Formation pressure monitoring method and device and formation pressure monitoring model construction method and device

Abstract

The invention discloses a formation pressure monitoring method and device and a formation pressure monitoring model construction method and device. The method comprises the steps that by collecting logging information and logging information of at least one adjacent well of a target block, the actual formation pressure of the adjacent well is determined according to the logging information of the adjacent well; determining logging parameters influencing the formation pressure according to the logging data of the adjacent well; optimal sensitive logging parameters are obtained from the determined logging parameters, and a deep learning network model is established in combination with the optimal sensitive logging parameters; and inputting a pre-established pressure data set into the deep learning network model for training, and using the trained deep learning network model as a formation pressure monitoring model and putting the formation pressure monitoring model into practical application. The actual formation pressure can be predicted more accurately according to the established formation pressure monitoring model, the precision and accuracy of formation pressure prediction are improved, the time of drilling work is shortened, the drilling efficiency is improved, and the practical value is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of drilling engineering, and particularly to a method and device for formation pressure monitoring and monitoring model construction. Background Art

[0002] Formation pressure includes formation pore pressure, formation fracture pressure, and formation collapse pressure. The prediction and monitoring of formation pressure play a very important role in the process of oil exploration. By predicting and monitoring the formation pressure, accidents such as blowouts and wellbore collapses can be effectively avoided, and the efficiency of oil and gas production can be improved. The accurate grasp of formation pressure not only plays an important role in the safety of drilling and shortening the drilling cycle, but also has a significant impact on improving the drilling rate, cementing, protecting the oil and gas reservoir, and reducing the drilling cost.

[0003] In recent years, after in-depth exploration of the monitoring methods of formation pressure, some calculation methods and models for formation pressure have been proposed. For example, Chinese Patent with publication number CN11499157A discloses a method for integrated monitoring of formation pressure while drilling. This method uses parameters such as drilling time, drilling speed, and drilling pressure to establish an ANN neural network model for monitoring formation pressure while drilling; Chinese Patent with publication number CN101025084A discloses a method for predicting formation pore pressure under the drill bit while drilling. This method uses well logging data such as formation density, acoustic time difference, and natural gamma to establish a prediction model for pore pressure; Chinese Patent with publication number CN115059448 discloses a method for monitoring formation pressure based on deep learning algorithms. This method uses logging parameters of adjacent wells, including well depth, mud circulation density, mechanical drilling speed, torque and other parameters to establish a GA-BP model, etc. Summary of the Invention

[0004] The monitoring of formation pressure plays an important role in the safety of drilling and shortening the drilling cycle. However, in the existing technologies for monitoring formation pressure while drilling, the parameters used for analysis are few, and the accurate pressure of the formation cannot be accurately predicted. Moreover, due to the small number of parameters and the large influence of lithology, the prediction is inaccurate, and effective data support cannot be provided for subsequent oil and gas production. Due to the inability to accurately predict the formation pressure, the drilling cycle is long and the oil and gas production efficiency is low.

[0005] In view of the above problems, the present invention is proposed to provide a method and device for monitoring formation pressure while drilling that can overcome or at least partially solve the above problems, accurately grasp the formation pressure during the drilling process, efficiently complete the drilling work, and reduce the occurrence of downhole complex accidents.

[0006] In a first aspect, an embodiment of the present invention provides a method for constructing a formation pressure monitoring model, including:

[0007] Collect the logging data and mud logging data of at least one adjacent well in the target block. According to the logging data of the adjacent well, determine the actual formation pressure of the adjacent well; according to the mud logging data of the adjacent well, determine the mud logging parameters affecting the formation pressure; the mud logging parameters include lithology-related parameters, drilling fluid-related parameters, and gas logging-related parameters.

[0008] Obtain the optimized sensitive mud logging parameters from the determined mud logging parameters affecting the formation pressure, and establish a deep learning network model based on the optimized sensitive mud logging parameters.

[0009] Train the deep learning network model with a pre-established pressure data set to obtain a formation pressure monitoring model; the pressure data set includes the actual formation pressure of the adjacent well and the optimized sensitive mud logging parameters.

[0010] In some alternative embodiments, after collecting the logging data and mud logging data of at least one adjacent well in the target block and before determining the actual formation pressure of the adjacent well, it further includes:

[0011] Perform outlier rejection and normalization processing on the logging parameters included in the logging data and the mud logging parameters included in the mud logging data.

[0012] In some alternative embodiments, the following formula is used to perform outlier rejection on the logging parameters and mud logging parameters: where n is the number of parameters, and x i is the i-th parameter data;

[0013] The following formula is used to perform normalization processing on the logging parameters and mud logging parameters: where represents the result of normalizing the i-th parameter data, x i is the i-th parameter data, MAX(X i ) is the maximum data among the i-th parameter data, and MIN(X i ) is the minimum data among the i-th parameter data.

[0014] In some alternative embodiments, determining the mud logging parameters affecting the formation pressure according to the mud logging data of the adjacent well includes:

[0015] Determine the mud logging parameters affecting the formation pressure from the mud logging data based on the causes of abnormal formation pressure and the variation law of mud logging data; where:

[0016] Lithology-related parameters include at least one of well depth, drilling time, drilling pressure, rotation speed, and torque; drilling fluid-related parameters include at least one of drilling fluid inlet flow rate, drilling fluid outlet flow rate, drilling fluid flow rate difference, drilling fluid inlet density, drilling fluid outlet density, drilling fluid density difference, drilling fluid inlet temperature, drilling fluid outlet temperature, drilling fluid temperature difference, drilling fluid inlet conductivity, drilling fluid outlet conductivity, and drilling fluid outlet conductivity difference; gas logging-related parameters include total hydrocarbon content.

[0017] In some alternative embodiments, obtaining preferred sensitive logging parameters from the determined logging parameters that affect formation pressure includes:

[0018] Determining the correlation coefficient between the logging parameters and the formation pressure based on the following formula: where n is the number of logging parameters, x i is the data of the i-th parameter, and y i is the data of the i-th formation pressure; R(X, Y) is the correlation coefficient;

[0019] According to the determined correlation coefficient, preferred sensitive logging parameters are selected from the determined logging parameters that affect formation pressure; the sensitive parameters include at least one of well depth, drilling time, drilling pressure, rotation speed, torque, drilling fluid inlet density, drilling fluid outlet density, drilling fluid inlet temperature, drilling fluid outlet temperature, drilling fluid temperature difference, and total hydrocarbon content.

[0020] In some alternative embodiments, establishing a deep learning network model based on the preferred sensitive logging parameters includes:

[0021] Establishing a deep learning network model with the preferred sensitive logging parameters as input parameters and the formation pressure as output parameters according to the preferred sensitive logging parameters;

[0022] The neural network learning model is a BP neural network model. The BP neural network model includes an input layer, a first hidden layer, a second hidden layer, and an output layer; the input layer is used to input the preferred sensitive logging parameters, the output layer is used to output the predicted formation pressure; the first hidden layer is used to extract features from the input sensitive logging parameters, and the second hidden layer is used to perform secondary feature extraction on the input sensitive logging parameters.

[0023] In some alternative embodiments, inputting a pre-established pressure data set into the deep learning network model for training, and using the trained deep learning network model as a formation pressure monitoring model includes:

[0024] Input the sensitivity logging parameters and the actual formation pressure included in the pressure dataset into the deep learning network model, output the predicted formation pressure, perform anti-normalization on the predicted formation pressure to obtain the anti-normalized predicted formation pressure, adjust the parameters of the deep learning network model based on the actual formation pressure and the predicted formation pressure output by the model, and repeat the model training process until the model meets the preset convergence condition to obtain a trained deep learning network model, which is used as the formation pressure monitoring model.

[0025] In some alternative embodiments, adjusting the parameters of the deep learning network model based on the actual formation pressure and the predicted formation pressure output by the model includes:

[0026] Determine the model error of the deep learning network model according to the actual formation pressure and the predicted formation pressure output by the model;

[0027] In the backpropagation stage, according to the model error, determine the gradient of each layer of the deep learning network layer by layer through the chain rule, and use the gradient descent method to adjust the weights and thresholds of the deep learning network model.

[0028] In a second aspect, an embodiment of the present invention provides a device for constructing a formation pressure monitoring model, including:

[0029] A data acquisition module, configured to collect logging data and mud logging data of at least one adjacent well in the target block;

[0030] A data analysis module, configured to determine the actual formation pressure of the adjacent well according to the logging data of the adjacent well; determine the logging parameters affecting the formation pressure according to the mud logging data of the adjacent well; the logging parameters include lithology-related parameters, drilling fluid-related parameters, and gas logging-related parameters;

[0031] A model establishment module, configured to obtain the optimized sensitivity logging parameters from the determined logging parameters affecting the formation pressure, establish a deep learning network model according to the optimized sensitivity logging parameters; train the deep learning network model with a pre-established pressure dataset to obtain a formation pressure monitoring model; the pressure dataset includes the actual formation pressure of the adjacent well and the optimized sensitive logging parameters.

[0032] In a third aspect, an embodiment of the present invention provides a formation pressure monitoring method, including:

[0033] Obtain the mud logging data of the target well to be monitored in the target block, and determine the sensitivity logging parameters according to the logging parameters included in the mud logging data;

[0034] Input the sensitivity logging parameters into the corresponding formation pressure monitoring model, and output the predicted formation pressure. The formation pressure monitoring model is trained by using the above-mentioned formation pressure monitoring model construction method.

[0035] Fourthly, an embodiment of the present invention provides a formation pressure monitoring device, including:

[0036] A data analysis module, configured to obtain well logging data of a target well to be monitored in a target block, and determine sensitive well logging parameters according to the well logging parameters included in the well logging data;

[0037] A pressure prediction module, configured to input the sensitive well logging parameters into a pre-established formation pressure monitoring model, and output a predicted formation pressure, where the formation pressure monitoring model is trained by using the above-mentioned formation pressure monitoring model construction method.

[0038] An embodiment of the present invention also provides a computer storage medium, in which computer-executable instructions are stored, and when the computer-executable instructions are executed by a processor, the formation pressure monitoring model construction method or the formation pressure monitoring method is implemented.

[0039] An embodiment of the present invention also provides a computer device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, where when the processor executes the program, the formation pressure monitoring model construction method or the formation pressure monitoring method is implemented.

[0040] The beneficial effects of the above technical solutions provided by the embodiments of the present invention at least include:

[0041] By collecting well logging data and well logging data of at least one adjacent well in the target block, determining the actual formation pressure of the adjacent well according to the well logging data of the adjacent well; determining well logging parameters affecting the formation pressure according to the well logging data of the adjacent well; where the well logging data includes basic well logging formation pressure data of the adjacent well; the well logging parameters include three categories: lithology parameters, drilling fluid parameters, and gas logging parameters, and a large number of parameters are selected, and the parameters are obtained not only considering the influence of lithology on the formation pressure, but also considering the influence of drilling fluid performance and hydrocarbon content on the formation pressure; obtaining preferred sensitive well logging parameters from the determined well logging parameters, and establishing a deep learning network model in combination with the preferred sensitive well logging parameters; inputting a pre-established pressure data set into the deep learning network model for training, and using the trained deep learning network model as a formation pressure monitoring model and putting it into actual application. The sensitive well logging parameters selected by using the correlation between well logging parameters and formation pressure can make the formation pressure predicted by the finally established formation pressure monitoring model closer to the actual pressure, improve the accuracy and precision of formation pressure prediction, shorten the time of drilling work, improve drilling efficiency, and have practical value.

[0042] Other features and advantages of the present invention will be set forth in the following description, and in part will be obvious from the description, or may be learned by practice of the present invention. The objectives and other advantages of the present invention may be realized and attained by the structure particularly pointed out in the written description, claims as well as the drawings.

[0043] The technical solutions of the present invention will be further described in detail below with reference to the drawings and embodiments. Description of the Drawings

[0044] The drawings are used to provide a further understanding of the present invention, and constitute a part of the description. Together with the embodiments of the present invention, they are used to explain the present invention, and do not constitute a limitation to the present invention. In the drawings:

[0045] Figure 1 It is a flowchart of the method for constructing a formation pressure monitoring model in Embodiment 1 of the present invention;

[0046] Figure 2 It is an example diagram of the training result of the formation pressure monitoring model in Embodiment 1 of the present invention;

[0047] Figure 3 It is a schematic structural diagram of the device for constructing a formation pressure monitoring model in Embodiment 1 of the present invention.

[0048] Figure 4 It is a flowchart of the formation pressure monitoring method in Embodiment 2 of the present invention;

[0049] Figure 5 It is an example diagram of the prediction result of the formation pressure monitoring model in Embodiment 2 of the present invention;

[0050] Figure 6 It is a schematic structural diagram of the formation pressure monitoring device in Embodiment 2 of the present invention. Detailed Embodiments

[0051] The prediction and monitoring of formation pressure play a very important role in the process of oil exploration. The accurate grasp of formation pressure not only has an important effect on the safety of drilling and shortening the drilling cycle, but also has a significant impact on improving the drilling rate, cementing, protecting the oil and gas reservoir, and reducing the drilling cost.

[0052] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be fully conveyed to those skilled in the art.

[0053] In the prior art, for the method of monitoring formation pressure while drilling, there are few parameters for analysis, and the accurate formation pressure cannot be accurately predicted. Moreover, due to the small number of parameters and the large influence of lithology, the prediction is inaccurate, and effective data support cannot be provided for subsequent exploitation. Since the formation pressure cannot be accurately predicted, the drilling cycle is long and the oil and gas exploitation efficiency is low.

[0054] To solve the problem of the long drilling cycle caused by inaccurate prediction of formation pressure in the prior art, the embodiments of the present invention provide a method for constructing a formation pressure monitoring model and a formation pressure monitoring method, which can solve the problem of inaccurate prediction of formation pressure in the prior art, shorten the drilling cycle, and provide technical support for oil and gas exploration work.

[0055] Embodiment 1

[0056] Embodiment 1 of the present invention provides a method for constructing a formation pressure monitoring model, and its process is as Figure 1 shown, including the following steps:

[0057] Step S101: Collect the logging data and mud logging data of at least one adjacent well in the target block. According to the logging data of the adjacent well, determine the actual formation pressure of the adjacent well; determine the mud logging parameters affecting the formation pressure according to the mud logging data of the adjacent well; the mud logging parameters include lithology-related parameters, drilling fluid-related parameters, and gas logging-related parameters.

[0058] Step S102: Obtain the preferred sensitive mud logging parameters from the determined mud logging parameters affecting the formation pressure, and establish a deep learning network model according to the preferred sensitive mud logging parameters.

[0059] Step S103: Train the established deep learning network model with a pre-established pressure data set to obtain a formation pressure monitoring model; the pressure data set includes the actual formation pressure of the adjacent well and the preferred sensitive mud logging parameters.

[0060] Preferably, in the above step S101, after collecting the logging data and mud logging data of at least one adjacent well in the target block and before determining the actual formation pressure of the adjacent well, it further includes: performing outlier rejection and normalization processing on the logging parameters included in the logging data and the mud logging parameters included in the mud logging data.

[0061] There will inevitably be outliers in the collected logging data and mud logging data. After collecting these drilling data, rejecting the outliers can eliminate the influence of the outliers on the measurement results.

[0062] Preferably, the following formula is used to reject outliers from the logging parameters and mud logging parameters: where n is the number of parameters, and x i is the data of the i-th parameter;

[0063] Normalize the logging parameters and mud logging parameters using the following formula: Wherein, represents the result of normalizing the i-th parameter data, and x i is the i-th parameter data, MAX(X i ) is the maximum data in the i-th parameter data, and MIN(X i ) is the minimum data in the i-th parameter data.

[0064] The purpose of normalization is to eliminate the dimension of the data and reduce the impact of different orders of magnitude among parameters on the subsequent model establishment.

[0065] Preferably, in the above step S101, determine the mud logging parameters affecting the formation pressure according to the mud logging data of adjacent wells, including:

[0066] Determine the mud logging parameters affecting the formation pressure from the mud logging data according to the causes of abnormal formation pressure and the variation law of mud logging data; wherein:

[0067] The lithology-related parameters include at least one of well depth, drilling time, drilling pressure, rotary speed, and torque; the drilling fluid-related parameters include at least one of the drilling fluid inlet flow rate, drilling fluid outlet flow rate, drilling fluid flow rate difference, drilling fluid inlet density, drilling fluid outlet density, drilling fluid density difference, drilling fluid inlet temperature, drilling fluid outlet temperature, drilling fluid temperature difference, drilling fluid inlet conductivity, drilling fluid outlet conductivity, and drilling fluid outlet conductivity difference; the gas logging-related parameters include total hydrocarbon content.

[0068] When selecting parameters, the embodiments of the present invention start from the perspectives of the impacts of lithology, drilling fluid, and total hydrocarbon content on the formation, select lithology-related parameters, drilling fluid-related parameters, and gas logging-related parameters, solve the drawback of relying on lithology parameters in traditional methods, and the more parameters are selected, the more accurately the formation pressure monitoring model established based on these parameters can predict the formation pressure, improving the exploration efficiency.

[0069] Preferably, in the above step S102, obtain the preferred sensitive mud logging parameters from the determined mud logging parameters affecting the formation pressure, including:

[0070] Determine the correlation coefficient between the mud logging parameters and the formation pressure based on the following formula: Wherein, n is the number of mud logging parameters, x i is the i-th parameter data, y i is the i-th formation pressure data; R(X, Y) is the correlation coefficient;

[0071] According to the determined correlation coefficients, the sensitive logging parameters are selected from the determined logging parameters affecting the formation pressure; the sensitive parameters include at least one of well depth, drilling time, drilling pressure, rotary speed, torque, inlet density of drilling fluid, outlet density of drilling fluid, inlet temperature of drilling fluid, outlet temperature of drilling fluid, temperature difference of drilling fluid, and total hydrocarbon content.

[0072] The larger the value of the correlation coefficient, the greater the correlation between the parameter and the formation pressure. The determination of the correlation can obtain the specific variation law between each logging parameter and the formation pressure. For example, when encountering a high-pressure brine layer or abnormal high formation pressure, the porosity will increase, and parameters such as drilling speed, rotary speed, conductivity, temperature, and total hydrocarbon content will increase; under normal pressure, the formation pressure will increase exponentially with the increase of well depth, and the abnormal high pressure situation will increase relatively fast with the increase of well depth. In this embodiment, the screening criteria for the correlation coefficient can be selected according to the actual situation. For example, it can be taken as 0.6. Of course, the geological characteristics of each exploration area are different, and the obtained geological data are also different, so the screening criteria are also different. During the actual exploration process, the screening criteria should be determined in combination with the actual geological data, and no limitation is made here.

[0073] Preferably, in the above step S102, a deep learning network model is established according to the selected sensitive logging parameters, including:

[0074] A deep learning network model is established with the selected sensitive logging parameters as input parameters and the formation pressure as output parameters according to the selected sensitive logging parameters;

[0075] Establishing a deep learning network model including the corresponding relationship between the sensitive logging parameters and the formation pressure by using the selected sensitive parameters can better fit the formation pressure according to the variation law between each parameter and the formation pressure, so that the formation pressure predicted by the model will be closer to the real formation pressure. Because the variation law between each parameter and the formation pressure is different, the structures of the deep learning network models established based on different sensitive parameters are different, and the predicted formation pressures are also different.

[0076] The neural network learning model is a BP neural network model. The BP neural network model includes an input layer, a first hidden layer, a second hidden layer, and an output layer; the input layer is used to input the selected sensitive logging parameters, the output layer is used to output the predicted formation pressure; the first hidden layer is used to extract features from the input sensitive logging parameters, and the second hidden layer is used to perform secondary feature extraction on the input sensitive logging parameters.

[0077] The ordinary BP neural network model includes an input layer, a hidden layer, and an output layer. The neural network model adopted in this method adds two hidden layers, performs multiple feature extractions on the sensitive logging parameters input into the model, can better linearly divide different types of logging parameters, more accurately fit the relationship between the sensitive logging parameters and the formation pressure, and the predicted formation pressure will also be more accurate.

[0078] Preferably, in the above step S103, the pre-established pressure data set is input into the deep learning network model for training, and the trained deep learning network model is used as the formation pressure monitoring model, including:

[0079] The sensitive logging parameters and the actual formation pressure included in the pressure data set are input into the deep learning network model, the predicted formation pressure is output, the anti-normalization is performed on the predicted formation pressure to obtain the anti-normalized predicted formation pressure, and based on the actual formation pressure and the predicted formation pressure output by the model, the parameters of the deep learning network model are adjusted, and the process of model training is repeated until the model meets the preset convergence conditions, and the trained deep learning network model is obtained as the formation pressure monitoring model.

[0080] In this embodiment, through the above formation pressure monitoring method, the logging parameters of Well Hetan 1 in Area A are collected for training, and the training results are as Figure 2 shown, Figure 2 It can be seen from [the figure] that the more the number of data sets participating in the training, the closer the predicted value of the formation pressure is to the true value of the formation pressure.

[0081] The more the number of adjacent wells participating in the training, the more complete the logging parameter data, the better the model training effect, the more accurate the formation pressure monitoring result, and this method has strong inclusiveness for the data quantity. The process of training the model is to continuously adjust the parameters of the model, and finally obtain a deep learning network model with better prediction effect. At the same time, adding an activation function between the hidden layer and the output layer of the deep learning network model can enable the deep learning network model to adapt to various complex data distributions and relationships, improve the expression ability and fitting ability of the network. The activation function added in this embodiment is: F(x) is the activation value, x is the output parameter data of the second activation layer, and also the input parameter data of the output layer.

[0082] The anti-normalization is performed on the predicted formation pressure through the following formula to obtain the anti-normalized predicted formation pressure Y new =Y new *(MAX(Y i ) - MIN(Y i )) + MIN(Y i ), where Y is the predicted formation pressure value, MAX(Y i) is the maximum value of the predicted formation pressure output by the model, MIN(Y i ) is the minimum value of the predicted formation pressure output by the model.

[0083] Preferably, based on the actual formation pressure and the predicted formation pressure output by the model, the parameters of the deep learning network model are adjusted, including:

[0084] Determine the model error of the deep learning network model according to the actual formation pressure and the predicted formation pressure output by the model;

[0085] In the backpropagation stage, according to the model error, the gradient of each layer of the deep learning network is determined layer by layer through the chain rule, and the weights and thresholds of the deep learning network model are adjusted using the gradient descent method.

[0086] To measure the generalization ability and optimization effect of this prediction model, the evaluation indexes select the coefficient of determination (R 2 ) and the mean relative error MRE to evaluate the effect of the formation pressure monitoring model: Where y i , are the actual value of the formation pressure and the predicted value of the formation pressure respectively.

[0087] Based on the same inventive concept, the embodiment of the present invention also provides a device for constructing a formation pressure monitoring model. This device can be set in a device capable of processing computer instructions. The structure of this device is as Figure 3 shown, including:

[0088] The data acquisition module 11 is used to collect well logging data and mud logging data of at least one adjacent well in the target block;

[0089] The data analysis module 12 is used to determine the actual formation pressure of the adjacent well according to the well logging data of the adjacent well; determine the mud logging parameters affecting the formation pressure according to the mud logging data of the adjacent well; the mud logging parameters include lithology-related parameters, drilling fluid-related parameters, and gas logging-related parameters;

[0090] The model establishment module 13 is used to obtain the preferred sensitive mud logging parameters from the determined mud logging parameters affecting the formation pressure, establish a deep learning network model according to the preferred sensitive mud logging parameters; train the deep learning network model with a pre-established pressure data set to obtain a formation pressure monitoring model; the pressure data set includes the actual formation pressure of the adjacent well and the preferred sensitive mud logging parameters.

[0091] Regarding the device for constructing a formation pressure monitoring model in the above embodiments, the specific ways in which each module performs operations have been described in detail in the embodiments related to this method, and will not be elaborated here.

[0092] The above method and device of this embodiment collect the logging data and mud logging data of at least one adjacent well in the target block. According to the logging data of the adjacent well, the actual formation pressure of the adjacent well is determined; according to the mud logging data of the adjacent well, the mud logging parameters affecting the formation pressure are determined; among them, the logging data includes the basic logging formation pressure data of the adjacent well; the mud logging parameters include three categories: lithology parameters, drilling fluid parameters, and gas logging parameters. This method selects a large number of parameters, and when obtaining the parameters, not only the influence of lithology on the formation pressure is considered, but also the influence of drilling fluid performance and hydrocarbon content on the formation pressure is considered; the optimized sensitive mud logging parameters are obtained from the determined mud logging parameters, and a deep learning network model is established in combination with the optimized sensitive mud logging parameters; the deep learning network model established according to the optimized sensitive parameters can better fit the corresponding relationship between the sensitive mud logging parameters and the formation pressure, so as to more accurately predict the formation pressure; the pre-established pressure data set is input into the deep learning network model for training. Through multiple iterative trainings, the optimization effect of the model is more prominent. Finally, the trained deep learning network model is used as the formation pressure prediction model. This model can improve the accuracy and precision of formation pressure prediction, shorten the time of drilling work, improve the drilling efficiency, and has practical value.

[0093] Embodiment 2

[0094] Embodiment 2 of the present invention provides a specific implementation process of a formation pressure monitoring method, and its process is as Figure 4 shown, including the following steps:

[0095] Step S201: Obtain the mud logging data of the target well to be monitored in the target block, and determine the sensitive mud logging parameters according to the mud logging parameters included in the mud logging data;

[0096] Step S202: Input the sensitive mud logging parameters into the corresponding formation pressure monitoring model, and output the predicted formation pressure. The formation pressure monitoring model is trained by using the above formation pressure monitoring model construction method.

[0097] For example, in this embodiment, the deep learning network model is trained according to the mud logging parameters of Well Hetan 1 in Area A collected in Embodiment 1 above. The trained neural learning network model is used as the formation pressure monitoring model, and the formation pressure monitoring model is applied to Well Hetan 101 in Area A. Well Hetan 101 in Area A is the well being drilled. The sensitive mud logging parameters of Well Hetan 101 in Area A are collected as the data of the test set, and the formation pressure of Well Hetan 101 in Area A being drilled is predicted by using the formation pressure monitoring model. The prediction result is as Figure 5 shown, Figure 5 The comparison result between the true value of the formation pressure and the predicted value of the formation pressure of Well Hetan 101 is given in, and from Figure 5It can be concluded that the richer the dataset, the more accurate the formation pressure prediction result. Experimental data has verified the practicability of this method, which can provide technical support for the exploration of oil and gas wells.

[0098] Table 1 is the error table of the application of this method in Well LT1, including the measured formation pressure value at the corresponding well depth, the predicted formation pressure value by machine learning, and the calculated error. It can be seen from Table 1 that the final errors are all within 10%, which proves the practical significance of this method, can be applied in the field of drilling engineering technology, accurately predict the formation pressure, and shorten the construction period.

[0099] Table 1

[0100]

[0101] Steps S201 to S202 realize the monitoring of the formation pressure by using the pre-established formation pressure monitoring model, and finally obtain the predicted value of the formation pressure. After comparing with the measured formation pressure value, this method can accurately predict the formation pressure at the corresponding well depth, and can provide an effective data basis and strong technical support for the construction work during the actual drilling process.

[0102] Based on the same inventive concept, an embodiment of the present invention provides a formation pressure monitoring device, which can be set in a device capable of processing computer instructions, and the structure of the device is as Figure 6 shown, including:

[0103] A data analysis module 21, configured to obtain the logging data of the target well to be monitored in the target block, and determine the sensitive logging parameters according to the logging parameters included in the logging data;

[0104] A pressure prediction module 22, configured to input the sensitive logging parameters into the pre-established formation pressure monitoring model, and output the predicted formation pressure, where the formation pressure monitoring model is trained by using the above-mentioned formation pressure monitoring model construction method.

[0105] Regarding the formation pressure monitoring device in the above embodiment, the specific manners in which each module performs operations have been described in detail in the embodiments related to this method, and will not be elaborated here.

[0106] In the above methods and devices of the embodiments of the present invention, by obtaining the logging data of the target well to be monitored in the target block, the sensitive logging parameters are determined according to the logging parameters included in the logging data; the sensitive logging parameters are input into the corresponding formation pressure monitoring model, and the predicted formation pressure is output. The formation pressure monitoring model is trained by using the above-mentioned formation pressure monitoring model construction method. The formation pressure monitoring models established by using different logging parameters are targeted and can accurately reflect the law of the change of sensitive logging parameters and formation pressure, so as to better predict the underground formation pressure, improve the drilling rate, and shorten the drilling cycle.

[0107] The embodiments of the present invention further provide a computer storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, the formation pressure monitoring model construction method or the formation pressure monitoring method is implemented.

[0108] The embodiments of the present invention further provide a computer device, which includes: a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the formation pressure monitoring model construction method or the formation pressure monitoring method is implemented.

[0109] Unless otherwise specifically stated, terms such as processing, computing, operating, determining, displaying, etc. may refer to the actions and / or processes of one or more processing or computing systems, or similar devices, which operate on and transform data represented as physical (such as electronic) quantities in the registers or memories of the processing system into other data similarly represented as physical quantities in the memories, registers, or other such information storage, transmission, or display devices of the processing system. Information and signals can be represented by any of a variety of different technologies and methods. For example, the data, instructions, commands, information, signals, bits, symbols, and chips mentioned throughout the above description can be represented by voltage, current, electromagnetic waves, magnetic fields or particles, optical fields or particles, or any combination thereof.

[0110] It should be understood that the specific order or hierarchy of steps in the disclosed processes is an example of an exemplary method. Based on design preferences, it should be understood that the specific order or hierarchy of steps in the process can be rearranged without departing from the scope of the present disclosure. The appended method claims present the elements of the various steps in an exemplary order and are not intended to be limited to the specific order or hierarchy.

[0111] In the foregoing detailed description, various features are combined in a single embodiment to simplify the present disclosure. This method of disclosure should not be interpreted as reflecting an intention that the embodiments of the claimed subject matter require more features than are expressly recited in each claim. On the contrary, as reflected in the appended claims, the invention lies in less than all of the features of a single disclosed embodiment. Accordingly, the appended claims are hereby expressly incorporated into the detailed description, with each claim standing on its own as a separate preferred embodiment of the invention.

[0112] Those skilled in the art should also understand that the various illustrative logical blocks, modules, circuits, and algorithm steps described in connection with the embodiments herein can be implemented as electronic hardware, computer software, or combinations thereof. To clearly illustrate the interchangeability of hardware and software, the various illustrative components, blocks, modules, circuits, and steps have been generally described in terms of their functionality. Whether such functionality is implemented as hardware or software depends upon the particular application and the design constraints imposed on the overall system. Skilled artisans may implement the described functionality in a flexible manner for each particular application, but such implementation decisions should not be interpreted as departing from the scope of the present disclosure.

[0113] The steps of a method or algorithm described in connection with the embodiments herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination thereof. The software module may be located in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium well known in the art. An exemplary storage medium is coupled to the processor such that the processor can read information from, and write information to, the storage medium. Of course, the storage medium may also be integral to the processor. The processor and the storage medium may be located in an ASIC. The ASIC may be located in a user terminal. Of course, the processor and the storage medium may also exist as discrete components in a user terminal.

[0114] For a software implementation, the techniques described in this application can be implemented using modules (e.g., procedures, functions, etc.) that perform the functions described herein. These software codes can be stored in a memory unit and executed by a processor. The memory unit may be implemented within the processor or outside the processor, and in the latter case, it is coupled to the processor in a communication manner by various means, which are well known in the art.

[0115] The foregoing description includes examples of one or more embodiments. Of course, it is not possible to describe all possible combinations of components or methods for the purpose of describing the above embodiments, but those of ordinary skill in the art should recognize that the various embodiments can be further combined and arranged. Accordingly, the embodiments described herein are intended to embrace all such changes, modifications, and variations that fall within the scope of the appended claims. Further, with respect to the term "comprising" used in the specification or claims, this term is inclusive in a manner similar to the term "including," as that term is interpreted when used as a transitional word in a claim. Additionally, any use of the term "or" in the claims or specification is to be meant "non-exclusive or."

Claims

1. A method for constructing a formation pressure monitoring model, characterized in that, it includes: Collect well logging data and mud logging data of at least one adjacent well in the target block, and determine the actual formation pressure of the adjacent well according to the well logging data of the adjacent well; Determine the mud logging parameters affecting the formation pressure according to the mud logging data of the adjacent well; the mud logging parameters include lithology-related parameters, drilling fluid-related parameters, and gas logging-related parameters; Obtain the preferred sensitive mud logging parameters from the determined mud logging parameters affecting the formation pressure, and establish a deep learning network model according to the preferred sensitive mud logging parameters; Train the deep learning network model with a pre-established pressure data set to obtain a formation pressure monitoring model; the pressure data set includes the actual formation pressure of the adjacent well and the preferred sensitive mud logging parameters.

2. The method according to claim 1, characterized in that, after collecting the well logging data and mud logging data of at least one adjacent well in the target block and before determining the actual formation pressure of the adjacent well, it further includes: Performing outlier rejection and normalization processing on the well logging parameters included in the well logging data and the mud logging parameters included in the mud logging data.

3. The method according to claim 2, characterized in that, The following formula is used to eliminate outliers from well logging parameters and mud logging parameters: where n is the number of parameters, and x i is the data of the i-th parameter; The logging parameters and mud logging parameters are normalized using the following formula: Among them, represents the result of normalizing the i-th parameter data, and x i is the i-th parameter data, MAX(X i ) is the maximum data in the i-th parameter data, and MIN(X i ) is the minimum data in the i-th parameter data.

4. The method according to claim 1, characterized in that, The determining the mud logging parameters affecting the formation pressure according to the mud logging data of the adjacent well includes: Determining the mud logging parameters affecting the formation pressure from the mud logging data according to the causes of abnormal formation pressure and the variation law of mud logging data; wherein: The lithology-related parameters include at least one of well depth, drilling time, drilling pressure, rotary speed, and torque; the drilling fluid-related parameters include at least one of drilling fluid inlet flow rate, drilling fluid outlet flow rate, drilling fluid flow rate difference, drilling fluid inlet density, drilling fluid outlet density, drilling fluid density difference, drilling fluid inlet temperature, drilling fluid outlet temperature, drilling fluid temperature difference, drilling fluid inlet conductivity, drilling fluid outlet conductivity, and drilling fluid outlet conductivity difference; the gas logging-related parameters include total hydrocarbon content.

5. The method according to claim 1, characterized in that, The obtaining the preferred sensitive mud logging parameters from the determined mud logging parameters affecting the formation pressure includes: Determining the correlation coefficient between the mud logging parameters and the formation pressure based on the following formula: Among them, n is the number of mud logging parameters, x i is the data of the i-th parameter, y i is the data of the i-th formation pressure; R(X, Y) is the correlation coefficient; According to the determined correlation coefficient, preferably select the sensitive mud logging parameters from the determined mud logging parameters affecting the formation pressure; the sensitive parameters include at least one of well depth, drilling time, drilling pressure, rotary speed, torque, drilling fluid inlet density, drilling fluid outlet density, drilling fluid inlet temperature, drilling fluid outlet temperature, drilling fluid temperature difference, and total hydrocarbon content.

6. The method according to claim 1, characterized in that, Establishing a deep learning network model according to the preferred sensitive mud logging parameters includes: Establishing a deep learning network model with the preferred sensitive mud logging parameters as input parameters and the formation pressure as output parameters according to the preferred sensitive mud logging parameters. The neural network learning model is a BP neural network model, which includes an input layer, a first hidden layer, a second hidden layer, and an output layer; the input layer is used to input the preferably sensitive mud logging parameters, and the output layer is used to output the predicted formation pressure; the first hidden layer is used to extract features from the input sensitive mud logging parameters, and the second hidden layer is used to perform secondary feature extraction on the input sensitive mud logging parameters.

7. The method according to claim 1, wherein a pre-established pressure data set is input into the deep learning network model for training, and the trained deep learning network model is used as a formation pressure monitoring model. It includes: The sensitive mud logging parameters and the actual formation pressure included in the pressure data set are input into the deep learning network model, the predicted formation pressure is output, the predicted formation pressure is anti-normalized to obtain the anti-normalized predicted formation pressure, and based on the actual formation pressure and the predicted formation pressure output by the model, the parameters of the deep learning network model are adjusted, and the process of model training is repeated until the model meets the preset convergence condition, and the trained deep learning network model is obtained as the formation pressure monitoring model.

8. The method according to claim 7, characterized in that the adjusting the parameters of the deep learning network model based on the actual formation pressure and the predicted formation pressure output by the model includes: determining the model error of the deep learning network model according to the actual formation pressure and the predicted formation pressure output by the model; in the backpropagation stage, according to the model error, the gradient of each layer of the deep learning network is determined layer by layer through the chain rule, and the weights and thresholds of the deep learning network model are adjusted using the gradient descent method.

9. A device for constructing a formation pressure monitoring model, characterized in that it includes: a data acquisition module, configured to collect well logging data and mud logging data of at least one adjacent well in the target block; a data analysis module, configured to determine the actual formation pressure of the adjacent well according to the well logging data of the adjacent well; determine the mud logging parameters affecting the formation pressure according to the mud logging data of the adjacent well; the mud logging parameters include lithology-related parameters, drilling fluid-related parameters, and gas logging-related parameters; a model establishment module, configured to obtain preferably sensitive mud logging parameters from the determined mud logging parameters affecting the formation pressure, establish a deep learning network model according to the preferably sensitive mud logging parameters; train the deep learning network model with a pre-established pressure data set to obtain a formation pressure monitoring model; the pressure data set includes the actual formation pressure of the adjacent well and the preferably sensitive mud logging parameters.

10. A formation pressure monitoring method, characterized in that it includes: obtaining the mud logging data of the target well to be monitored in the target block, and determining the sensitive mud logging parameters according to the mud logging parameters included in the mud logging data; inputting the sensitive mud logging parameters into the corresponding formation pressure monitoring model, and outputting the predicted formation pressure, where the formation pressure monitoring model is trained by using the formation pressure monitoring model construction method according to any one of claims 1-8.

11. A formation pressure monitoring device, characterized in that it includes: A data analysis module, configured to obtain well logging data of a target well to be monitored in a target block, and determine sensitive well logging parameters according to the well logging parameters included in the well logging data; A pressure prediction module, configured to input the sensitive well logging parameters into a pre-established formation pressure monitoring model, and output a predicted formation pressure, where the formation pressure monitoring model is trained by using the formation pressure monitoring model construction method according to any one of claims 1-8.

12. A computer storage medium, characterized in that, the computer storage medium stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, the formation pressure monitoring model construction method according to any one of claims 1-8 or the formation pressure monitoring method according to claim 10 is implemented.

13. A computer device, characterized in that, it includes: a memory, a processor, and a computer program stored on the memory and executable on the processor, and when the processor executes the program, the formation pressure monitoring model construction method according to any one of claims 1-8 or the formation pressure monitoring method according to claim 10 is implemented.

Images