Prediction method and device for airlock in production well, equipment, medium and product

Abstract

The invention discloses a prediction method, device and equipment for airlock in a production well, a medium and a product, and relates to the field of oil and gas field development and production. The method comprises the steps that first production parameters corresponding to a first production well are collected, the first production well comprises first liquid to be collected and first gas generated in the collection process, and the first production parameters are used for indicating the production state of the first production well in the process of collecting the first liquid based on an infusion pump; the pump fullness degree corresponding to the first production well is predicted based on the first production parameters through a first prediction model, a first prediction result is obtained, the first prediction result is used for indicating the fullness degree of first liquid in the infusion pump, and the first prediction model is a pre-trained machine learning model used for predicting the gas influence degree of the collection equipment; and determining an airlock state corresponding to the first production well based on the first prediction result, wherein the airlock state is used for indicating the degree of influence of the first gas on the production performance of the first production well.

Description

Technical Field

[0001] The present application relates to the field of oil and gas field development and production, and in particular to a method, device, equipment, medium and product for predicting gas lock in a production well. Background Art

[0002] In the development of oil and gas fields using Carbon Capture, Utilization and Storage (CCUS), carbon dioxide (CO2) is injected into underground reservoirs to displace crude oil, a process known as CO2 Enhanced Oil Recovery (EOR). CCUS production wells generally have a high gas-liquid ratio. As CCUS production well development enters the middle and late stages, the CO2 gas drive front gradually approaches the production well, resulting in an increase in the gas ratio in the wellbore, which may cause gas impact or gas lock.

[0003] In the related technology, production data of conventional high gas-liquid ratio wells are collected and analyzed, including oil production, water content, gas production, etc. When the gas ratio increases, these parameters usually change, thereby indirectly reflecting the possibility of gas lock by observing the changes in production data.

[0004] However, the oil and gas environment in the oil and gas fields where CCUS is applied is complex. It is difficult to accurately analyze the gas lock situation by simply observing the production data. Introducing more complex data makes it difficult to ensure the efficiency and timeliness of the analysis, and the efficiency of gas lock analysis is low. Summary of the invention

[0005] The embodiment of the present application provides a method, device, equipment, medium and product for predicting gas lock in a production well, which can improve the efficiency of gas lock analysis. The technical solution is as follows.

[0006] In one aspect, a method for predicting gas lock in a production well is provided, the method comprising:

[0007] Acquire a first production parameter corresponding to a first production well, wherein the first production well includes a first liquid to be collected and a first gas generated during the collection process, and the first production parameter is used to indicate a production state of the first production well during the process of collecting the first liquid based on a liquid pump, and the liquid pump is a collection device for collecting the first liquid;

[0008] Predicting the filling degree of the pump corresponding to the first production well based on the first production parameter by a first prediction model to obtain a first prediction result, wherein the first prediction result is used to indicate the filling degree of the first liquid in the liquid pump, and the filling degree of the first liquid in the liquid pump is used to characterize the degree to which the liquid pump is affected by the first gas, and the first prediction model is a pre-trained machine learning model for predicting the degree to which the acquisition equipment is affected by the gas;

[0009] A gas lock state corresponding to the first production well is determined based on the first prediction result, and the gas lock state is used to indicate the extent to which the production performance of the first production well is affected by the first gas.

[0010] In another aspect, a device for predicting gas lock in a production well is provided, the device comprising:

[0011] an acquisition module, configured to acquire a first production parameter corresponding to a first production well, wherein the first production well includes a first liquid to be collected and a first gas generated during the collection process, and the first production parameter is used to indicate a production state of the first production well during the process of collecting the first liquid based on a liquid pump, and the liquid pump is a collection device for collecting the first liquid;

[0012] a processing module, configured to predict the filling degree of the pump corresponding to the first production well based on the first production parameter by using a first prediction model to obtain a first prediction result, wherein the first prediction result is used to indicate the filling degree of the first liquid in the liquid pump, and the filling degree of the first liquid in the liquid pump is used to characterize the degree to which the liquid pump is affected by the first gas, wherein the first prediction model is a pre-trained machine learning model for predicting the degree to which a collection device is affected by gas;

[0013] The processing module is further used to determine a gas lock state corresponding to the first production well based on the first prediction result, and the gas lock state is used to indicate the degree to which the production performance of the first production well is affected by the first gas.

[0014] In some embodiments, the processing module is further configured to determine that gas lock occurs in the first production well when the fullness level indicated by the first prediction result reaches a first preset range.

[0015] In some embodiments, the processing module is further used to dynamically determine the first preset degree range based on the production status of the first production well.

[0016] In some embodiments, the acquisition module is further used to acquire sample production parameters, where the sample production parameters are production parameters collected within a historical time period for the first production well;

[0017] The processing module is further used to calculate a reference fullness based on the sample production parameters, wherein the reference fullness is used to indicate the degree to which the liquid pump is affected by the first gas when the liquid pump is in the production state indicated by the sample production parameters;

[0018] The processing module is further used to predict the pump filling degree corresponding to the first production well based on the sample production parameters through a candidate prediction model to obtain a sample prediction result;

[0019] The processing module is further used to determine a predicted loss value based on a difference between the sample prediction result and the reference fullness;

[0020] The processing module is further used to train the candidate prediction model based on the prediction loss value to obtain the first prediction model.

[0021] In some embodiments, the processing module is further configured to:

[0022] calculating a deviation factor and a dissolved gasoline ratio of the first gas based on the sample production parameters;

[0023] The reference filling level is calculated based on the deviation factor, the dissolved gasoline ratio, and the sample production parameters.

[0024] In some embodiments, the first production well comprises a carbon capture utilization and storage CCUS production well, and the first gas comprises natural gas containing carbon dioxide CO2;

[0025] The first production parameters include at least:

[0026] the oil production of the first production well;

[0027] the water content of the first production well;

[0028] the gas production of the first production well;

[0029] The molar content of CO2 in the first production well;

[0030] Dynamic liquid level parameters of the first production well;

[0031] a pump stroke of the liquid pump;

[0032] pump strokes of the liquid pump;

[0033] The pumping depth of the liquid pump.

[0034] On the other hand, a computer device is provided, which includes a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set are loaded and executed by the processor to implement a method for predicting gas lock in a production well as described in any of the above-mentioned embodiments of the present application.

[0035] On the other hand, a computer-readable storage medium is provided, wherein at least one instruction, at least one program, a code set or an instruction set is stored in the storage medium, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by a processor to implement a method for predicting gas lock in a production well as described in any of the above-mentioned embodiments of the present application.

[0036] On the other hand, a computer program product or a computer program is provided, the computer program product or the computer program comprising computer instructions, the computer instructions being stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the method for predicting gas lock in a production well as described in any of the above embodiments.

[0037] The beneficial effects brought by the technical solution provided by the embodiment of the present application include at least:

[0038] By collecting production parameters to characterize the production status of the production well, a pre-trained first prediction model is used to predict the pump fullness based on the production parameters, so that the gas lock state is determined according to the pump fullness. This allows a machine learning model that can predict the degree to which the acquisition equipment is affected by gas to participate in the prediction of the gas lock state, thereby reducing the cost of human analysis and improving the automation of gas lock prediction. In addition, since the model operation takes a short time, the timeliness of the gas lock prediction can be guaranteed, thereby improving the efficiency of the gas lock prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0040] Figure 1 is a schematic diagram of an air lock prediction system provided by an exemplary embodiment of the present application;

[0041] Figure 2 is a flow chart of a method for predicting gas lock in a production well provided by an exemplary embodiment of the present application;

[0042] Figure 3 It is a schematic diagram of an intelligent early warning process of a CCUS production well gas lock condition provided by an exemplary embodiment of the present application;

[0043] Figure 4 is a schematic diagram of the pump filling degree provided by an exemplary embodiment of the present application;

[0044] Figure 5 is a schematic diagram of a long short-term memory structure provided by an exemplary embodiment of the present application;

[0045] Figure 6 is a schematic diagram of an attention mechanism provided by an exemplary embodiment of the present application;

[0046] Figure 7 This is a schematic diagram of the Attention-LSTM neural network structure provided by an exemplary embodiment of the present application;

[0047] Figure 8 is a structural block diagram of a prediction device for gas lock in a production well provided by an exemplary embodiment of the present application;

[0048] Fig. 9 It is a structural block diagram of a terminal provided by an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0049] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

[0050] It should be understood that although the terms first, second, etc. may be used in the present disclosure to describe various information, such information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the present disclosure, the first parameter may also be referred to as the second parameter, and similarly, the second parameter may also be referred to as the first parameter. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0051] In the development of oil and gas fields where CCUS is applied, CO2 is injected into underground reservoirs to displace crude oil, a process known as CO2 enhanced oil recovery. CCUS production wells generally have a high gas-liquid ratio. As the development of CCUS production wells enters the middle and late stages, the CO2 gas drive front gradually approaches the production wells, resulting in an increase in the gas ratio in the wellbore, which may cause gas impact or gas lock. In related technologies, production data of oil wells, including oil production, water content, gas production, etc., are collected and analyzed for conventional high gas-liquid ratio wells. When the gas ratio increases, these parameters usually change, thereby indirectly reflecting the possibility of gas lock by observing changes in production data. However, the oil and gas environment in oil and gas fields where CCUS is applied is complex, and it is difficult to accurately analyze the gas lock situation by simply observing the production data. The introduction of more complex data makes it difficult to ensure analysis efficiency and timeliness, and the efficiency of gas lock analysis is low.

[0052] For high gas-liquid ratio wells, the free gas, dissolved gas and condensate gas in the pump cavity occupy a part of the volume, reducing the effective working space of the pump and the filling degree of the pump, thereby reducing the amount of liquid that can be pumped out per stroke of the pump and reducing the pump efficiency. When the gas-liquid ratio in the pump is too high, the ball valve in the oil pump cannot work normally, and even gas lock conditions may occur. For CCUS production wells, it is particularly important to predict the filling degree of the oil pump, warn the gas lock state of the oil pump in advance, and take effective anti-gas measures in time to improve the production efficiency of CCUS production wells. However, traditional oil and gas well production monitoring and management methods often find it difficult to predict and warn the pump filling degree and gas lock conditions, especially when CO2 is used as a displacement medium, its fluid properties are different from those of conventional high gas-liquid ratio wells. Therefore, related technologies are often not sufficient to effectively predict and solve problems such as production interruptions and capacity reduction caused by CO2 gas locks.

[0053] Traditional methods for diagnosing and predicting the working conditions of pumping wells fail to fully consider the unique production conditions of CCUS, especially the complex gas-liquid two-phase flow characteristics in production wells during CO2 flooding. These traditional methods often make delayed diagnosis of the current working conditions based on current production data. Gas lock not only affects the production efficiency of production wells, but may also damage wellbore and ground facilities, increasing operating costs.

[0054] Therefore, the pump filling degree prediction and gas lock condition warning for CCUS production wells are particularly important. An accurate and timely warning mechanism is the basis for ensuring the smooth operation of CCUS production, optimizing the CO2 injection strategy, and taking early adjustment measures to avoid gas lock.

[0055] The method for predicting gas lock in a production well provided in an embodiment of the present application collects production parameters to characterize the production status of the production well, and uses a pre-trained first prediction model to predict the pump fullness based on the production parameters, thereby determining the gas lock status according to the pump fullness, so that a machine learning model capable of predicting the degree to which the collection equipment is affected by gas is involved in the prediction of the gas lock status, thereby freeing up human analysis costs and improving the automation of gas lock prediction. In addition, since the model operation takes a short time, the timeliness of the gas lock prediction can be guaranteed, thereby improving the efficiency of the gas lock prediction.

[0056] First, the air lock prediction system of this application is introduced. Please refer to Figure 1 , which shows a schematic diagram of an air lock prediction system provided by an exemplary embodiment of the present application, and the air lock prediction system includes: a terminal 110.

[0057] The terminal 110 is deployed with a first prediction model for participating in the prediction of the airlock state.

[0058] In some embodiments, the terminal 110 stores production parameters corresponding to the production wells, which are used as model inputs to participate in gas lock prediction.

[0059] Illustratively, the terminal 110 obtains a first production parameter corresponding to a first production well, which includes a first liquid to be collected and gas generated during the collection process. The first production parameter is used to indicate the production status of the first production well during the process of collecting the first liquid based on a liquid pump, and the liquid pump is a collection device for collecting the first liquid.

[0060] Terminal 110 predicts the degree of fullness of the pump corresponding to the first production well based on the first production parameter through the first prediction model, and obtains a first prediction result. The first prediction result is used to indicate the degree of fullness of the first liquid in the pumping pump. The degree of fullness of the first liquid in the pumping pump is used to characterize the degree to which the pumping pump is affected by the first gas. The first prediction model is a pre-trained machine learning model for predicting the degree to which the collection equipment is affected by the gas.

[0061] The terminal 110 determines a gas lock state corresponding to the first production well based on the first prediction result, where the gas lock state is used to indicate the degree to which the production performance of the first production well is affected by the first gas.

[0062] In some embodiments, the terminal 110 may be connected to the first production well, and the terminal 110 may obtain the first production parameter by collecting the production status in real time.

[0063] The terminal 110 may also be a remote analysis device corresponding to the first production well. The implementation environment also includes a server 120 and a communication network 130 . The terminal 110 and the server 120 perform data transmission via the communication network 130 .

[0064] The server 120 is provided with a data acquisition device for acquiring production parameters corresponding to the first production well, and storing the acquired production parameters in the server 120 .

[0065] When it is necessary to predict the gas lock in the production well, the terminal 110 reads the first production parameter from the server 120, predicts the pump fullness corresponding to the first production well based on the first production parameter through the first prediction model, obtains a first prediction result, and determines the gas lock state corresponding to the first production well based on the first prediction result. The gas lock state is used to indicate the degree to which the production performance of the first production well is affected by the first gas.

[0066] Optionally, the above interaction methods are only illustrative examples and are not limited in this application.

[0067] The above-mentioned terminal is optional. The terminal can be a desktop computer, a laptop computer, a mobile phone, a tablet computer, an e-book reader, a Moving Picture Experts Group Audio Layer III (MP3) player, a Moving Picture Experts Group Audio Layer IV (MP4) player, a smart TV, a smart car, and other forms of terminal devices, which are not limited to the embodiments of the present application.

[0068] It is worth noting that the above-mentioned servers can be independent physical servers, or they can be server clusters or distributed systems composed of multiple physical servers. They can also be cloud servers that provide basic cloud computing services such as cloud services, cloud security, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDN), as well as big data and artificial intelligence platforms.

[0069] Among them, cloud technology refers to a hosting technology that unifies hardware, software, network and other resources within a wide area network or local area network to achieve data computing, storage, processing and sharing.

[0070] In some embodiments, the above server can also be implemented as a node in a blockchain system.

[0071] It should be noted that the information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data used for analysis, stored data, displayed data, etc.) and signals involved in this application are all authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards in the relevant regions. For example, the production data involved in this application are all obtained with full authorization.

[0072] To further explain, before collecting relevant user data (for example, production parameters involved in this application, etc.) and during the process of collecting relevant user data, this application can display a prompt interface, pop-up window or output voice prompt information. The prompt interface, pop-up window or voice prompt information is used to prompt the user that its relevant data is currently being collected, so that this application only starts to execute the relevant steps of obtaining user-related data after obtaining the user's confirmation operation on the prompt interface or pop-up window, otherwise (that is, when the user's confirmation operation on the prompt interface or pop-up window is not obtained), the relevant steps of obtaining user-related data are terminated, that is, the user's relevant data is not obtained. In other words, all user data collected by this application are collected with the user's consent and authorization, and the collection, use and processing of relevant user data need to comply with relevant laws, regulations and standards in the relevant regions.

[0073] For illustration, please refer to Figure 2 , which shows a flowchart of a method for predicting gas lock in a production well provided by an exemplary embodiment of the present application. The method can be executed by a terminal, or by a server, or by both a terminal and a server. The present application embodiment takes the method executed by a terminal as an example for explanation. Figure 2 As shown, the method comprises the following steps:

[0074] Step 210, obtaining a first production parameter corresponding to the first production well.

[0075] The first production well includes a first liquid to be collected and a first gas generated during the collection process. The first production parameter is used to indicate the production status of the first production well during the collection of the first liquid based on a pumping pump. The pumping pump is a collection device for collecting the first liquid.

[0076] Optionally, the first production well comprises a CCUS oil well, the first gas comprises natural gas containing CO2, the first liquid is oil to be collected from the CCUS production well, and the liquid extraction pump is an oil extraction pump for collecting oil.

[0077] An oil well is a channel for extracting underground oil and natural gas, which allows underground oil and gas resources to be extracted to the surface through the channel.

[0078] Oil pumps are collection equipment installed in oil wells to lift crude oil from underground reservoirs to the surface.

[0079] Optionally, the types of the oil pump include but are not limited to a rod pump, a submersible electric pump, a gas lift, etc., which are not limited in the embodiments of the present application.

[0080] In some embodiments, the first production parameters include the oil production of the first production well, the water content of the first production well, the gas production of the first production well, the molar content of CO2 in the first production well, the dynamic liquid level parameters of the first production well, the pump stroke of the liquid pump, the pump strokes of the liquid pump, and the pumping depth of the liquid pump.

[0081] Oil production refers to the volume of crude oil produced by an oil well within a preset time period and is a key indicator for measuring the production capacity of an oil well.

[0082] Water content refers to the proportion of water in the fluid produced by an oil well. The water content directly affects the economic benefits of the oil well and the difficulty of crude oil processing.

[0083] Gas production refers to the amount of associated natural gas produced during the production process of an oil well, in addition to crude oil. Gas production refers to the output of this natural gas.

[0084] CO2 molar content In CCUS projects, the molar ratio of CO2 in the fluid produced by the oil well is an important parameter for measuring the efficiency of CO2 capture and utilization.

[0085] The dynamic liquid level is the free surface of the liquid in the oil well. It changes with the production of the oil well. The height of the dynamic liquid level can reflect the pressure state of the oil layer and the production status of the oil well.

[0086] The pump stroke refers to the distance the pump piston or plunger moves during one up and down reciprocating motion, which affects the amount of crude oil that can be pumped out in each stroke.

[0087] Pump strokes refer to the number of reciprocating times of the oil pump per minute. The speed of the strokes affects the oil production rate of the oil well.

[0088] The pump depth refers to the installation depth of the oil well pump in the oil well, that is, the depth from the wellhead to the pump.

[0089] Step 220: predicting the pump fullness corresponding to the first production well based on the first production parameter using the first prediction model to obtain a first prediction result.

[0090] The first prediction result is used to indicate the degree of fullness of the first liquid in the liquid pump, and the degree of fullness of the first liquid in the liquid pump is used to characterize the degree to which the liquid pump is affected by the first gas. The first prediction model is a pre-trained machine learning model for predicting the degree to which the collection device is affected by the gas.

[0091] The first prediction result is the pump filling degree of the liquid pump under the production state corresponding to the first production parameter, that is, the volume ratio of the liquid in each stroke of the liquid pump, which can reflect the degree to which the liquid pump is affected by the gas.

[0092] Optionally, the first prediction model is a Long Short-Term Memory (LSTM) model improved using an attention mechanism, which is capable of capturing long-term dependencies in time series data.

[0093] In some embodiments, a feature vector of the first production parameter is extracted by a feature extraction module of the first prediction model, and the weight of the feature vector is calculated using an attention mechanism. The feature weight is multiplied by the feature vector to participate in the prediction analysis, and the first prediction result is output.

[0094] Before using the first prediction model, a model training process is also included.

[0095] Optionally, the model training process includes the following steps:

[0096] The first step is to obtain sample production parameters.

[0097] The sample production parameters are production parameters collected during a historical period of time for the first production well.

[0098] For CCUS oil wells, the sample production parameters include oil production, water content, gas production, CO2 molar content, dynamic liquid level parameters, pump stroke, pump stroke times, pumping depth and other production parameters.

[0099] In the second step, a reference filling degree is calculated based on sample production parameters.

[0100] The reference filling level is used to indicate the degree to which the liquid pump is affected by the first gas under the production state indicated by the sample production parameters.

[0101] The reference fill level is the pump fill level that the pump corresponds to based on sample production parameters.

[0102] When calculating the reference fullness, it is necessary to first calculate the actual gas-liquid ratio of the fluid entering the liquid pump. The calculation of the actual fluid gas-liquid ratio requires attention to the gas deviation factor and the dissolved gasoline ratio. Therefore, the calculation process of the reference fullness includes calculating the deviation factor and the dissolved gasoline ratio of the first gas based on the sample production parameters, and calculating the reference fullness based on the deviation factor, the dissolved gasoline ratio and the sample production parameters.

[0103] Specifically, the critical temperature and critical gas pressure of natural gas can be calculated based on the situation that natural gas does not contain CO2. Since the natural gas in CCUS oil wells contains CO2, the critical temperature of natural gas needs to be corrected according to the molar fraction of CO2, thereby providing a basis for accurately calculating the gas deviation factor and dissolved gasoline ratio.

[0104] In the third step, the candidate prediction model is used to predict the pump fullness corresponding to the first production well based on the sample production parameters to obtain the sample prediction result.

[0105] Schematically, the feature vector of the sample production parameters is extracted through the feature extraction module of the candidate prediction model, and the weight of the feature vector is calculated using the attention mechanism. The feature weight is multiplied by the feature vector to participate in the prediction analysis, and the sample prediction result is output.

[0106] Optionally, the candidate prediction model is an LSTM model.

[0107] The production data of the first production well can be collected as sample production parameters at preset time intervals during the historical time period. Since the LSTM model can capture long-term dependencies in time series data, each time a prediction is made based on the latest collected sample production parameters, it will rely on the sample production parameters that have been input before and jointly participate in the prediction, thereby improving the model's prediction accuracy.

[0108] In the fourth step, the predicted loss value is determined based on the difference between the sample prediction result and the reference fullness.

[0109] Schematically, the sample prediction results output by the model are compared with the reference fullness for error, different output values ​​are derived through a preset loss function, and the error is transmitted to all neurons using the back propagation algorithm.

[0110] The fifth step is to train the candidate prediction model based on the prediction loss value to obtain the first prediction model.

[0111] In some embodiments, the model parameters are iteratively adjusted based on the prediction loss value until the preset training requirements are met to obtain a first prediction model.

[0112] Model parameters include model weights and bias values, etc.

[0113] Optionally, the preset training requirement includes at least one of the number of training times reaching a preset number threshold, or the predicted loss value reaching a preset loss threshold, or the predicted loss convergence, etc., which is not limited in the embodiments of the present application.

[0114] In schematic form, the weights and bias values ​​of the model are updated using the stochastic gradient descent method based on the predicted loss value, the adjustment amplitude of each model parameter is calculated based on the learning rate, the training steps of prediction loss calculation and model parameter adjustment are repeated based on each round of model input until a preset number threshold is reached, the model network parameters are saved, and the first prediction model is obtained.

[0115] Step 230: Determine the gas lock state corresponding to the first production well based on the first prediction result.

[0116] The gas lock state is used to indicate the extent to which the production performance of the first production well is affected by the first gas.

[0117] Optionally, different air lock states are determined according to a value range of the pump fullness in the first prediction result.

[0118] In some embodiments, when the fullness level indicated by the first prediction result reaches a first preset level range, it is determined that gas lock occurs in the first production well.

[0119] Optionally, the first preset degree range may be a fixed range determined based on the pump fullness when gas lock occurs in the first production well during a historical time period, or may be a dynamic range determined according to the state of the first production well.

[0120] For example, when the oil production, water content, gas production, etc. corresponding to the first production well change, the pump filling degree corresponding to the gas lock occurring in the first production well may also change.

[0121] Illustratively, an oil well in which the pump filling degree indicated by the first prediction result is less than or equal to 10% is recorded as a "gas lock" state, which is used to indicate that the oil well has experienced gas lock.

[0122] The oil wells whose pump filling degree indicated by the first prediction result is between 10% and 30% are recorded as "severe gas impact" status, which is used to indicate that the oil wells are currently severely affected by gas.

[0123] The oil wells whose pump filling degree indicated by the first prediction result is between 30% and 60% are recorded as "gas-affected" status, which is used to indicate that the current oil well is moderately affected by gas.

[0124] The oil wells whose pump filling degree indicated by the first prediction result is between 60% and 80% are recorded as "slightly gas-affected" status, which is used to indicate that the oil wells are currently slightly affected by gas.

[0125] The oil wells whose pump filling degree reaches 80% indicated by the first prediction result are recorded as being in a state of "almost no gas influence", which is used to indicate that the oil wells are currently affected by gas to a negligible degree.

[0126] In some embodiments, warning information is generated and displayed according to the gas lock state, and the warning information is used to prompt that a gas lock occurs in the first production well.

[0127] For example, when it is determined that the oil well is in a "gas lock" state, the warning information is displayed.

[0128] In some embodiments, different warning information may be generated and displayed for different airlock states.

[0129] Indicatively, when an oil well is in a "severe gas impact" state, an early warning message may be generated to indicate that a gas lock is about to occur in the first production well.

[0130] It is worth noting that the above-mentioned classification of air lock states and degree ranges, as well as the method of generating early warning information are only illustrative examples and are not limited in this application.

[0131] To summarize, the method provided in the embodiment of the present application characterizes the production status of the production well by collecting production parameters, and uses a pre-trained first prediction model to predict the pump fullness based on the production parameters, thereby determining the gas lock state according to the pump fullness, so that a machine learning model capable of predicting the degree to which the collection equipment is affected by gas is involved in the prediction of the gas lock state, thereby freeing up human analysis costs and improving the automation of gas lock prediction. In addition, since the model operation takes a short time, the timeliness of the gas lock prediction can be guaranteed, thereby improving the efficiency of the gas lock prediction.

[0132] In some embodiments, the prediction method for gas lock in production wells provided in the embodiments of the present application calculates the liquid filling degree in the downhole pump based on the comprehensive consideration of the gas-liquid ratio of oil well production, the physical properties of natural gas containing CO2, and the gas-liquid fluid state in the downhole pump. This method combines the long short-term memory network (LSTM) and the attention mechanism (AttentionMechanism), which can capture long-term dependencies in time series data. Then, the Attention-LSTM algorithm is used to build an intelligent prediction model for the gas impact degree of production wells, and the gas impact degree is used to predict whether the oil well has a gas lock risk.

[0133] For illustration, please refer to Figure 3 , Figure 3 FIG. 1 is a schematic diagram of an intelligent early warning process of a CCUS production well gas lock condition provided by an exemplary embodiment of the present application. Figure 3 As shown, the method comprises the following steps:

[0134] Step 310: Collect CCUS oil well production parameters.

[0135] The production parameters used to participate in model training in the training phase are sample production parameters, and the production parameters used to participate in model prediction in the application phase are first production parameters.

[0136] Schematically, the oil well production parameters of CCUS production wells during the development process are collected, including parameters such as oil production, water content, gas production, CO2 molar content, dynamic liquid level, pump stroke, pump stroke times, and pumping depth.

[0137] Oil production refers to the volume of crude oil produced by an oil well within a preset time period and is a key indicator for measuring the production capacity of an oil well.

[0138] Water content refers to the proportion of water in the fluid produced by an oil well. The water content directly affects the economic benefits of the oil well and the difficulty of crude oil processing.

[0139] Gas production refers to the amount of associated natural gas produced during the production process of an oil well, in addition to crude oil. Gas production refers to the output of this natural gas.

[0140] CO2 molar content In CCUS projects, the molar ratio of CO2 in the fluid produced by the oil well is an important parameter for measuring the efficiency of CO2 capture and utilization.

[0141] The dynamic liquid level is the free surface of the liquid in the oil well. It changes with the production of the oil well. The height of the dynamic liquid level can reflect the pressure state of the oil layer and the production status of the oil well.

[0142] The pump stroke refers to the distance the pump piston or plunger moves during one up and down reciprocating motion, which affects the amount of crude oil that can be pumped out in each stroke.

[0143] Pump strokes refer to the number of reciprocating times of the oil pump per minute. The speed of the strokes affects the oil production rate of the oil well.

[0144] The pump depth refers to the installation depth of the oil well pump in the oil well, that is, the depth from the wellhead to the pump.

[0145] Step 320, calculating the CO2-containing natural gas deviation factor and dissolved gas-oil ratio of the CCUS oil well.

[0146] When the CO2 content of natural gas is high, a large error may occur when using the conventional natural gas deviation factor calculation method. Optionally, the Wichert-Aziz-Cranmerf method is used in the embodiment of the present application to correct the deviation factor of natural gas containing CO2.

[0147] Next, the specific calculation process of the deviation factor of the CO2-containing natural gas and the dissolved gas-oil ratio is introduced, which specifically includes the following steps (1) to (7).

[0148] (1) When natural gas does not contain CO2, the calculation method of the critical pressure and critical temperature of natural gas can refer to the following formula 1 and formula 2:

[0149] p pc=4.67+0.10γ g -0.26γ g 2 Formula 1,

[0150] T pc =93.3+180.6γ g -6.9γ g 2 Formula 2,

[0151] Among them, p pc is the critical pressure of natural gas, in MPa; T pc is the critical temperature of natural gas, in K; γ g is the relative density of natural gas, without unit.

[0152] (2) When natural gas contains CO2, the critical temperature of natural gas needs to be corrected. Please refer to the following formula 3 and formula 4:

[0153] T pc ′=T pc -ε formula 3,

[0154]

[0155] Where, ε is the critical temperature correction coefficient, in K; is the mole fraction of CO2 in natural gas, unitless; the corrected critical temperature of natural gas is T pc ′, unit is K.

[0156] (3) Calculation of the quasi-comparative pressure p pr , simulated comparison temperature T pr , quasi-contrast density ρ pr Please refer to the following formulas 5 to 7:

[0157]

[0158]

[0159]

[0160] (4) To calculate the deviation factor Z of natural gas containing CO2, please refer to the following formula 8:

[0161]

[0162] (5) For the effective molecular weight M0 of the ground degassed crude oil, if the relative density of crude oil γ0 ≤ 0.8251, please refer to the following formula 9 to calculate M0:

[0163] M0=630-10D Formula 9,

[0164] Where D is the API gravity of crude oil and has no unit.

[0165] If the relative density of crude oil γ0>0.8251, please refer to the following formula 10 to calculate M0:

[0166] M0=73110D -1.562 Formula 10.

[0167] (6) Calculate the mole fraction of natural gas. Please refer to the following formula 11 to calculate the mole fraction of natural gas:

[0168]

[0169] Among them, y g is the mole fraction of natural gas, P is the pressure, and θ is the temperature parameter.

[0170] like Please refer to the following formula 12 to calculate the mole fraction of natural gas:

[0171]

[0172] (7) Please refer to the following formula 13 to calculate the dissolved oil-gas ratio of natural gas:

[0173]

[0174] Step 330, calculating the oil well pump degree.

[0175] The pump filling degree β indicates the degree to which the pump is filled with liquid during operation. The lower the pump filling degree β, the greater the degree to which the oil well pump is affected by gas.

[0176] Next, the calculation of the pump fullness is specifically introduced, and the calculation is carried out in combination with the above formulas 1 to 13, which specifically includes the following steps (1) to (2).

[0177] (1) When calculating the pump filling degree, you first need to calculate the actual gas-liquid ratio R of the fluid entering the pump. Please refer to the following formula 14:

[0178]

[0179] Among them, R p is the gas-liquid ratio of oil well production, in m 3 / m 3 ; R s is the dissolved gas-oil ratio, in m 3 / m 3 ;f w is the moisture content, no unit; T inis the temperature inside the pump, in °C; T0 is the standard temperature, in °C; P in is the pressure inside the pump, in MPa; P0 is the standard pressure, in MPa; Z is the deviation factor of natural gas containing CO2, without unit.

[0180] (2) To calculate the pump filling degree based on R, please refer to the following formula 15:

[0181]

[0182] Wherein, K is the clearance ratio of the oil well pump and has no unit.

[0183] For illustration, please refer to Figure 4 , Figure 4 is a schematic diagram of the pump filling degree provided by an exemplary embodiment of the present application, such as Figure 4 As shown, the oil pump includes an upper dead center 410 and a lower dead center 420, wherein the upper dead center 410 is the highest point that the piston or plunger of the oil pump can reach during the upward stroke, and the lower dead center 420 is the lowest point that the piston or plunger can reach during the downward stroke. p Refers to the volume discharged by the pump; V g Refers to the gas in the pump; V l ' refers to the fluid in the pump that can be discharged; V l Refers to the fluid in the pump; V s Refers to the liquid within the anti-surge distance. If the gas volume in the pump (V g ) is larger, it will reduce the amount of fluid that can be discharged from the pump (V i ) and the liquid within the impact distance (V s ), thus reducing pump efficiency. Reduced pump filling may result in pump stroke loss, affecting the production efficiency of the oil well. Therefore, optimizing the design and operating parameters of the pump to reduce gas entering the pump can improve the pump filling degree and overall production efficiency.

[0184] Step 340 , predict the pump fullness based on the prediction model.

[0185] In some embodiments, based on the Attention-LSTM pump fullness prediction, an LSTM model with an attention mechanism added is constructed, and the collected CCUS well production parameters and pump fullness calculation results are used as inputs of the model to train the pump fullness prediction model.

[0186] Specifically, the Attention mechanism is used to improve the LSTM model. The oil well production parameters are input into the hidden layer of the LSTM network for feature extraction. The Attention mechanism is used to calculate the weight of the feature vector. The feature weight is then multiplied by the feature vector extracted from the hidden layer and input into the Dropout layer and the fully connected layer to obtain the model output.

[0187] The oil well production data collected in step 310 is used as training data, and the pump fullness calculated in step 330 is used as prediction data, and a training set and a test set are divided.

[0188] During the training of the Attention-LSTM model, the parameter x is generated each time. i As the input of the model, the predicted result of the pump filling coefficient is output; the error of the model output value is compared with the actual value, the different output values ​​are derived through the loss function, and the error is transmitted to all neurons using the back propagation algorithm.

[0189] Use the stochastic gradient descent method to update the weights and bias values ​​of the Attention-LSTM model, and calculate the adjustment amplitude of each model parameter based on the learning rate.

[0190] Repeat the above training steps until the set number of training rounds is reached, complete the training, save the parameters of the Attention-LSTM network, and use the Attention-LSTM network model with trained parameters to predict and analyze the fullness of the oil pump.

[0191] For illustration, please refer to Figure 5 , Figure 5 is a schematic diagram of a long short-term memory structure provided by an exemplary embodiment of the present application, such as Figure 5 As shown, the input information x t Input LSTM model 510, LSTM will maintain cell memory based on state vector and hidden state vector, where c t-1 、c t is the state vector, which controls the state or memory of the entire LSTM unit. It will be updated according to the input at each moment, thereby maintaining the memory of the LSTM unit in real time; t-1 、h t is the hidden state vector, which is the output state of the current LSTM unit. It is the actual working state vector, that is, h is generally used. t To do some specific tasks; x t is the input information; tanh is the activation function.

[0192] For illustration, please refer to Figure 6 , Figure 6 is a schematic diagram of an attention mechanism provided by an exemplary embodiment of the present application, such as Figure 6As shown, the prediction model includes a feature encoder 610, which is used to extract the features of the model input to obtain multiple feature vectors, and the attention module 620 combines the multiple feature vector outputs to obtain the final prediction result 630. Among them, y is the final output; y1, y2, y3, y4 are output features; h0, h1, h2, h3, h4 are feature encoders; x1, x2, x3, x4 are input features.

[0193] For illustration, please refer to Figure 7 , Figure 7 is a schematic diagram of the Attention-LSTM neural network structure provided by an exemplary embodiment of the present application, such as Figure 7 As shown, the oil well parameters are input into the LSTM according to the time step, and the attention mechanism layer 710 calculates the weight of the feature vector based on the feature vector of the long-term input of the LSTM, and then multiplies the feature weight and the feature vector extracted by the hidden layer and inputs the regularization layer 720 (Dropout) and the fully connected layer 730 to obtain the model output 740.

[0194] Step 350: Intelligently warn of CCUS production well gas lock conditions.

[0195] Illustratively, an oil well in which the pump filling degree indicated by the first prediction result is less than or equal to 10% is recorded as a "gas lock" state, which is used to indicate that the oil well has experienced gas lock.

[0196] The oil wells whose pump filling degree indicated by the first prediction result is between 10% and 30% are recorded as "severe gas impact" status, which is used to indicate that the oil wells are currently severely affected by gas.

[0197] The oil wells whose pump fullness indicated by the first prediction result is between 30% and 60% are recorded as "gas-affected" status, which is used to indicate that the current oil well is moderately affected by gas.

[0198] The oil wells whose pump filling degree indicated by the first prediction result is between 60% and 80% are recorded as "slightly gas-affected" status, which is used to indicate that the oil wells are currently slightly affected by gas.

[0199] The oil wells whose pump filling degree reaches 80% indicated by the first prediction result are recorded as being in a "almost no gas impact" state, indicating that the oil wells are currently affected by gas to a negligible degree.

[0200] Compared with the traditional abnormal working condition classification method, which only divides the impact of gas on the pump into "gas impact" and "gas lock" working conditions based on whether the valve ball can be opened normally, the method provided in the embodiment of the present application further refines it into 5 types according to the pump fullness. As the pump fullness decreases and the gas impact level gradually increases, it helps oilfield developers to promptly discover seriously gas-affected wells and warn of the gas lock state of the oil pump in advance.

[0201] To sum up, the method provided in the embodiment of the present application is that the CCUS production well has the characteristics of high gas-liquid ratio and high CO2 content. When the CO2 content in the natural gas is high, the use of the conventional natural gas deviation factor calculation method will result in a large error. The present application adopts the Wichert-Aziz-Cranmerf method to correct the deviation factor of the natural gas containing CO2, which can more accurately calculate the degree of fullness in the pump.

[0202] The method provided in the embodiment of the present application uses the Attention mechanism to optimize LSTM, which can better capture long-term dependencies and important information in time series data and improve the diagnostic accuracy of the model.

[0203] Figure 8 is a structural block diagram of a prediction device for gas lock in a production well provided by an exemplary embodiment of the present application, such as Figure 8 As shown, the device includes the following parts:

[0204] An acquisition module 810 is used to acquire a first production parameter corresponding to a first production well, wherein the first production well includes a first liquid to be acquired and a first gas generated during the acquisition process, and the first production parameter is used to indicate a production state of the first production well during the acquisition of the first liquid by a liquid pump, and the liquid pump is an acquisition device for acquiring the first liquid;

[0205] Processing module 820, used for predicting the filling degree of the pump corresponding to the first production well based on the first production parameter by using a first prediction model to obtain a first prediction result, wherein the first prediction result is used to indicate the filling degree of the first liquid in the liquid pump, and the filling degree of the first liquid in the liquid pump is used to characterize the degree to which the liquid pump is affected by the first gas, wherein the first prediction model is a pre-trained machine learning model for predicting the degree to which the acquisition equipment is affected by the gas;

[0206] The processing module 820 is further used to determine a gas lock state corresponding to the first production well based on the first prediction result, and the gas lock state is used to indicate the degree to which the production performance of the first production well is affected by the first gas.

[0207] In some embodiments, the processing module 820 is further configured to determine that gas lock occurs in the first production well when the fullness level indicated by the first prediction result reaches a first preset range.

[0208] In some embodiments, the processing module 820 is further configured to dynamically determine the first preset degree range based on the production status of the first production well.

[0209] In some embodiments, the acquisition module 810 is further used to acquire sample production parameters, where the sample production parameters are production parameters collected within a historical time period for the first production well;

[0210] The processing module 820 is further configured to calculate a reference fullness based on the sample production parameter, wherein the reference fullness is used to indicate the degree to which the liquid pump is affected by the first gas under the production state indicated by the sample production parameter;

[0211] The processing module 820 is further configured to predict the pump filling degree corresponding to the first production well based on the sample production parameters by using a candidate prediction model to obtain a sample prediction result;

[0212] The processing module 820 is further configured to determine a predicted loss value based on a difference between the sample prediction result and the reference fullness;

[0213] The processing module 820 is further used to train the candidate prediction model based on the prediction loss value to obtain the first prediction model.

[0214] In some embodiments, the processing module 820 is further configured to:

[0215] calculating a deviation factor and a dissolved gasoline ratio of the first gas based on the sample production parameters;

[0216] The reference filling level is calculated based on the deviation factor, the dissolved gasoline ratio, and the sample production parameters.

[0217] In some embodiments, the first production well comprises a carbon capture utilization and storage CCUS production well, and the first gas comprises natural gas containing carbon dioxide CO2;

[0218] The first production parameters include at least:

[0219] the oil production of the first production well;

[0220] the water content of the first production well;

[0221] the gas production of the first production well;

[0222] The molar content of CO2 in the first production well;

[0223] Dynamic liquid level parameters of the first production well;

[0224] a pump stroke of the liquid pump;

[0225] pump strokes of the liquid pump;

[0226] The pumping depth of the liquid pump.

[0227] To summarize, the device provided in the embodiment of the present application characterizes the production status of the production well by collecting production parameters, and uses a pre-trained first prediction model to predict the pump fullness based on the production parameters, thereby determining the gas lock state according to the pump fullness, so that a machine learning model capable of predicting the degree to which the collection equipment is affected by gas is involved in the prediction of the gas lock state, thereby freeing up human analysis costs and improving the automaticity of gas lock prediction. In addition, since the model operation takes a short time, the timeliness of the gas lock prediction can be guaranteed, thereby improving the efficiency of the gas lock prediction.

[0228] It should be noted that the prediction device for gas lock in production wells provided in the above embodiment is only illustrated by the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the equipment can be divided into different functional modules to complete all or part of the functions described above.

[0229] Fig. 9 The structure block diagram of a terminal 900 provided by an exemplary embodiment of the present application is shown. The terminal 900 may be: a smart phone, a tablet computer, an MP3 player, an MP4 player, a notebook computer or a desktop computer. The terminal 900 may also be called a user equipment, a portable terminal, a laptop terminal, a desktop terminal or other names.

[0230] Typically, the terminal 900 includes a processor 901 and a memory 902 .

[0231] The processor 901 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 901 may be implemented in at least one hardware form of digital signal processing (DSP), field-programmable gate array (FPGA), and programmable logic array (PLA). The processor 901 may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a central processing unit (CPU); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor 901 may be integrated with a graphics processing unit (GPU), which is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 901 may also include an artificial intelligence (AI) processor, which is used to process computing operations related to machine learning.

[0232] The memory 902 may include one or more computer-readable storage media, which may be non-transitory. The memory 902 may also include a high-speed random access memory, and a non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 902 is used to store at least one instruction, which is used to be executed by the processor 901 to implement the prediction method for gas lock in the production well provided by the method embodiment of the present application.

[0233] In some embodiments, the terminal 900 further includes some other components 903, and the type and quantity of the other components 903 can be selected based on the functional requirements of the terminal 900. Those skilled in the art will appreciate that Fig. 9 The structure shown in the figure does not constitute a limitation on the terminal 900, and the terminal 900 may include more or less components than those shown in the figure, or combine some components, or adopt a different component arrangement.

[0234] The embodiment of the present application also provides a computer device, which can be implemented as follows: Figure 1The computer device includes a processor and a memory, wherein at least one instruction, at least one program, code set or instruction set is stored in the memory, and the at least one instruction, at least one program, code set or instruction set is loaded and executed by the processor to implement the prediction method for gas lock in the production well provided by the above-mentioned method embodiments.

[0235] An embodiment of the present application also provides a computer-readable storage medium, on which is stored at least one instruction, at least one program, code set or instruction set, and the at least one instruction, at least one program, code set or instruction set is loaded and executed by a processor to implement the prediction method for gas lock in production wells provided in the above-mentioned method embodiments.

[0236] The embodiments of the present application also provide a computer program product or a computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the prediction method for gas lock in a production well provided by the above-mentioned various method embodiments.

[0237] Optionally, the computer readable storage medium may include: Read Only Memory (ROM), Random Access Memory (RAM), Solid State Drives (SSD) or optical disks, etc. Among them, the random access memory may include resistance random access memory (ReRAM) and dynamic random access memory (DRAM). The serial numbers of the above embodiments of the present application are only for description and do not represent the advantages and disadvantages of the embodiments.

[0238] A person skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware or by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.

[0239] The above description is only an optional embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for predicting gas lock in production wells, It is characterized in that The method comprises: Acquire a first production parameter corresponding to a first production well, wherein the first production well includes a first liquid to be collected and a first gas generated during the collection process, and the first production parameter is used to indicate a production state of the first production well during the process of collecting the first liquid based on a liquid pump, and the liquid pump is a collection device for collecting the first liquid; Predicting the filling degree of the pump corresponding to the first production well based on the first production parameter by a first prediction model to obtain a first prediction result, wherein the first prediction result is used to indicate the filling degree of the first liquid in the liquid pump, and the filling degree of the first liquid in the liquid pump is used to characterize the degree to which the liquid pump is affected by the first gas, and the first prediction model is a pre-trained machine learning model for predicting the degree to which the acquisition equipment is affected by the gas; A gas lock state corresponding to the first production well is determined based on the first prediction result, and the gas lock state is used to indicate the extent to which the production performance of the first production well is affected by the first gas.

2. The method according to claim 1, It is characterized in that The determining the gas lock state corresponding to the first production well based on the first prediction result includes: When the fullness indicated by the first prediction result reaches a first preset range, it is determined that gas lock occurs in the first production well.

3. The method according to claim 2, It is characterized in that When the fullness indicated by the first prediction result reaches a first preset range, before determining that the first production well has gas lock, the method further includes: The first preset degree range is dynamically determined based on the production status of the first production well.

4. The method according to any one of claims 1 to 3, It is characterized in that Before obtaining the first prediction result by predicting the pump filling degree corresponding to the first production well based on the first production parameter using the first prediction model, the method further includes: Acquiring sample production parameters, where the sample production parameters are production parameters collected within a historical time period for the first production well; Calculating a reference fullness based on the sample production parameters, the reference fullness being used to indicate the extent to which the liquid pump is affected by the first gas when the liquid pump is in the production state indicated by the sample production parameters; Predicting the pump fullness corresponding to the first production well based on the sample production parameters using a candidate prediction model to obtain a sample prediction result; determining a predicted loss value based on a difference between the sample prediction result and the reference fullness; The candidate prediction model is trained based on the prediction loss value to obtain the first prediction model.

5. The method according to claim 4, It is characterized in that The calculating a reference filling degree based on the sample production parameters comprises: calculating a deviation factor and a dissolved gasoline ratio of the first gas based on the sample production parameters; The reference filling level is calculated based on the deviation factor, the dissolved gasoline ratio, and the sample production parameters.

6. The method according to any one of claims 1 to 3, It is characterized in that The first production well includes a carbon capture utilization and storage CCUS production well, and the first gas includes carbon dioxide CO 2 of natural gas; The first production parameters include at least: the oil production of the first production well; the water content of the first production well; the gas production of the first production well; The first production well contains CO 2 Molar content; Dynamic liquid level parameters of the first production well; a pump stroke of the liquid pump; pump strokes of the liquid pump; The pumping depth of the liquid pump.

7. A prediction device for gas lock in production wells, It is characterized in that The device comprises: an acquisition module, configured to acquire a first production parameter corresponding to a first production well, wherein the first production well includes a first liquid to be collected and a first gas generated during the collection process, and the first production parameter is used to indicate a production state of the first production well during the process of collecting the first liquid based on a liquid pump, and the liquid pump is a collection device for collecting the first liquid; a processing module, configured to predict the filling degree of the pump corresponding to the first production well based on the first production parameter by using a first prediction model to obtain a first prediction result, wherein the first prediction result is used to indicate the filling degree of the first liquid in the liquid pump, and the filling degree of the first liquid in the liquid pump is used to characterize the degree to which the liquid pump is affected by the first gas, wherein the first prediction model is a pre-trained machine learning model for predicting the degree to which a collection device is affected by gas; The processing module is further used to determine a gas lock state corresponding to the first production well based on the first prediction result, and the gas lock state is used to indicate the degree to which the production performance of the first production well is affected by the first gas.

8. A computer device, It is characterized in that The computer device includes a processor and a memory, wherein the memory stores at least one computer program, and the at least one computer program is loaded and executed by the processor to implement the method for predicting gas lock in a production well as described in any one of claims 1 to 6.

9. A computer-readable storage medium, It is characterized in that The storage medium stores at least one computer program, which is loaded and executed by a processor to implement the method for predicting gas lock in a production well as described in any one of claims 1 to 6.

10. A computer program product, It is characterized in that It comprises a computer program, which, when executed by a processor, implements the method for predicting gas lock in a production well as claimed in any one of claims 1 to 6.

Images