A method and device for determining the downhole leakage rate

Abstract

This specification provides a method and device for determining the downhole leakage rate. The method includes: obtaining the depth-domain seismic data volume of the target well and the engineering parameters of the drilled wells; extracting the variance attribute of the depth of the drilled well trajectory according to the depth-domain seismic data volume and the engineering parameters of the drilled wells; extracting the variance value of the target depth of the drilled well according to the variance attribute; forming a planar graph of the variance value of the target depth according to the variance value of the target depth; determining the connectivity index of the drilled well according to the planar graph of the variance value of the target depth; training a full-feedback neural network model according to the engineering parameters of the drilled wells, the variance value of the target depth of the drilled well, and the connectivity index of the drilled well to obtain a leakage rate prediction model; obtaining the dataset of the wells to be drilled in the target well, and inputting the dataset of the wells to be drilled into the leakage rate prediction model to obtain the leakage rate of the wells to be drilled in the target well. Based on the above method, the downhole leakage rate can be predicted accurately and reliably.

Description

Technical Field

[0001] This specification belongs to the technical field of oil and gas drilling operations, and particularly relates to a method and device for determining the downhole leakage rate. Background Art

[0002] Downhole leakage not only causes losses of drilling fluid materials, but also takes up a large amount of drilling time, and at the same time seriously threatens drilling safety. Accurately predicting and judging the potential leakage risk of drilling and quantifying the leakage risk are crucial for safe and efficient drilling. However, based on the existing technology, it is impossible to accurately and reliably quantify the leakage risk of discontinuous formations (such as fractured leakage), so it is impossible to effectively guide drilling development work.

[0003] In response to the above problems, no effective solution has been proposed yet. Summary of the Invention

[0004] This specification provides a method and device for determining the downhole leakage rate, which can accurately and reliably determine the downhole leakage rate.

[0005] On the one hand, an embodiment of this specification provides a method for determining the downhole leakage rate, including:

[0006] Obtain the depth-domain seismic data volume and the drilled well engineering parameters of the target well;

[0007] Extract the variance attribute of the drilled well trajectory depth according to the depth-domain seismic data volume and the drilled well engineering parameters;

[0008] Extract the variance value of the target depth of the drilled well according to the variance attribute;

[0009] Form a planar graph of the variance value of the target depth according to the variance value of the target depth;

[0010] Determine the connectivity index of the drilled well according to the planar graph of the variance value of the target depth;

[0011] Train a full-feedback neural network model according to the drilled well engineering parameters, the variance value of the target depth of the drilled well, and the connectivity index of the drilled well to obtain a leakage rate prediction model;

[0012] Obtain the dataset of the well to be drilled in the target well, and input the dataset of the well to be drilled into the leakage rate prediction model to obtain the leakage rate of the well to be drilled in the target well.

[0013] Further, before obtaining the depth-domain seismic data volume of the target well, it further includes:

[0014] Obtain the seismic data volume, drilling data, logging data, and mud logging data of the target well in the seismic work area; among them, the drilling data of the drilled well includes wellhead coordinates and wellbore trajectory data;

[0015] Convert the seismic data volume according to the drilling data, logging data, and mud logging data to obtain the depth-domain seismic data volume of the target well.

[0016] Further, extracting the variance attribute of the drilled well trajectory depth according to the depth-domain seismic data volume and the drilled well engineering parameters includes:

[0017] Extract the variance attribute data volume of the depth-domain seismic data volume;

[0018] Import the wellhead coordinates and the wellbore trajectory data into the depth-domain seismic data volume to form target trajectory data;

[0019] Extract the variance attribute of the variance attribute data volume along the target trajectory data to form a two-dimensional array of depth and variance;

[0020] Take the two-dimensional array of depth and variance as the variance attribute of the drilled well trajectory depth.

[0021] Further, extract the variance value of the target depth of the drilled well according to the following formula:

[0022] V≥(V max -V min )×0.5

[0023] where V is the variance value of the depth to be extracted, V max is the maximum variance along the well trajectory, and V min is the minimum variance along the well trajectory.

[0024] Further, before determining the connectivity index of the drilled well according to the planar graph of the variance value of the target depth, it also includes:

[0025] Obtain the wellbore and multiple fractures in the drilled well;

[0026] Determine the degree of connection between the wellbore and each fracture among the multiple fractures, the degree of connection between each fracture among the multiple fractures, and the development degree of each fracture among the multiple fractures;

[0027] Obtain the relative positions of the wellbore and each fracture among the multiple fractures;

[0028] Form an imaging set of the wellbore and the relative positions according to the wellbore and the relative positions;

[0029] Determine a connectivity determination criterion based on the relative position imaging set, the degree of intersection between the wellbore and each fracture among the multiple fractures, the degree of intersection between each pair of fractures among the multiple fractures, and the development degree of each fracture among the multiple fractures, where the connectivity determination criterion is used to obtain a connectivity index of the drilled well in combination with the planar graph.

[0030] Furthermore, the drilled well engineering parameters include at least one of the following: drilling fluid density, bottom hole circulating pressure loss, formation pore pressure, hole size of each section of the drilled well, depth of the drilled well, depth of lost circulation in the drilled well, and lost circulation rate.

[0031] Furthermore, training the full-feedback neural network model according to the drilled well engineering parameters, the variance value of the target depth of the drilled well, and the connectivity index of the drilled well to obtain a lost circulation rate prediction model includes:

[0032] Calculate the pressure difference according to the drilling fluid density, the bottom hole circulating pressure loss, and the formation pore pressure according to the following formula:

[0033] ΔP = P s + P anu - P pore

[0034] where ΔP is the difference between the hydrostatic pressure of the liquid column in the wellbore and the formation pore pressure at a certain depth of the wellbore, P s is the static hydrostatic pressure of the drilling fluid with a specific density in the non-circulating state, P anu is the bottom hole circulating pressure loss, and P pore is the formation pore pressure;

[0035] Train the full-feedback neural network model according to the pressure difference, the variance value of the target depth of the drilled well, and the connectivity index of the drilled well to obtain a lost circulation rate prediction model.

[0036] Furthermore, train the full-feedback neural network model according to the following formula:

[0037] LR i = f(ΔP i , V i , N Index )

[0038] where LR i is the lost circulation rate at depth i, ΔP i is the difference between the hydrostatic pressure of the liquid column in the wellbore and the formation pore pressure at depth i, V i is the variance value at depth i, and N Index is the connectivity index at depth i.

[0039] On the other hand, the embodiments of the present specification further provide a device for determining the downhole leakage rate, including:

[0040] An acquisition module, configured to acquire the depth-domain seismic data volume of the target well and the engineering parameters of the drilled well;

[0041] A variance extraction module, configured to extract the variance attribute of the depth of the drilled well trajectory according to the depth-domain seismic data volume and the engineering parameters of the drilled well; and extract the variance value of the target depth of the drilled well according to the variance attribute;

[0042] A connectivity index extraction module, configured to form a planar graph of the variance value of the target depth according to the variance value of the target depth; and determine the connectivity index of the drilled well according to the planar graph of the variance value of the target depth;

[0043] A model training module, configured to train a full-feedback neural network model according to the engineering parameters of the drilled well, the variance value of the target depth of the drilled well, and the connectivity index of the drilled well, to obtain a leakage rate prediction model;

[0044] An output module, configured to acquire the dataset of the well to be drilled in the target well, input the dataset of the well to be drilled into the leakage rate prediction model, and obtain the leakage rate of the well to be drilled in the target well.

[0045] On yet another aspect, the present application further provides a computer-readable storage medium, on which computer instructions are stored, and when the computer-readable storage medium executes the instructions, the above method for determining the downhole leakage rate is implemented.

[0046] A method and device for determining the downhole leakage rate provided in this specification. First, obtain the depth-domain seismic data volume of the target well and the engineering parameters of the drilled wells. The engineering parameters of the drilled wells are obtained through an objective method, reducing the errors caused by human subjective judgment. Secondly, according to the depth-domain seismic data volume and the engineering parameters of the drilled wells, extract the variance attribute of the depth of the drilled well trajectory. According to the variance attribute, extract the variance value of the target depth of the drilled well. By obtaining the variance value of the target depth, it can lay a data foundation for accurately and quickly obtaining the connectivity index in the subsequent steps. Further, according to the variance value of the target depth, form a planar graph of the variance value of the target depth. According to the planar graph of the variance value of the target depth, determine the connectivity index of the drilled well. By obtaining the connectivity index, it can characterize the channels and reservoir spaces, thus laying a foundation for obtaining a leakage rate prediction model that meets the engineering requirements. Further, train a full-feedback neural network model according to the engineering parameters of the drilled wells, the variance value of the target depth of the drilled well, and the connectivity index of the drilled well to obtain a leakage rate prediction model. Through the above training method, the problems of difficult parameter acquisition, many assumed variables, and large prediction errors in traditional analytical methods can be overcome. Finally, obtain the dataset of the wells to be drilled in the target well, and input the dataset of the wells to be drilled into the leakage rate prediction model to obtain the leakage rate of the wells to be drilled in the target well. Through the above solution, the downhole leakage rate can be accurately determined. Description of the Drawings

[0047] To more clearly illustrate the embodiments of this specification, the drawings required for use in the embodiments will be briefly introduced below. The drawings described below are only some embodiments recorded in this specification. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0048] Figure 1 It is a schematic flowchart of the method for determining the downhole leakage rate provided by an embodiment of this specification;

[0049] Figure 2 It is a schematic diagram of an embodiment of applying the method for determining the downhole leakage rate provided by an embodiment of this specification in a scenario example;

[0050] Figure 3 It is a schematic diagram of an embodiment of applying the method for determining the downhole leakage rate provided by an embodiment of this specification in a scenario example;

[0051] Figure 4 It is a schematic diagram of the structural composition of the device for determining the downhole leakage rate provided by an embodiment of this specification;

[0052] Figure 5It is a schematic diagram of the structural composition of an electronic device provided by an embodiment of this specification. Detailed implementation manners

[0053] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of this specification. Obviously, the described embodiments are only a part of the embodiments of this specification, rather than all the embodiments. Based on the embodiments in this specification, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of this specification.

[0054] Considering that since the birth of oil drilling, downhole losses have always been a problem plaguing the industry. Downhole losses, especially severe losses, not only cause losses of drilling fluid materials, but also take up a large amount of drilling time, resulting in economic losses and posing a serious threat to drilling safety. Therefore, accurately predicting and judging the potential loss risk of drilling and quantifying the loss risk are crucial for safe and efficient drilling. At the current stage, the industry mostly uses the loss pressure to evaluate the loss risk, that is, it is considered that when the liquid column pressure in the wellbore reaches a certain value, formation losses will occur, and the pressure value at this time is the loss pressure. The application of traditional loss pressure faces various difficulties. It can neither characterize the loss rate (scale) of drilling fluid nor accurately determine the critical pressure value for the occurrence of losses in discontinuous formations (such as fractured losses).

[0055] Furthermore, considering the continuous advancement of oil and gas exploration and development in recent years, two other methods for analyzing and predicting downhole losses have gradually emerged based on the concepts of machine learning and geological engineering integration. Among them, the method for predicting downhole losses based on machine learning mainly uses the actual engineering data and actual downhole loss situations that have occurred in drilled wells, trains the data set using corresponding machine learning algorithms, and then uses the trained model to analyze the loss of the predicted well. The biggest problem with this method is that it simply divides the downhole loss process into the result of the action of engineering parameters, completely separating the core elements of loss occurrence such as downhole loss channels and drilling fluid storage spaces. Therefore, when this method is applied outside the oil and gas fields to which the training data set belongs, it is difficult to obtain accurate and reliable prediction results. The analysis and prediction method based on the concept of geological engineering integration mainly conducts seismic profile analysis for the target oil and gas field, and qualitatively analyzes the drilling loss risk by identifying the unconformity information of the designed drilling track passing through the seismic profile. Due to the multi-solution nature of seismic profile interpretation based on graphics, it is difficult to guarantee the prediction accuracy of downhole loss risk, and the influence of engineering parameters is not considered in the existing methods. Therefore, the application scope of the existing methods has limitations, and the application effect is difficult to meet the engineering requirements.

[0056] In summary, the currently known methods for predicting the leakage rate have the following problems: The traditional leakage pressure analysis method is based on the theory of continuous formation overcoming the horizontal minimum principal in-situ stress, which is not suitable for analyzing the fracture leakage risk and scale, and cannot give a quantitative prediction result of the leakage rate; while the analysis methods based on machine learning or geological engineering integration do not organically combine the three elements of leakage occurrence (leakage driving force element, channel element, and reservoir space element) according to the leakage occurrence process.

[0057] Based on the above ideas, this specification proposes a method for determining the downhole leakage rate. First, obtain the depth-domain seismic data volume and the engineering parameters of the drilled wells for the target well; secondly, according to the depth-domain seismic data volume and the engineering parameters of the drilled wells, extract the variance attribute of the depth of the drilled well trajectory, and according to the variance attribute, extract the variance value of the target depth of the drilled well; further, according to the variance value of the target depth, form a planar graph of the variance value of the target depth, and according to the planar graph of the variance value of the target depth, determine the connectivity index of the drilled well; further, train a full-feedback neural network model according to the engineering parameters of the drilled wells, the variance value of the target depth of the drilled well, and the connectivity index of the drilled well to obtain a leakage rate prediction model; finally, obtain the data set of the wells to be drilled for the target well, and input the data set of the wells to be drilled into the fracture leakage rate prediction model to obtain the leakage rate of the wells to be drilled in the target well. Refer to Figure 1 As shown, the embodiments of this specification provide a method for determining the downhole leakage rate. Specifically, when implemented, the method may include the following content.

[0058] S101: Obtain the depth-domain seismic data volume and the engineering parameters of the drilled wells for the target well.

[0059] In some embodiments, before obtaining the depth-domain seismic data volume of the target well, it further includes:

[0060] Obtain the seismic data volume in the seismic work area of the target well, the drilling data, logging data, and mud logging data of the drilled wells; wherein, the drilling data of the drilled wells includes wellhead coordinates and wellbore trajectory data;

[0061] Convert the seismic data volume according to the drilling data, logging data, and mud logging data to obtain the depth-domain seismic data volume.

[0062] In some embodiments, the above seismic work area may be a three-dimensional work area established in the area of the well to be drilled; the above target well may be a well located in the area of the well to be drilled; the above seismic data volume may be three-dimensional data, which refers to the time-domain seismic data volume formed by using artificial seismic methods to record seismic reflection waves during the oil and gas field exploration process.

[0063] In some embodiments, the drilling data of the drilled well may at least include: wellhead coordinates, wellbore trajectory data, where the wellbore trajectory data may at least include: measured depth, well inclination angle, and azimuth angle.

[0064] In some embodiments, to convert the seismic data volume into a depth-domain seismic data volume according to the drilling data, logging data, and mud logging data as described above, the following steps may be taken: Based on the measured depth data in the drilling data of the drilled well, combined with the layer velocity obtained from the logging data and mud logging data, determine the time-depth relationship of a single well. Then, based on the time-depth relationship of the single well, establish a velocity scale for time-depth conversion calibration. Finally, based on the velocity scale for time-depth conversion calibration, convert the time-domain seismic data volume into a depth-domain seismic data volume.

[0065] In some embodiments, the above-mentioned drilled well engineering parameters may include at least one of the following: drilling fluid density, bottom-hole circulating pressure loss, formation pore pressure, hole size for each drilling section of the drilled well, drilled well depth, depth of lost circulation in the drilled well, and loss rate.

[0066] In some embodiments, by converting the seismic data volume from the time domain to the depth domain and obtaining the drilled well engineering parameters, it can lay a foundation for accurately and quickly extracting the variance attribute of the depth of the drilled well trajectory in the subsequent process. At the same time, by obtaining the engineering parameters, it can lay a foundation for improving the practicality of the prediction model in the subsequent process, and can effectively meet the engineering needs.

[0067] S102: Extract the variance attribute of the depth of the drilled well trajectory according to the depth-domain seismic data volume and the drilled well engineering parameters.

[0068] In some embodiments, when extracting the variance attribute of the depth of the drilled well trajectory according to the depth-domain seismic data volume and the drilled well engineering parameters as described above, the specific implementation may include:

[0069] S1: Extract the variance attribute data volume of the depth-domain seismic data volume;

[0070] S2: Import the wellhead coordinates and the wellbore trajectory data into the depth-domain seismic data volume to form target trajectory data;

[0071] S3: Extract the variance attribute of the variance attribute data volume along the target trajectory data to form a two-dimensional array of depth and variance;

[0072] S4: Use the two-dimensional array of depth and variance as the variance attribute of the depth of the drilled well trajectory.

[0073] In some embodiments, the variance attribute data volume of the above-mentioned deep-domain seismic data volume can be extracted in the following manner: Relying on professional software, according to the pre-set selection range and variance calculation formula of the variance attribute data volume, the seismic data volume is calculated along the depth, and finally the variance attribute data volume is formed, thus completing the extraction of the variance attribute data volume. Among them, the pre-set selection range of the variance attribute data volume can be: 3 traces are taken in each of the X and Y directions, and 12 to 15 sampling points are taken vertically. It should be noted that the calculation range method of the variance attribute data volume is not limited to the above example. Those skilled in the art may make other changes under the inspiration of the technical essence of the embodiments of this specification. However, as long as the functions and effects achieved are the same or similar to those of the embodiments of this specification, they should all be covered within the protection scope of the embodiments of this specification.

[0074] In some embodiments, since the above-mentioned variance attribute data volume is three-dimensional data, it is further necessary to convert the above-mentioned variance attribute data volume into two-dimensional data, that is, the variance attribute data volume can be converted into the variance attribute of the depth of the drilled well trajectory. The variance attribute of the depth of the drilled well trajectory can be extracted in the following manner: Import the measured depth, well inclination angle, azimuth angle, and wellhead coordinates in the above-mentioned wellbore trajectory data into the software carrying the seismic data volume to form target trajectory data, extract the variance attribute of the variance attribute data volume along the target trajectory data, and form a two-dimensional array of depth and variance. Among them, the two-dimensional array of depth and variance can be represented by the following matrix: Among them, the first column is the depth data, and the second column is the corresponding variance data. The formed two-dimensional array of depth and variance is used as the variance attribute of the depth of the drilled well trajectory to be extracted. By extracting the variance attribute of the depth of the drilled well trajectory, it can lay a data foundation for further determining the variance value of the target depth subsequently.

[0075]

[0076] S103: Extract the variance value of the target depth of the drilled well according to the variance attribute.

[0077] S104: Form a planar graph of the variance value of the target depth according to the variance value of the target depth.

[0078] In some embodiments, the variance value of the target depth of the drilled well can be extracted according to the following formula:

[0079] V≥(V max -V min )×0.5

[0080] Among them, V is the variance value of the depth to be extracted, V max is the maximum variance along the well trajectory, and V min is the minimum variance along the well trajectory.

[0081] In some embodiments, after extracting the variance value of the target depth of the drilled well, in specific implementation, it may further include: determining the corresponding target depth according to the variance value of the target depth of the drilled well, then determining the depth-domain seismic data volume corresponding to the target depth and slicing it, and extracting the planar imaging graph of the variance value of the target depth. When slicing, the depth needs to be corrected according to the following formula:

[0082] TVD 切片 = TVD V - RKB

[0083] Wherein, TVD 切片 represents the elevation true vertical depth of the slice of the variance attribute data volume, TVD V represents the track true vertical depth corresponding to the target depth, and RKB represents the difference between the track depth reference point of the drilled well and the depth reference point of the seismic data volume.

[0084] In some embodiments, by performing depth correction, the planar imaging graph of the variance value of the target depth can be obtained more accurately, thereby laying a foundation for accurately determining the connectivity index subsequently.

[0085] S105: Determine the connectivity index of the drilled well according to the planar graph of the variance value of the target depth.

[0086] In some embodiments, before determining the connectivity index of the drilled well according to the planar graph of the variance value of the target depth, in specific implementation, it may further include:

[0087] S1: Obtain the wellbore and multiple fractures in the drilled well;

[0088] S2: Determine the degree of connection between the wellbore and each of the multiple fractures, the degree of connection between each of the multiple fractures, and the development degree of each of the multiple fractures;

[0089] S3: Obtain the relative positions of the wellbore and each of the multiple fractures;

[0090] S4: Form an imaging set of the wellbore and the relative positions according to the wellbore and the relative positions;

[0091] S5: Determine the connectivity determination criterion according to the imaging set of the relative positions, the degree of connection between the wellbore and each of the multiple fractures, the degree of connection between each of the multiple fractures, and the development degree of each of the multiple fractures, wherein the connectivity determination criterion is used to obtain the connectivity index of the drilled well in combination with the planar graph.

[0092] In some embodiments, the relative position imaging sets of the wellbore and each of the multiple fractures, the degree of connection between the wellbore and each of the multiple fractures, the degree of connection between each of the multiple fractures, and the development degree of each of the multiple fractures can be integrated to form various types of connectivity index charts. Each type of connectivity index chart is sorted and numbered, which is used as the determination criterion for connectivity. Then, the planar graphic slice of the variance value at the target depth obtained is compared and analyzed with different types of connectivity index charts to determine the connectivity index of the target depth of the drilled well. The connectivity index is used to characterize the channels and storage spaces for drilling fluid loss.

[0093] In some embodiments, by obtaining the connectivity index to characterize the channels and storage spaces, the quantitative prediction of the fractured loss rate can be realized, overcoming the problems of difficult parameter acquisition, many assumed variables, and large prediction errors in the analytical method.

[0094] S106: Train the full-feedback neural network model according to the drilled well engineering parameters, the variance value of the target depth of the drilled well, and the connectivity index of the drilled well to obtain a loss rate prediction model.

[0095] In some embodiments, the above-mentioned training of the full-feedback neural network model according to the drilled well engineering parameters, the variance value of the target depth of the drilled well, and the connectivity index of the drilled well to obtain a loss rate prediction model may specifically include:

[0096] S1: Calculate the pressure difference according to the drilling fluid density, the bottom-hole circulating pressure loss, and the formation pore pressure according to the following formula:

[0097] ΔP = P s + P anu - P pore

[0098] where ΔP is the difference between the hydrostatic pressure in the wellbore at a certain depth of the wellbore and the formation pore pressure, P s is the static hydrostatic pressure of the drilling fluid with a specific density in the non-circulating state, P anu is the bottom-hole circulating pressure loss, and P pore is the formation pore pressure;

[0099] S2: Train the full-feedback neural network model according to the pressure difference, the variance value of the target depth of the drilled well, and the connectivity index of the drilled well to obtain a loss rate prediction model.

[0100] In some embodiments, the above-mentioned drilling fluid density (ρ mud ) can be used to determine the static hydrostatic pressure, and the above-mentioned bottom-hole circulating pressure loss (P anu) can represent the pressure value added to the bottom hole due to the annular circulation friction under certain drilling rates, surface mud pump displacements, and wellbore size conditions.

[0101] In some embodiments, the fully feedback neural network model can be trained according to the following formula:

[0102] LR i = f(ΔP i , V i , N Index )

[0103] Where LR i refers to the leakage rate at depth i, ΔP i represents the difference between the liquid column pressure in the wellbore and the formation pore pressure at depth i, V i represents the variance value at depth i, N Index represents the connectivity index at depth i.

[0104] In some embodiments, the leakage rate at the above-mentioned depth i can be the leakage rate at the actual leakage depth of the drilled well. Among them, the selection principles for the leakage depth and leakage rate are as follows: When there are multiple leakages in the same well section, select the data at the first discovered leakage as the valid data; the data with lost circulation plugging materials added to the drilling fluid in the drilling process shall not be selected. According to the above principles, the leakage rate at the actual leakage depth of the drilled well is screened, that is, LR i is used as the target data set in the prediction model, and the target data set is used to verify the training result.

[0105] In some embodiments, the number of hidden layers of the above-mentioned feedforward neural network can be selected from 3 to 5, the output layer is 1 layer, and the training method can be selected from the medium network training function (Levenberg-Marquardt training method), the gradient descent backpropagation algorithm training function, or the gradient descent backpropagation algorithm training function with adaptive adjustment of the learning rate and additional momentum factor. Among them, the training error target is less than 0.01, and the training learning rate is 0.05. It should be noted that the training method of the model is not limited to the above examples. Those skilled in the art may make other changes under the inspiration of the technical essence of the embodiments of this specification, but as long as the functions and effects achieved are the same or similar to those of the embodiments of this specification, they should all be covered within the protection scope of the embodiments of this specification.

[0106] In some embodiments, by screening the loss depth and loss rate, the prediction accuracy of the model can be effectively improved. At the same time, the geological factors and engineering factors causing losses are organically combined, and a fracture leakage prediction method that conforms to the downhole objective conditions is established starting from the three elements of leakage occurrence (leakage dynamic element (i.e., pressure difference), channel element, and reservoir space element (i.e., connectivity index)), solving the problem that traditional mechanical analysis means cannot accurately predict the leakage risk of fractured formations, achieving a higher-precision quantitative prediction of the fracture leakage scale, and providing an important guiding basis for drilling engineering design and construction.

[0107] S107: Obtain the dataset of the well to be drilled for the target well, and input the dataset of the well to be drilled into the loss rate prediction model to obtain the loss rate of the well to be drilled in the target well.

[0108] In some embodiments, the above dataset of the well to be drilled may include the pressure difference at a certain depth, the variance at a certain depth, and the connectivity index at a certain depth. Input the dataset of the well to be drilled into the loss rate prediction model to obtain the loss rate of the well to be drilled in the target well. The acquisition method of the above dataset of the well to be drilled may refer to the acquisition of the engineering parameters, variance values, and connectivity indices of the drilled wells, which will not be elaborated in this specification.

[0109] In some embodiments, after inputting the dataset of the well to be drilled into the loss rate prediction model to obtain the loss rate of the well to be drilled in the target well, in specific implementation, it may further include:

[0110] Judge whether the loss rate is greater than a preset threshold;

[0111] When the loss rate is greater than the preset threshold, send a risk prompt to the target drilling equipment.

[0112] In some embodiments, after inputting the dataset of the well to be drilled into the fracture loss rate prediction model to obtain the loss rate of the well to be drilled in the target well, in specific implementation, it may further include:

[0113] When the loss rate is less than the preset threshold, guide the drilling development work based on the loss rate.

[0114] In some embodiments, the above target drilling equipment may be the equipment that is currently drilling, and may include a complete set of surface equipment for drilling, special drilling tools, and drilling instruments. It should be noted that the target drilling equipment is not limited to the above examples. Those skilled in the art may make other changes under the inspiration of the technical essence of the embodiments of this specification, but as long as the functions and effects achieved are the same or similar to those of the embodiments of this specification, they should be covered within the protection scope of the embodiments of this specification.

[0115] In some embodiments, by quantifying the leakage risk, drilling operations can be carried out safely and efficiently.

[0116] The above method will be described below with reference to a specific embodiment. However, it should be noted that this specific embodiment is only for better illustrating the present application and does not constitute an improper limitation of the present application.

[0117] Before specific implementation, first, obtain the drilled well engineering parameters, the seismic data volume of the target well in the seismic work area, the drilling data, logging data, and mud logging data of the drilled well. Among them, the drilling data of the drilled well includes wellhead coordinates and wellbore trajectory data. Secondly, convert the seismic data volume according to the drilling data, logging data, and mud logging data to obtain the depth-domain seismic data volume of the target well. Further, extract the variance attribute data volume of the depth-domain seismic data volume, import the wellhead coordinates and the wellbore trajectory data into the depth-domain seismic data volume to form target trajectory data, extract the variance attribute along the target trajectory data of the variance attribute data volume to form a two-dimensional array of depth and variance, and use the two-dimensional array of depth and variance as the variance attribute of the depth of the drilled well trajectory. Further, according to the variance attribute of the depth of the drilled well trajectory, extract the variance value of the target depth of the drilled well, and then, according to the variance value of the target depth, form a planar graph of the variance value of the target depth, and then compare and analyze the planar graph of the variance value of the target depth with the pre-determined connectivity determination criterion to determine the connectivity index of the drilled well. Further, collect and screen the leakage rate at the actual downhole leakage depth of the drilled well, use the difference between the liquid column pressure in the wellbore and the formation pore pressure at the target depth determined according to the drilled well engineering parameters, the variance value of the target depth of the drilled well, and the connectivity index of the drilled well as the training set, and use the leakage rate at the actual downhole leakage depth of the drilled well as the verification set to train the fully feedback neural network model, and finally obtain the leakage rate prediction model. During specific implementation, obtain the dataset of the well to be drilled of the target well, input the dataset of the well to be drilled into the leakage rate prediction model to obtain the leakage rate of the well to be drilled in the target well, and then, according to the leakage rate, safely and efficiently guide the drilling and development work. Through the above method, the rate of downhole leakage can be determined more accurately, the leakage risk can be quantified, which can provide an important basis for the optimization design of drilling, and has an important role in reducing downhole complex situations, reducing the construction period and economic losses.

[0118] In a specific scenario example, the method for determining the downhole leakage rate provided by the embodiments of the present specification can be applied to improve the accuracy of predicting the downhole leakage rate. During specific implementation, refer to Figure 2 as shown, including:

[0119] Preparation of the seismic work area corresponding to the well to be drilled: Taking the scenario where there are 10 drilled wells in a certain oil and gas area as an example, different degrees of downhole leakage occurred during the actual drilling of these 10 wells, and there are existing 3D time-domain seismic data and engineering parameter records of the 10 drilled wells in the area.

[0120] 3D depth-domain seismic data after time-depth conversion and depth correction: According to the logging, well logging, and geological stratification data of the 10 drilled wells, perform time-depth relationship conversion on the 3D time-domain seismic data to obtain the 3D depth-domain seismic data.

[0121] Extraction of variance attributes from the 3D depth-domain seismic data: Relying on professional software, according to the pre-set selection range of the variance attribute data volume and the variance calculation formula, perform depth-wise calculation on the seismic data volume, and finally form the variance attribute data volume, thereby completing the extraction of the variance attribute data volume. Among them, the pre-set selection range of the variance attribute data volume can be: 3 traces are taken in each of the X and Y directions, and 12 - 15 sampling points are taken vertically.

[0122] Variance attributes along the depth of the drilled well trajectory: After extracting the variance attribute data volume, import the wellhead coordinates and wellbore trajectory data (measured depth, well inclination angle, and azimuth angle) of the 10 drilled wells into the 3D depth-domain seismic data volume to form the target trajectory data. Based on the variance attribute data volume along the target trajectory data, extract the variance attributes along the trajectory to form a depth-variance two-dimensional array: Among them, the first column is different depth data, and the second column is the variance data corresponding to different depth data.

[0123]

[0124] Lateral slicing of the abnormal points of the variance attributes along the trajectory of the drilled wells in the oil and gas area to form a legend of the relative position relationship and fracture communication relationship between the wellbore (shaft) and fractures at different depths. Comprehensively analyze the position relationship between the wellbore (shaft) and fractures, the degree of connection between fractures, the development degree of fractures, and the connection degree between the shaft and fractures, and perform sorting. Assign connectivity indices to the area according to the sorting. For example, Figure 3 As shown, 9 typical cases of shaft-fracture-development-crosslinking are formed, that is, the connectivity determination criteria, corresponding to Figure 3 the 9 small figures in, sorted in order of crosslinking degree from 0 to 8, and the corresponding connectivity indices are 0 to 8.

[0125] Extraction of the fracture connectivity attributes of the formation to be drilled and determination of the connectivity index: Compare and analyze the obtained lateral slicing imaging graph with the above connectivity determination criteria to determine the connectivity of the formation drilled by the shaft at the corresponding depth of the drilled well.

[0126] Drilled well engineering parameters: The main parameters are as follows: the density of the drilling fluid used along the well depth of the drilled well, the bottom-hole circulating pressure loss during drilling, the formation pore pressure data along the well depth, the hole sizes and depths of each section of the drilled well, and the depth and loss rate of actual losses occurred in the drilled well. Among them, data such as the density of the drilling fluid, the formation pore pressure, and the hole size and depth of the drilled well. When losses occur downhole, the bottom-hole pressure difference is calculated by the following formula:

[0127] ΔP = P s + P anu - P pore

[0128] Where, ΔP is the difference between the hydrostatic pressure in the wellbore at a certain depth of the wellbore and the formation pore pressure, P s is the static hydrostatic pressure of the drilling fluid with a specific density in the non-circulating state, P anu is the bottom-hole circulating pressure loss, P pore is the formation pore pressure.

[0129] Depth and loss rate of losses: The selection principles for the depth and loss rate of losses are as follows: When losses occur in multiple places in the same well section, the data at the first discovery of losses are selected as valid data; the data of adding lost circulation materials in the drilling fluid in advance during drilling are not selected. According to the above principles, the loss rate at the actual loss depth of the drilled well is screened as the target data set in the prediction model, and the target data set is used to verify the training results.

[0130] The full-feedback neural network model is trained according to the following formula:

[0131] LR i = f(ΔP i , V i , N Index )

[0132] Where, LR i is the loss rate at depth i, ΔP i is the difference between the hydrostatic pressure in the wellbore and the formation pore pressure at depth i, V i is the variance value at depth i, N Index is the connectivity index at depth i.

[0133] The training data set is as follows: The first column represents ΔP i the difference between the hydrostatic pressure in the wellbore and the formation pore pressure, the second column represents V i the variance value, and the third column represents N Index the connectivity index.

[0134]

[0135] Target data set (LRi )As follows:

[0136]

[0137] Feedforward neural network training and prediction of the scale of drilling fluid loss: Based on the training set established above, train the fully feedback neural network model, and verify the training results of the model based on the target data set. The example training results show that when the accuracy is greater than 95%, obtain the data set of the well to be drilled: Import the well trajectory data of the well to be drilled into the seismic data volume, and extract the variance attribute along the well trajectory. Use the following formula to preliminarily screen the depth of the risk points, that is, at V max is the maximum variance along the well trajectory, V min is the minimum variance along the well trajectory. After that, substitute the maximum variance along the well trajectory and the minimum variance along the well trajectory into the following formula to obtain the depth corresponding to the variance value, and use this depth as the preliminarily screened depth of the risk point.

[0138] V≥(V max -V min )×0.5

[0139] where V is the variance value of the depth to be extracted, V max is the maximum variance along the well trajectory, and V min is the minimum variance along the well trajectory.

[0140] Based on the engineering parameters and connectivity index to be adopted according to the preliminarily screened point depth, establish the analysis data set of the well to be drilled as follows:

[0141]

[0142] Use the completed prediction model to input the formed analysis data set of the well to be drilled into the prediction model to predict the leakage rate, and the comparison between the obtained results and the actual situation is as follows:

[0143]

[0144] From the analysis of the above results, it can be seen that for the first two depth points, the predicted leakage rates are 0.001 and 0.002 respectively. Such leakage rates at this scale cannot be recognized in engineering practice, and the corresponding engineering practice shows no leakage, that is, the leakage rate is considered to be 0, with a coincidence degree of 100% with the actual value. The 3rd and 4th predicted results are 125 and 85 respectively, and these two are the specific leakage rates, whose leakage magnitudes are in full agreement with the corresponding engineering practice. It is very difficult to accurately measure the specific leakage rate in actual engineering, especially for working conditions with a leakage rate greater than 80 cubic meters per hour. Therefore, in engineering practice, micro-leakage (less than 10 cubic meters per hour), small leakage (10 - 30 cubic meters per hour), medium leakage (30 - 60 cubic meters per hour), large leakage (greater than 60 cubic meters per hour), and lost circulation leakage are often used to evaluate the leakage rate (there will be slight differences in the classification methods in different oil and gas regions). Therefore, if the predicted result falls within the corresponding classification interval, it can represent the leakage scale and the impacts caused. In addition, the accuracy of leakage prediction is considered from two aspects. One is whether there is leakage at the predicted location, and the other is whether the leakage scale falls within the corresponding partition. Evaluated from these two perspectives, the prediction accuracy of this example is 100%. Currently, in the industry, a fracture leakage prediction accuracy reaching 80% is considered to be of high accuracy. Therefore, the prediction accuracy in this application is relatively high. Using the method described in this specification to analyze and predict the leakage scale of the well to be drilled under the designed trajectory and designed engineering parameters has a relatively high accuracy, can provide an important basis for well drilling optimization design, and plays an important role in reducing downhole complex situations, shortening the construction period, and reducing economic losses.

[0145] Although this specification provides the method operation steps or device structures as described in the following embodiments or as shown in the attached Figure 4 drawings, based on routine or non-creative labor, more or fewer operation steps or module units may be included in the described method or device. In steps or structures where there is no necessary causal relationship logically, the execution order of these steps or the module structure of the device is not limited to the execution order or module structure shown in the embodiments of this specification or the drawings. When the described method or module structure is applied to actual devices, servers, or terminal products, it can be executed sequentially or in parallel according to the method or module structure shown in the embodiments or the drawings (for example, in an environment of parallel processors or multi-threaded processing, and even including an implementation environment of distributed processing and server clusters).

[0146] Based on the above method for determining the downhole leakage rate, this specification also presents an embodiment of a device for determining the downhole leakage rate. As Figure 4 shown, the device for determining the downhole leakage rate may specifically include the following modules:

[0147] An acquisition module 401, configured to obtain the depth domain seismic data volume of the target well and the engineering parameters of the drilled well;

[0148] The variance extraction module 402 is configured to extract the variance attribute of the depth of the drilled well trajectory according to the depth-domain seismic data volume and the drilled well engineering parameters; and extract the variance value of the target depth of the drilled well according to the variance attribute.

[0149] The connectivity index extraction module 403 is configured to form a planar graph of the variance value of the target depth according to the variance value of the target depth; and determine the connectivity index of the drilled well according to the planar graph of the variance value of the target depth.

[0150] The model training module 404 is configured to train a full-feedback neural network model according to the drilled well engineering parameters, the variance value of the target depth of the drilled well, and the connectivity index of the drilled well, so as to obtain a lost circulation rate prediction model.

[0151] The output module 405 is configured to obtain the dataset of the well to be drilled of the target well, input the dataset of the well to be drilled into the lost circulation rate prediction model, and obtain the lost circulation rate of the well to be drilled in the target well.

[0152] In some embodiments, before the above-mentioned acquisition module 401 is specifically implemented, it can be used to: acquire the seismic data volume of the target well in the seismic work area, the drilling data, logging data, and mud logging data of the drilled well; wherein, the drilling data of the drilled well includes the wellhead coordinates and the wellbore trajectory data; and convert the seismic data volume according to the drilling data, logging data, and mud logging data to obtain the depth-domain seismic data volume of the target well.

[0153] In some embodiments, when the above-mentioned variance extraction module 402 is specifically implemented, it can be used to: extract the variance attribute data volume of the depth-domain seismic data volume; import the wellhead coordinates and the wellbore trajectory data into the depth-domain seismic data volume to form target trajectory data; extract the variance attribute of the variance attribute data volume along the target trajectory data to form a two-dimensional array of depth and variance; and use the two-dimensional array of depth and variance as the variance attribute of the depth of the drilled well trajectory. The variance value of the target depth of the drilled well is extracted according to the following formula:

[0154] V≥(V max -V min )×0.5

[0155] wherein, V is the variance value of the depth to be extracted, V max is the maximum variance along the well trajectory, and V min is the minimum variance along the well trajectory.

[0156] In some embodiments, before the specific implementation of the above-mentioned connectivity index extraction module 403, it can be used to: obtain the wellbore and multiple fractures in the drilled well; determine the degree of connection between the wellbore and each fracture among the multiple fractures, the degree of connection between each fracture among the multiple fractures, and the development degree of each fracture among the multiple fractures; obtain the relative positions of the wellbore and each fracture among the multiple fractures; form an imaging set of the wellbore and the relative positions according to the wellbore and the relative positions; determine a connectivity determination criterion according to the relative position imaging set, the degree of connection between the wellbore and each fracture among the multiple fractures, the degree of connection between each fracture among the multiple fractures, and the development degree of each fracture among the multiple fractures, wherein the connectivity determination criterion is used to combine the planar graph to obtain the connectivity index of the drilled well.

[0157] In some embodiments, when the above-mentioned model training module 404 is specifically implemented, it can be used to: The engineering parameters of the drilled well include at least one of the following: drilling fluid density, bottom hole circulating pressure loss, formation pore pressure, hole size of each section of the drilled well, drilled well depth, depth of lost circulation in the drilled well, lost circulation rate. Calculate the pressure difference according to the drilling fluid density, the bottom hole circulating pressure loss, and the formation pore pressure according to the following formula:

[0158] ΔP = P s + P anu - P pore

[0159] where, ΔP is the difference between the hydrostatic pressure of the liquid column in the wellbore at a certain depth of the wellbore and the formation pore pressure, P s is the static hydrostatic pressure of the drilling fluid with a specific density in the non-circulating state, P anu is the bottom hole circulating pressure loss, P pore is the formation pore pressure;

[0160] Train the full-feedback neural network model according to the pressure difference, the variance value of the target depth of the drilled well, and the connectivity index of the drilled well to obtain a lost circulation rate prediction model.

[0161] Train the full-feedback neural network model according to the following formula:

[0162] LR i = f(ΔP i , V i , N Index )

[0163] where, LR i is the lost circulation rate at depth i, ΔP i is the difference between the hydrostatic pressure of the liquid column in the wellbore and the formation pore pressure at depth i, V i is the variance value at depth i, N Indexis the connectivity index at depth i.

[0164] It should be noted that the units, devices, or modules described in the above embodiments can be specifically implemented by computer chips or entities, or by products with certain functions. For the convenience of description, the above devices are described by dividing them into various modules according to their functions. Of course, when implementing this specification, the functions of each module can be implemented in the same or multiple software and / or hardware, or the modules that implement the same function can be implemented by a combination of multiple sub-modules or sub-units. The device embodiments described above are only illustrative. For example, the division of the units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the coupling or direct coupling or communication connection shown or discussed with each other can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be in an electrical, mechanical or other form.

[0165] As can be seen from the above, based on the device for determining the downhole leakage rate provided by the embodiments of this specification, on the one hand, all the data required in this specification come from the commonly used data in the oilfield, without the need to use any supporting tools, and the parameters involved are all objectively obtained data, reducing the errors caused by human subjective judgment. Therefore, it has strong technical and economic advantages and is easy to promote and apply. On the other hand, the geological factors and engineering factors that cause leakage are organically combined. Starting from the three elements of leakage occurrence (determining the pressure difference based on engineering parameters, taking the pressure difference as the dynamic element of leakage, the connectivity index representing the channel element, and the reservoir space element), a fracture leakage prediction method that conforms to the downhole objective conditions is established, solving the problem that traditional mechanical analysis methods cannot accurately predict the leakage risk of fractured formations, achieving a relatively high-precision quantitative prediction of the fracture leakage scale, providing an important guiding basis for drilling engineering design and construction, and the obtained leakage rate prediction model can meet the engineering needs.

[0166] An embodiment of this specification also provides an electronic device, including a processor and a memory for storing executable instructions of the processor. When specifically implemented, the processor may execute the following steps according to the instructions: obtaining depth-domain seismic data volume of a target well and engineering parameters of drilled wells; extracting the variance attribute of the depth of the drilled well trajectory according to the depth-domain seismic data volume and the engineering parameters of the drilled wells; extracting the variance value of the target depth of the drilled well according to the variance attribute; forming a planar graph of the variance value of the target depth according to the variance value of the target depth; determining the connectivity index of the drilled well according to the planar graph of the variance value of the target depth; training a full-feedback neural network model according to the engineering parameters of the drilled wells, the variance value of the target depth of the drilled well, and the connectivity index of the drilled well to obtain a leakage rate prediction model; obtaining a dataset of wells to be drilled in the target well, and inputting the dataset of wells to be drilled into the leakage rate prediction model to obtain the leakage rate of the wells to be drilled in the target well.

[0167] To be able to more accurately complete the above instructions, refer to Figure 5 As shown, an embodiment of this specification also provides another specific electronic device. Among them, the electronic device includes a network communication port 501, a processor 502, and a memory 503. The above structures are connected by internal cables so that each structure can perform specific data interactions.

[0168] Among them, the network communication port 501 can specifically be used to obtain the depth-domain seismic data volume of the target well and the engineering parameters of the drilled wells.

[0169] The processor 502 can specifically be used to extract the variance attribute of the depth of the drilled well trajectory according to the depth-domain seismic data volume and the engineering parameters of the drilled wells; extract the variance value of the target depth of the drilled well according to the variance attribute; form a planar graph of the variance value of the target depth according to the variance value of the target depth; determine the connectivity index of the drilled well according to the planar graph of the variance value of the target depth; train a full-feedback neural network model according to the engineering parameters of the drilled wells, the variance value of the target depth of the drilled well, and the connectivity index of the drilled well to obtain a leakage rate prediction model; obtain a dataset of wells to be drilled in the target well, and input the dataset of wells to be drilled into the leakage rate prediction model to obtain the leakage rate of the wells to be drilled in the target well.

[0170] The memory 503 can specifically be used to store the corresponding instruction programs.

[0171] In this embodiment, the network communication port 501 can be bound to different communication protocols, so as to send or receive different data. For example, the network communication port can be a port responsible for web data communication, or a port responsible for FTP data communication, or a port responsible for mail data communication. In addition, the network communication port can also be a physical communication interface or communication chip. For example, it can be a wireless mobile network communication chip, such as GSM, CDMA, etc.; it can also be a Wifi chip; it can also be a Bluetooth chip.

[0172] In this embodiment, the processor 502 can be implemented in any suitable manner. For example, the processor can take the form of, for example, a microprocessor or a processor and a computer-readable medium storing computer-readable program code (such as software or firmware) executable by the (micro)processor, logic gates, switches, application specific integrated circuit (ASIC), programmable logic controller, and embedded microcontroller, etc. This specification does not make a limitation.

[0173] In this embodiment, the memory 503 can include multiple levels. In a digital system, anything that can store binary data can be a memory; in an integrated circuit, a circuit without a physical form but with a storage function is also called a memory, such as RAM, FIFO, etc.; in a system, a storage device with a physical form is also called a memory, such as a memory stick, TF card, etc.

[0174] The embodiment of this specification also provides a computer storage medium based on the above method for determining the downhole leakage rate. The computer storage medium stores computer program instructions, which when executed, implement: obtaining the depth-domain seismic data volume and the drilled well engineering parameters of the target well; extracting the variance attribute of the depth of the drilled well trajectory according to the depth-domain seismic data volume and the drilled well engineering parameters; extracting the variance value of the target depth of the drilled well according to the variance attribute; forming a planar graph of the variance value of the target depth according to the variance value of the target depth; determining the connectivity index of the drilled well according to the planar graph of the variance value of the target depth; training a full-feedback neural network model according to the drilled well engineering parameters, the variance value of the target depth of the drilled well, and the connectivity index of the drilled well to obtain a leakage rate prediction model; obtaining the data set of the well to be drilled in the target well, and inputting the data set of the well to be drilled into the leakage rate prediction model to obtain the leakage rate of the well to be drilled in the target well.

[0175] In this embodiment, the above storage medium includes, but is not limited to, random access memory (RAM), read-only memory (ROM), cache, hard disk drive (HDD), or memory card. The memory can be used to store computer program instructions. The network communication unit can be set according to the standards stipulated by the communication protocol and is used as an interface for network connection communication.

[0176] Although this specification provides method operation steps as described in the embodiments or flowcharts, more or fewer operation steps may be included based on conventional or non-creative means. The step sequences listed in the embodiments are only one of the ways of the execution sequences of numerous steps and do not represent the only execution sequence. When the actual device or client product is executed, it can be executed in the method sequence shown in the embodiments or the drawings or executed in parallel (for example, in an environment of parallel processors or multi-threaded processing, or even in a distributed data processing environment). The term "comprise", "include" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, product or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, product or device. Without further limitation, there is no exclusion of additional identical or equivalent elements in the process, method, product or device comprising the said elements. The terms such as first, second are used to denote names and do not denote any particular sequence.

[0177] Those skilled in the art also know that, in addition to implementing the controller in the form of pure computer-readable program code, the method steps can be logically programmed to enable the controller to be implemented in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, embedded microcontrollers, etc. to achieve the same functions. Therefore, such a controller can be regarded as a hardware component, and the devices included therein for implementing various functions can also be regarded as the structures within the hardware component. Or even, the devices for implementing various functions can be regarded as either software modules for implementing the method or the structures within the hardware component.

[0178] This specification can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, classes, etc. that perform specific tasks or implement specific abstract data types. This specification can also be practiced in a distributed computing environment where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0179] From the description of the above embodiments, those skilled in the art can clearly understand that this specification can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of this specification can essentially be embodied in the form of a software product, which can be stored in a storage medium such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to enable a computer device (which can be a personal computer, mobile terminal, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of this specification.

[0180] The various embodiments in this specification are described in a progressive manner. For the same or similar parts between the various embodiments, reference can be made to each other, and the key points of each embodiment are the differences from other embodiments. This specification can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multi-processor systems, microprocessor-based systems, set-top boxes, programmable electronic devices, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, and so on.

[0181] Although this specification is depicted through embodiments, those of ordinary skill in the art know that this specification has many variations without departing from the spirit of this specification, and it is hoped that the appended claims will include these variations without departing from the spirit of this specification.

Claims

1. A method for determining the downhole leakage rate, characterized in that Including: Obtaining the depth-domain seismic data volume of the target well and the drilled well engineering parameters; wherein, the method for obtaining the depth-domain seismic data volume of the target well is as follows: obtaining the seismic data volume of the target well in the seismic work area, the drilling data, logging data, and mud logging data of the drilled well, and the drilling data of the drilled well including the wellhead coordinates and wellbore trajectory data; converting the seismic data volume according to the drilling data, logging data, and mud logging data to obtain the depth-domain seismic data volume of the target well; Extracting the variance attribute of the drilled well trajectory depth according to the depth-domain seismic data volume and the drilled well engineering parameters; wherein, extracting the variance attribute of the drilled well trajectory depth according to the depth-domain seismic data volume and the drilled well engineering parameters includes: extracting the variance attribute data volume of the depth-domain seismic data volume; importing the wellhead coordinates and the wellbore trajectory data into the depth-domain seismic data volume to form target trajectory data; extracting the variance attribute of the variance attribute data volume along the target trajectory data to form a two-dimensional array of depth and variance; using the two-dimensional array of depth and variance as the variance attribute of the drilled well trajectory depth; Extracting the variance value of the drilled well target depth according to the variance attribute; Forming a planar graph of the variance value of the target depth according to the variance value of the target depth; Comparing and analyzing the planar graph of the variance value of the target depth with the connectivity determination criterion to determine the connectivity index of the drilled well; wherein, the method for determining the connectivity determination criterion is as follows: obtaining the relative positions of the wellbore and each fracture among multiple fractures; forming an imaging set of the wellbore and the relative positions according to the wellbore and the relative positions; determining the connectivity determination criterion according to the imaging set of the relative positions, the degree of intersection between the wellbore and each fracture among the multiple fractures, the degree of intersection between each fracture among the multiple fractures, and the development degree of each fracture among the multiple fractures; Training a full-feedback neural network model according to the drilled well engineering parameters, the variance value of the drilled well target depth, and the connectivity index of the drilled well to obtain a leakage rate prediction model; Obtaining the dataset of the well to be drilled in the target well, and inputting the dataset of the well to be drilled into the leakage rate prediction model to obtain the leakage rate of the well to be drilled in the target well.

2. The method according to claim 1, characterized in that, Extracting the variance value of the drilled well target depth according to the following formula: V≥(V max -V min )×0.5 Among them, V is the variance value of the depth to be extracted, and V max is the maximum variance along the well trajectory, and V min is the minimum variance along the well trajectory.

3. The method according to claim 1, wherein The drilled well engineering parameters include at least one of the following: drilling fluid density, bottom-hole circulating pressure loss, formation pore pressure, hole size of each drilled well section, drilled well depth, depth of drilled well leakage, leakage rate.

4. The method according to claim 3, wherein Training a full-feedback neural network model according to the drilled well engineering parameters, the variance value of the drilled well target depth, and the connectivity index of the drilled well to obtain a leakage rate prediction model, including: Calculating the pressure difference according to the drilling fluid density, the bottom-hole circulating pressure loss, and the formation pore pressure according to the following formula: ΔP = P s + P anu - P pore where ΔP is the difference between the hydrostatic pressure of the liquid column in the wellbore at a certain depth and the formation pore pressure, P s is the static hydrostatic pressure of a drilling fluid with a specific density in the non-circulating state, P anu is the circulating pressure loss at the bottom of the well, P pore is the formation pore pressure; Training a full-feedback neural network model according to the pressure difference, the variance value of the drilled well target depth, and the connectivity index of the drilled well to obtain a leakage rate prediction model.

5. The method according to claim 4, wherein The full-feedback neural network model is trained according to the following formula: LR i = f(ΔP i , V i , N Index ) Among them, LR i is the leakage rate at depth i, and ΔP i is the difference between the liquid column pressure in the wellbore and the formation pore pressure at depth i. V i is the variance value at depth i, and N Index is the connectivity index at depth i.

6. A device for determining the downhole leakage rate, characterized in that Including: An acquisition module, which is used to acquire the depth-domain seismic data volume of the target well and the engineering parameters of the drilled wells; wherein, the acquisition method of the depth-domain seismic data volume of the target well is as follows: acquire the seismic data volume of the target well in the seismic work area, the drilling data, logging data and mud logging data of the drilled wells, and the drilling data of the drilled wells includes wellhead coordinates and wellbore trajectory data; convert the seismic data volume according to the drilling data, logging data and mud logging data to obtain the depth-domain seismic data volume of the target well; A variance extraction module, which is used to extract the variance attribute of the depth of the drilled well trajectory according to the depth-domain seismic data volume and the engineering parameters of the drilled wells; extract the variance value of the target depth of the drilled well according to the variance attribute; wherein, extracting the variance attribute of the depth of the drilled well trajectory according to the depth-domain seismic data volume and the engineering parameters of the drilled wells includes: extracting the variance attribute data volume of the depth-domain seismic data volume; importing the wellhead coordinates and the wellbore trajectory data into the depth-domain seismic data volume to form target trajectory data; extracting the variance attribute of the variance attribute data volume along the target trajectory data to form a two-dimensional array of depth and variance; using the two-dimensional array of depth and variance as the variance attribute of the depth of the drilled well trajectory; A connectivity index extraction module, which is used to form a planar graph of the variance value of the target depth according to the variance value of the target depth; compare and analyze the planar graph of the variance value of the target depth with the connectivity determination standard to determine the connectivity index of the drilled wells; wherein, the determination method of the connectivity determination standard is as follows: acquire the relative positions of the wellbore and each of the multiple fractures; form an imaging set of the wellbore and the relative positions according to the wellbore and the relative positions; determine the connectivity determination standard according to the imaging set of the relative positions, the degree of intersection between the wellbore and each of the multiple fractures, the degree of intersection between each of the multiple fractures, and the development degree of each of the multiple fractures; A model training module, which is used to train the full-feedback neural network model according to the engineering parameters of the drilled wells, the variance value of the target depth of the drilled wells, and the connectivity index of the drilled wells to obtain a lost circulation rate prediction model; An output module, which is used to acquire the dataset of the wells to be drilled in the target well, and input the dataset of the wells to be drilled into the lost circulation rate prediction model to obtain the lost circulation rate of the wells to be drilled in the target well.

7. A computer-readable storage medium, characterized in that, It stores computer instructions, and when the instructions are executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images