. In order to avoid relying on subjective experience and simplified assumptions, the well spacing optimization method proposed in Example 1 above is used to evaluate and screen multiple candidate well combination, and the specific implementation is as follows:

[0128] First, the regional geological modeling results and the coordinates of 16 candidate well sites are obtained, including the expected functions (injection / production), formation information, and fault boundary distance. Based on the regional permeability field and the structure layer data, the interwell geological connectivity factor is constructed, which is uniformly converted into a standard input feature matrix.

[0129] Then, the injection-production effect graph is constructed. The nodes are 16 candidate wells, and the edges are defined according to the well spacing and the connectivity strength The edge weight is defined as:

[0130]

[0131] The number of edges in the injection-production effect graph is 68, and the average connectivity of the nodes is 8.5. The edge weight represents the potential effect under the control of geology.

[0132] After that, the response prediction model is used to predict the overall potential production and structure balance index (imbalance intervention index) of the well group. The response prediction model can use a graph attention network structure, and the node features include formation thickness, well type, fault proximity factor, etc. The model aims to output the potential production and structure balance index corresponding to each well spacing combination parameter.

[0133] After that, according to the model prediction results, the objective function value is calculated using the following formula:

[0134]

[0135] Where, is the structure response standard deviation (reflecting the unevenness of injection-production).

[0136] Repeat the above process to obtain the objective function values of 30 candidate well spacing schemes, and sort them. Finally, the well spacing scheme with the highest objective function value is selected as the optimal well spacing scheme.

[0137] The optimal well spacing scheme improves the predicted production by about 8.5% compared to the average configuration, and the balance index is improved by 22%.

[0138] The new zone well spacing optimization result is derived, and an injection-production topology structure diagram as shown in Figure 7 is generated. The diagram adopts a "central injection well, peripheral production well" structure mode, combines the injection-production effect strength of edge width visualization, and directly reflects the injection-production connectivity pattern under the designed well spacing. The diagram structure and the target function value are submitted to the well pattern design system as a new well deployment scheme reference.

[0139] Those skilled in the art will understand that embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage, etc.) containing computer usable program code.

[0140] The present application is described with reference to flowcharts and / or block diagrams according to the methods, devices (systems), and computer program products of embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and combinations of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing apparatus to produce a machine, so that the instructions, which are executed via the processor of the computer or other programmable data processing apparatus, generate a means for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 The flow or flows and / or blocks in the flowcharts and / or block diagrams Figure 1 The means for performing the functions specified in the flowcharts and / or block diagrams.

[0141] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including instruction means, which implement the functions specified in the flowcharts and / or block diagrams. Figure 1 The flow or flows and / or blocks in the flowcharts and / or block diagrams Figure 1 The means for performing the functions specified in the flowcharts and / or block diagrams.

[0142] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus, so that a series of operation steps are performed on the computer or other programmable data processing apparatus to produce a computer-implemented process, so that the instructions executed on the computer or other programmable data processing apparatus provide a means for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 The flow or flows and / or blocks in the flowcharts and / or block diagrams Figure 1steps of the functions specified in the one or more blocks.

[0143] The above detailed description has been presented for the purposes of simplicity and clarity. It is not intended to be an exhaustive description of the application. Those skilled in the art will recognize that many modifications and changes can be made to the specific implementation described without departing from the scope of the application. Thus, the scope of the application should not be limited by any of the above described embodiments.

Claims

1. A well spacing optimization method based on injection-production interaction diagram modeling, characterized in that, include: Establish corresponding injection-production action diagrams for a set of well spacing schemes; The injection-production action diagram is input into a pre-established response prediction model to obtain the predicted values ​​of the overall development response index of the well group. Based on the predicted values ​​of the overall development response indicators of the well group, the objective function value corresponding to the well spacing scheme is calculated. Select another well spacing scheme from the pre-set multiple sets of optional well spacing schemes, and return to the injection-production action diagram establishment step to perform the next loop until the objective function values ​​corresponding to all well spacing schemes in the pre-set multiple sets of optional well spacing schemes are obtained; Sort the objective function values ​​corresponding to all well spacing schemes to determine the optimal well spacing scheme; The process of establishing the injection-production action map includes: Collect well group structure and injection-production response related data and preprocess them to construct structured time series data; Each well in the well group is constructed as a node, and the inter-well response relationship is used as the edge. The inter-well edge weight is calculated based on historical production and injection data or geological relationship data to form an injection and production action map. For well pairs with existing historical data, the response intensity between the injection well and the production well is calculated as the well boundary weight; For candidate wells in the new area, the inter-well boundary weights are defined as a geological distance weighting function or a structural connectivity score; the process of establishing the response prediction model includes: A graph neural network model is constructed using the injection-production action map as input. The input features of each node in the graph neural network model include well type, historical data statistics or geological parameters. The edge weights are used as weights for graph convolution or attention calculation to model the intensity of inter-well intervention. The training objective of the model is to predict the overall development response index under different well group structures. The loss function is the mean square error between the predicted index and the actual observed value.

2. The well spacing optimization method based on injection-production interaction diagram modeling according to claim 1, characterized in that, The formula for calculating the objective function value is as follows: ； in, The objective function value, The total output predicted by the model, The energy consumption required per unit of output as predicted by the model; The average moisture content or imbalance intervention index, This represents the task weighting coefficient.

3. The well spacing optimization method based on injection-production interaction diagram modeling according to claim 1, characterized in that, The process of determining the optimal well spacing scheme includes: The well spacing scheme corresponding to the maximum value among the objective function values ​​is selected as the optimal well spacing scheme.

4. A well spacing optimization method based on injection-production interaction diagram modeling according to any one of claims 1-3, characterized in that, Also includes: The optimal well spacing scheme and its corresponding injection-production action diagram are structured and output to other systems or platforms.

5. A well spacing optimization device based on injection-production interaction diagram modeling, characterized in that, include: The graph modeling unit is used to create a corresponding injection-production interaction diagram for a set of well spacing schemes; The prediction unit is used to input the injection-production action diagram into a pre-established response prediction model to obtain the predicted value of the overall development response index of the well group; The evaluation unit is used to calculate the objective function value corresponding to the well spacing scheme based on the predicted value of the overall development response index of the well group. The loop unit is used to determine whether the objective function value calculation of all the pre-set multiple sets of optional well spacing schemes has been completed. If so, the optimization unit is driven; otherwise, another set of well spacing schemes is selected from the pre-set multiple sets of optional well spacing schemes and the graph modeling unit is driven to perform the next loop. And, an optimization unit, used to sort the objective function values ​​corresponding to all well spacing schemes and determine the optimal well spacing scheme; The process of establishing the injection-production action map includes: Collect well group structure and injection-production response related data and preprocess them to construct structured time series data; Each well in the well group is constructed as a node, and the inter-well response relationship is used as the edge. The inter-well edge weight is calculated based on historical production and injection data or geological relationship data to form an injection and production action map. For well pairs with existing historical data, the response intensity between the injection well and the production well is calculated as the well boundary weight; For candidate wells in the new area, the inter-well boundary weight is defined as a geological distance weighting function or a structural connectivity score; The process of establishing the response prediction model includes: A graph neural network model is constructed using the injection-production action map as input. The input features of each node in the graph neural network model include well type, historical data statistics or geological parameters. The edge weights are used as weights for graph convolution or attention calculation to model the intensity of inter-well intervention. The training objective of the model is to predict the overall development response index under different well group structures. The loss function is the mean square error between the predicted index and the actual observed value.

6. The well spacing optimization device based on injection-production interaction diagram modeling according to claim 5, characterized in that, Also includes: The output unit is used to output the optimal well spacing scheme and its corresponding injection-production action diagram in a structured manner to other systems or platforms.

7. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the well spacing optimization method based on injection-production action diagram modeling as described in any one of claims 1-4.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the well spacing optimization method based on injection-production action diagram modeling as described in any one of claims 1-4.

Images