Method and system for simulating oil well failure and workover process of workover rig

Abstract

The invention relates to the technical field of oil exploitation, in particular to an oil well failure and workover rig workover process simulation method and system, and the method comprises the steps: carrying out the collaborative association mining of static parameter data and dynamic production historical data of a target oil well, and obtaining multi-source associated data; based on the multi-source associated data, knowledge-guided symptom identification is carried out on the target oil well, and a potential failure symptom data combination is obtained; performing feature clustering coding on the potential failure symptom data combination to obtain an oil well failure mode feature library; based on the oil well failure mode feature library, performing matching research and judgment on the current working condition data to obtain a failure mode record; based on the failure mode record, quantitative adaptation evaluation and strategy construction are carried out on the failure risk, and a simulated well repair scheme is obtained; performing logic deduction on the key operation node data based on the well repair simulation scheme to obtain a simulation effect; the efficiency of the oil well failure and workover rig workover process simulation method can be improved.

Description

Technical Field

[0001] This invention relates to the field of oil extraction technology, and in particular to a method and system for simulating oil well failure and well workover process. Background Technology

[0002] Existing technologies for simulating oil well failures and well workover processes lack in-depth integration and analysis of static parameters and dynamic production data, making it difficult to fully capture the inherent correlations between multi-source data. This results in a lack of accuracy and comprehensiveness in identifying potential failure signs. Furthermore, the construction of failure modes lacks systematic feature clustering and knowledge association support, leading to insufficient reliability in condition matching and judgment, and failing to provide accurate basis for well workover plan formulation.

[0003] Existing technologies lack a dynamic and adaptable risk quantification assessment mechanism in the well workover plan generation stage. Strategy development often fails to meet the actual operating conditions of oil wells, and the simulation of key operational nodes during the workover process is not detailed enough, making it difficult to predict the operation's effectiveness and resource consumption in advance. Therefore, improving the accuracy of oil well failure identification, the adaptability of workover plans, and the effectiveness of the simulation process have become urgent problems to be solved. Summary of the Invention

[0004] This invention provides a method and system for simulating oil well failure and well workover process, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a method for simulating oil well failure and well workover process, comprising: S1. Collaborative correlation mining is performed on the static parameter data and dynamic production history data of the target oil well to obtain multi-source correlation data of the target oil well; S2. Based on multi-source correlation data, knowledge-guided symptom identification is performed on the target oil well to obtain a combination of potential failure symptom data for the target oil well. S3. Perform feature clustering encoding on the combination of potential failure symptom data to obtain the oil well failure mode feature library of the target oil well; S4. Based on the oil well failure mode feature library, match and analyze the current operating data of the target oil well to obtain the failure mode record of the target oil well. S5. Based on the failure mode record, the failure risk of the target oil well is quantitatively adapted and evaluated to obtain the quantitative characterization of the target oil well. Based on the quantitative characterization, a strategy is constructed for the target oil well to obtain a simulated well workover plan for the target oil well. S6. Based on the simulated well workover scheme, logical deduction is performed on the key operation node data of the target oil well to obtain the simulation effect of the target oil well.

[0006] In a preferred embodiment, the step of collaboratively correlated mining of the static parameter data and dynamic production history data of the target oil well to obtain multi-source correlated data of the target oil well includes: Acquire static parameter data and dynamic production history data of the target oil well. Static parameter data includes well completion structure data and formation physical property data, while dynamic production history data includes production data, pressure data, and well workover operation data. By performing time-series alignment between static parameter data and dynamic production history data, the spatiotemporal aligned data of the target oil well is obtained; Feature coupling mining is performed on spatiotemporally aligned data to obtain multi-source correlation data of the target oil well.

[0007] In a preferred embodiment, the step of identifying potential failure symptom data of the target oil well based on multi-source correlation data to obtain a combination of potential failure symptom data of the target oil well includes: Knowledge condensation is performed on historical failure cases of the target oil well to obtain the knowledge characteristics of the target oil well; Based on prior knowledge features, the key parameters in multi-source correlated data are mapped to their importance to obtain feature analysis guidance for the target oil well; Based on feature analysis guidance, anomaly detection is performed on the time-series evolution process in multi-source correlated data to obtain the primary anomaly pattern of the target oil well. The intensity of the primary anomaly patterns is calibrated to obtain a combination of potential failure symptom data for the target oil well.

[0008] In a preferred embodiment, the step of mapping the importance of key parameters in multi-source correlated data based on knowledge features to obtain feature analysis guidance for the target oil well includes: Logical deconstruction of knowledge features yields a dynamic parameter constraint framework for the target oil well; Based on a dynamic parameter constraint framework, the correlation of multi-source associated data is measured to obtain the parameter correlation degree of multi-source associated data. The variance contribution of the parameter correlation is evaluated to obtain the contribution weights of the multi-source correlation data; Core nodes are identified to determine the correlation degree of parameters, thereby obtaining the dominant parameters of multi-source correlated data. Based on the dominant parameters and contribution weights, online attention allocation is performed on multi-source correlated data to obtain characteristic analysis guidance for target oil wells.

[0009] In a preferred embodiment, the step of performing feature clustering encoding on the combination of potential failure symptom data to obtain a well failure mode feature library for the target oil well includes: The potential failure symptom data combination is vectorized to obtain the standardized symptom vector of the target oil well; The dimensionality reduction mapping of the standardized symptom vector yields the dimensionality reduction feature representation of the target oil well; Based on the dimensionality reduction feature representation, the symptom similarity of the standardized symptom vectors is calculated. The formula for calculating the symptom similarity is as follows: ; in, For the first The and the first The similarity value of each symptom For the first The t-th eigenvalue of a standardized symptom vector For the first The t-th eigenvalue of a standardized symptom vector The number of feature dimensions for the standardized symptom vector; Based on symptom similarity, spatial clustering is performed on the dimensionality-reduced feature representation to obtain a preliminary cluster division of the target oil wells; Pattern induction was performed on the preliminary cluster classification to obtain the failure modes of the target oil well; Abstract failure modes are encapsulated through knowledge association to obtain a feature library of oil well failure modes for the target oil well.

[0010] In a preferred embodiment, the step of matching and analyzing the current operating data of the target oil well based on the oil well failure mode feature library to obtain the failure mode record of the target oil well includes: Real-time monitoring of the current operating data of the target oil well; Based on the oil well failure mode feature library, pattern matching is performed on the current operating data to obtain candidate failure modes of the target oil well. Contextual semantic fusion is performed on candidate failure modes to obtain the comprehensive failure mode of the target oil well; The integrity of the comprehensive failure modes is verified to obtain the failure mode record of the target oil well.

[0011] In a preferred embodiment, the step of quantitatively adapting and assessing the failure risk of the target oil well based on failure mode records to obtain a quantitative characterization of the target oil well, and then constructing a strategy for the target oil well based on the quantitative characterization to obtain a simulated well workover plan for the target oil well, includes: Key indicators are extracted from failure mode records to obtain the risk characteristics of the target oil well; By dynamically assigning weights to risk characteristics, preliminary risk indicators for the target oil well are obtained. The initial risk indicators are adjusted for suitability to obtain a quantitative risk characterization of the target oil well. Based on risk quantification, the historical case database of the target oil well is searched and adapted to obtain candidate strategies for the target oil well. The candidate strategies are dynamically programmed to obtain a simulated well workover plan for the target oil well.

[0012] In a preferred embodiment, the step of dynamically weighting the risk characteristics to obtain preliminary risk indicators for the target oil well includes: A dual assessment of risk characteristics yields a basic score for those characteristics. Based on the current operating data of the target oil well, the basic score is corrected for deviation to obtain the corrected score of risk characteristics; Cross-dimensional correlation analysis was performed on the corrected scores to obtain the interaction coefficients between risk characteristics; Based on the interaction influence coefficient, weights are assigned to the current working condition data to obtain the scenario-adaptive weights of the risk characteristics; Based on the revised score and scenario adaptation weights, the risk characteristics are comprehensively quantified to obtain the preliminary risk indicators of the target oil well.

[0013] In a preferred embodiment, the step of logically deducing the key operational node data of the target oil well based on the simulated well workover scheme to obtain the simulation effect of the target oil well includes: Key nodes of the simulated well workover scheme are extracted to obtain the discretized operation nodes of the target oil well; Path planning is performed on the discretized operation nodes to obtain the initial operation path of the target oil well; Resource consumption simulations were performed on the initial proposed path to obtain feasible operational paths for the target oil well; Based on feasible operation paths, the execution status of discrete operation nodes is reproduced by time-series simulation to obtain the simulation effect of the target oil well.

[0014] To address the aforementioned problems, the present invention also provides a simulation system for oil well failure and well workover rig processes, the system comprising: The multi-source data association mining module is used to collaboratively associate and mine the static parameter data and dynamic production history data of the target oil well to obtain the multi-source association data of the target oil well. The knowledge-guided symptom identification module is used to identify the symptom of a target oil well based on multi-source correlation data, and obtain a combination of potential failure symptom data of the target oil well. The failure feature encoding module is used to perform feature clustering encoding on the combination of potential failure symptom data to obtain the oil well failure mode feature library of the target oil well; The matching and analysis module is used to match and analyze the current operating data of the target oil well based on the oil well failure mode feature library to obtain the failure mode record of the target oil well. The scheme construction module is used to quantitatively adapt and assess the failure risk of the target oil well based on the failure mode record, obtain the quantitative characterization of the target oil well, and construct a strategy for the target oil well based on the quantitative characterization to obtain the simulated well workover scheme for the target oil well. The simulation effect generation module is used to logically deduce the key operational node data of the target oil well based on the simulated well workover scheme, and obtain the simulation effect of the target oil well.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention, through a method and system for simulating oil well failure and workover processes, utilizes the collaborative correlation mining of static parameter data and dynamic production history data. Combined with knowledge-guided symptom identification and feature clustering coding, it accurately constructs an oil well failure mode feature library, significantly improving the accuracy and efficiency of failure mode matching and judgment. This enables early detection and precise location of potential oil well failure risks. Based on quantitative adaptation assessment and dynamic strategy construction using failure mode records, the simulated workover scheme more closely matches actual oil well conditions, effectively ensuring the scheme's relevance and feasibility.

[0016] 2. By logically deducing and temporally reproducing the key operational nodes of the simulated well workover scheme, the simulation effect of the well workover process is fully presented, providing detailed reference for well workover operations and reducing blind operations in actual well workover processes. The entire technical process realizes intelligent processing across the entire chain from data mining and failure identification to scheme construction and effect simulation, significantly improving the overall efficiency of oil well failure prevention and well workover operations, and reducing operating costs and risks. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating a method for simulating oil well failure and well workover process according to an embodiment of the present invention. Figure 2 This is a functional block diagram of an oil well failure and well workover process simulation system provided in an embodiment of the present invention; The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0018] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0019] This application provides a method for simulating oil well failure and well workover process. The execution subject of this method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the method for simulating oil well failure and well workover process can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0020] Reference Figure 1 The diagram shown is a flowchart illustrating a method for simulating oil well failure and well workover process according to an embodiment of the present invention. In this embodiment, the method for simulating oil well failure and well workover process includes: S1. Collaborative correlation mining is performed on the static parameter data and dynamic production history data of the target oil well to obtain multi-source correlation data of the target oil well; In this embodiment of the invention, the step of collaboratively correlated mining of the static parameter data and dynamic production history data of the target oil well to obtain multi-source correlated data of the target oil well includes: Acquire static parameter data and dynamic production history data of the target oil well. Static parameter data includes well completion structure data and formation physical property data, while dynamic production history data includes production data, pressure data, and well workover operation data. By performing time-series alignment between static parameter data and dynamic production history data, the spatiotemporal aligned data of the target oil well is obtained; Feature coupling mining is performed on spatiotemporally aligned data to obtain multi-source correlation data of the target oil well.

[0021] The system collects static parameter data and dynamic production history data for the target oil well. The static parameter data specifically collects well completion structure data and formation physical property data. Well completion structure data includes information reflecting the fixed structure after the oil well is built, such as casing specifications, cementing quality, and wellhead equipment type. Formation physical property data includes physical properties of the formation itself, such as formation porosity, permeability, and oil saturation. The dynamic production history data focuses on collecting production data, pressure data, and well workover operation data. Production data records the crude oil production of the oil well at different time periods. Pressure data includes pressure information such as bottom hole pressure and wellhead pressure that change over time during production. Well workover operation data records in detail the operation time, operation content, and problems handled for each past well workover.

[0022] Using time as a unified benchmark, the collected static parameter data and dynamic production history data are time-series aligned. Static parameter data, which does not change over time, is set as fixed data that remains stable at all time points. Dynamic production history data is sorted and organized according to its recorded timestamps, so that the static parameter data and dynamic production history data correspond one-to-one in the same time dimension, ultimately forming spatiotemporal aligned data for the target oil well, ensuring that the two types of data can be matched and correlated in the time dimension in subsequent analysis.

[0023] Feature coupling mining is carried out on the obtained spatiotemporal aligned data. First, the features in the static parameter data and the features in the dynamic production history data are sorted out separately. The intrinsic correlation between different features is calculated by Pearson correlation coefficient. For example, the correlation between casing specifications and production and pressure data in oil well completion structure, and the correlation between permeability and production changes and well workover frequency in formation physical parameters. By systematically sorting out the mutual influence and interaction between different features, the originally scattered static and dynamic data features are organically combined, and finally multi-source correlation data that can comprehensively reflect the multi-faceted attribute correlation of the target oil well is obtained.

[0024] The beneficial effects are that by clarifying the specific content of static and dynamic data collection, the comprehensiveness and relevance of data acquisition are ensured; the time-series alignment operation enables different types of data to form an effective correspondence in the time dimension; and the feature coupling mining deeply explores the intrinsic relationship between data. The resulting multi-source correlation data can provide comprehensive and accurate basic support for the subsequent identification of potential failure symptoms, and greatly improve the reliability and accuracy of subsequent analysis.

[0025] S2. Based on multi-source correlation data, knowledge-guided symptom identification is performed on the target oil well to obtain a combination of potential failure symptom data for the target oil well. In this embodiment of the invention, the step of identifying potential failure symptom data of the target oil well based on multi-source correlation data to obtain a combination of potential failure symptom data of the target oil well includes: Knowledge condensation is performed on historical failure cases of the target oil well to obtain the knowledge characteristics of the target oil well; Based on prior knowledge features, the key parameters in multi-source correlated data are mapped to their importance to obtain feature analysis guidance for the target oil well; Based on feature analysis guidance, anomaly detection is performed on the time-series evolution process in multi-source correlated data to obtain the primary anomaly pattern of the target oil well. The intensity of the primary anomaly patterns is calibrated to obtain a combination of potential failure symptom data for the target oil well.

[0026] The method of mapping the importance of key parameters in multi-source correlated data based on knowledge features to obtain feature analysis guidance for target oil wells includes: Logical deconstruction of knowledge features yields a dynamic parameter constraint framework for the target oil well; Based on a dynamic parameter constraint framework, the correlation of multi-source associated data is measured to obtain the parameter correlation degree of multi-source associated data. The variance contribution of the parameter correlation is evaluated to obtain the contribution weights of the multi-source correlation data; Core nodes are identified to determine the correlation degree of parameters, thereby obtaining the dominant parameters of multi-source correlated data. Based on the dominant parameters and contribution weights, online attention allocation is performed on multi-source correlated data to obtain characteristic analysis guidance for target oil wells.

[0027] By reviewing all historical failure cases of the target oil well, key information such as the cause of failure, manifestation, influencing factors, and handling methods in each case is comprehensively extracted. This information is then classified, organized, and refined in depth, removing duplicate and redundant content, and summarizing core information with commonalities and regularities to form knowledge characteristics that reflect the key attributes related to the failure of the target oil well.

[0028] The extracted knowledge features are decomposed and analyzed to clarify the types of parameters involved, the logical relationships between parameters, and the reasonable range of parameter values. These elements are then systematically integrated to construct a dynamic parameter constraint framework that can constrain and guide multi-source correlated data, ensuring that subsequent parameter analysis has clear basis and standards.

[0029] Based on the dynamic parameter constraint framework, the interrelationships between various parameters in multi-source correlated data are analyzed one by one. By comparing the synchronicity and correlation of different parameters in the process of change, the degree of mutual influence between various parameters is determined, and thus the parameter correlation degree of multi-source correlated data is obtained.

[0030] Based on the obtained parameter correlation, we analyze the role of each parameter in the overall data variation, calculate the contribution ratio of the variance of each parameter to the overall data variance, and determine the importance of each parameter according to the contribution ratio, thereby obtaining the contribution weight of multi-source correlated data.

[0031] Based on the parameter correlation analysis, parameters that have the widest and deepest impact on other parameters and occupy a core position in the data correlation network are selected. These parameters are identified as the dominant parameters of multi-source correlation data, clarifying the core focus of subsequent analysis.

[0032] By combining the determined dominant parameters and the contribution weights of each parameter, all parameters in the multi-source correlation data are prioritized. Higher attention priority is assigned to the dominant parameters and parameters with large contribution weights, while relatively lower attention priority is assigned to parameters with small contribution weights. This forms an online attention allocation scheme for multi-source correlation data, thus providing guidance for the characteristic analysis of the target oil well.

[0033] Based on the parameter priority specified in the feature analysis guidelines, we focus on tracking the temporal changes of high-priority parameters in multi-source correlated data, comparing the differences and trends of parameter values ​​in different time periods, and comparing them with the parameter change patterns under normal production conditions to identify parameter changes that deviate from the normal pattern, thereby obtaining the primary anomaly pattern of the target oil well.

[0034] A detailed analysis of the identified primary anomaly patterns was conducted. Based on factors such as the degree of deviation of the anomaly parameters, the scope of influence, and the duration, a unified intensity calibration standard was established. According to this standard, the failure symptoms corresponding to each primary anomaly pattern were classified into intensity levels. The symptom data of different intensity levels were integrated to finally obtain the potential failure symptom data combination of the target oil well.

[0035] The beneficial effects are that by extracting targeted knowledge features from historical failure cases, and then forming precise feature analysis guidelines through a series of steps such as logical deconstruction and correlation measurement, anomaly detection and symptom intensity calibration can be carried out based on these guidelines. This ensures that the acquisition process of potential failure symptom data combinations is systematic, accurate and efficient, and can comprehensively and accurately capture the potential failure risks of target oil wells, laying a solid foundation for subsequent failure mode analysis and well workover plan formulation.

[0036] S3. Perform feature clustering encoding on the combination of potential failure symptom data to obtain the oil well failure mode feature library of the target oil well; In this embodiment of the invention, the step of performing feature clustering encoding on the combination of potential failure symptom data to obtain a well failure mode feature library for the target oil well includes: The potential failure symptom data combination is vectorized to obtain the standardized symptom vector of the target oil well; The dimensionality reduction mapping of the standardized symptom vector yields the dimensionality reduction feature representation of the target oil well; Based on the dimensionality reduction feature representation, the symptom similarity of the standardized symptom vectors is calculated. The formula for calculating the symptom similarity is as follows: ; in, For the first The and the first The similarity value of each symptom For the first The t-th eigenvalue of a standardized symptom vector For the first The t-th eigenvalue of a standardized symptom vector The number of feature dimensions for the standardized symptom vector; Based on symptom similarity, spatial clustering is performed on the dimensionality-reduced feature representation to obtain a preliminary cluster division of the target oil wells; Pattern induction was performed on the preliminary cluster classification to obtain the failure modes of the target oil well; Abstract failure modes are encapsulated through knowledge association to obtain a feature library of oil well failure modes for the target oil well.

[0037] For the obtained combination of potential failure symptom data, each failure symptom is converted into a numerical form according to a unified standard to ensure that the symptom data of different types and magnitudes are consistent and comparable. Then, these values ​​are arranged in a preset order to form a vector, and finally the standardized symptom vector of the target oil well is obtained.

[0038] The standardized symptom vectors are analyzed, and redundant and repetitive feature dimensions are removed. Key features that can reflect the core attributes of failure symptoms are retained. By simplifying the number of vector dimensions, the complexity of data processing is reduced, while the core information of the data is preserved, thus obtaining a dimensionality-reduced feature representation of the target oil well.

[0039] Based on the dimensionality reduction feature representation, the numerical differences of any two standardized symptom vectors in each core feature dimension are compared. The similarity between the two vectors is measured by calculating the degree of fit between them in space. Specifically, by comparing the degree of fit of the corresponding dimension values, the degree of correlation between the two standardized symptom vectors is determined, and thus the symptom similarity of the standardized symptom vectors is obtained.

[0040] No. The first standardized symptom vector The eigenvalues ​​are derived from the standardized symptom vector obtained by vectorizing the combination of potential failure symptom data of the target oil well. The first standardized symptom vector The feature values ​​also originate from this standardized symptom vector. The number of feature dimensions in the standardized symptom vector is the total number of feature dimensions determined during the vectorization transformation of the potential failure symptom data combination.

[0041] Calculate the first The and the first The similarity between the standardized symptom vectors is determined by summing the products of the corresponding eigenvalues ​​of the two standardized symptom vectors and dividing by the first standardized symptom vector. The square root of the sum of squares of the eigenvalues ​​of each dimension of the standardized symptom vector and the... The product of the square roots of the sum of the squares of the eigenvalues ​​of each dimension of the standardized symptom vector is used to perform spatial clustering on the dimensionality-reduced feature representation, providing a basis for constructing an oil well failure mode feature library.

[0042] When the eigenvalues ​​of the corresponding dimensions of two standardized symptom vectors are closer and their trends are consistent, the sum of their products will be larger. At the same time, the product of the square roots of the sum of the squares of the eigenvalues ​​of each dimension of the two vectors will change relatively smoothly, resulting in a larger similarity value. Conversely, when the eigenvalues ​​of the corresponding dimensions of the two vectors are more different or their trends are contradictory, the similarity value will be smaller.

[0043] Based on the calculated similarity of the symptoms, the dimensionality reduction feature representations with high similarity are grouped into one category, while the dimensionality reduction feature representations with low similarity are divided into different categories. In this way, all dimensionality reduction feature representations are grouped and classified to form multiple relatively independent feature sets, thus obtaining the preliminary clustering of the target oil well.

[0044] In-depth analysis of each feature set in the preliminary cluster division is conducted to extract the common features and patterns of all dimensionality-reduced feature representations within each set, summarize the typical manifestations that can represent the feature set of that type, clarify the failure types and characteristics corresponding to each category, and thus obtain the failure modes of the target oil well.

[0045] By associating abstract failure modes with previously condensed knowledge features, supplementing each failure mode with historical cases, failure causes, and impact range, and systematically organizing and encapsulating each failure mode, a complete and detailed feature set is formed, ultimately resulting in an oil well failure mode feature library for the target oil well.

[0046] The beneficial effects are that the data structure is simplified by vectorization and dimensionality reduction mapping while retaining core information, and the failure symptoms are scientifically classified based on spatial clustering of symptom similarity. Then, a complete oil well failure mode feature library is formed by pattern induction and knowledge association encapsulation, which provides accurate and systematic basis for matching and judging the current working condition data, and greatly improves the efficiency and accuracy of failure mode identification.

[0047] S4. Based on the oil well failure mode feature library, match and analyze the current operating data of the target oil well to obtain the failure mode record of the target oil well. In this embodiment of the invention, the step of matching and analyzing the current operating data of the target oil well based on the oil well failure mode feature library to obtain the failure mode record of the target oil well includes: Real-time monitoring of the current operating data of the target oil well; Based on the oil well failure mode feature library, pattern matching is performed on the current operating data to obtain candidate failure modes of the target oil well. Contextual semantic fusion is performed on candidate failure modes to obtain the comprehensive failure mode of the target oil well; The integrity of the comprehensive failure modes is verified to obtain the failure mode record of the target oil well.

[0048] Real-time data acquisition equipment is used to continuously monitor the current operating data of the target oil well. The monitoring content covers dynamic changes in production data, pressure data, temperature data, etc. during the production process, ensuring that the operating data at each time point can be completely and accurately captured and recorded, providing real and effective data support for subsequent pattern matching.

[0049] The current operating condition data obtained from real-time monitoring is compared one by one with the established oil well failure mode feature library. Matching analysis is performed according to the degree of consistency between the various features of the operating condition data and the features of different failure modes in the feature library. Several failure modes with high similarity to the features of the current operating condition data are selected, and these selected failure modes are determined as candidate failure modes for the target oil well.

[0050] The contextual information corresponding to each candidate failure mode is analyzed in depth, including the oil well production stage to which the mode applies, the corresponding formation conditions, and the range of variation of relevant parameters. Combined with semantic information such as the current actual production environment and operating status of the target oil well, multiple candidate failure modes are fused to eliminate contradictions and redundant information between different candidate modes and integrate them into a comprehensive failure mode that can fully and accurately reflect the current failure state of the target oil well.

[0051] By comparing the comprehensive failure mode with the complete characteristic system of oil well failure, check whether the comprehensive failure mode covers all possible failure-related information under the current operating conditions, including key contents such as the core characteristics of failure, influencing factors, and potential consequences. Confirm that there are no missing information, logical loopholes, or discrepancies with the actual operating conditions. After a comprehensive verification, the comprehensive failure mode is officially identified as the failure mode record of the target oil well.

[0052] The beneficial effects include ensuring the timeliness and accuracy of current operating data through real-time monitoring, quickly identifying candidate failure modes based on pattern matching of the failure mode feature library, improving the comprehensiveness and relevance of failure modes through contextual semantic fusion, and ensuring the reliability of failure mode records through integrity verification. The entire process is progressive and can accurately and efficiently determine the failure mode of the target oil well, providing an accurate basis for subsequent risk assessment and well workover plan formulation.

[0053] S5. Based on the failure mode record, the failure risk of the target oil well is quantitatively adapted and evaluated to obtain the quantitative characterization of the target oil well. Based on the quantitative characterization, a strategy is constructed for the target oil well to obtain a simulated well workover plan for the target oil well. In this embodiment of the invention, the step of quantitatively adapting and assessing the failure risk of the target oil well based on failure mode records to obtain a quantitative characterization of the target oil well, and constructing a strategy for the target oil well based on the quantitative characterization to obtain a simulated well workover plan for the target oil well, includes: Key indicators are extracted from failure mode records to obtain the risk characteristics of the target oil well; By dynamically assigning weights to risk characteristics, preliminary risk indicators for the target oil well are obtained. The initial risk indicators are adjusted for suitability to obtain a quantitative risk characterization of the target oil well. Based on risk quantification, the historical case database of the target oil well is searched and adapted to obtain candidate strategies for the target oil well. The candidate strategies are dynamically programmed to obtain a simulated well workover plan for the target oil well.

[0054] The dynamic weighting of risk characteristics to obtain preliminary risk indicators for the target oil well includes: A dual assessment of risk characteristics yields a basic score for those characteristics. Based on the current operating data of the target oil well, the basic score is corrected for deviation to obtain the corrected score of risk characteristics; Cross-dimensional correlation analysis was performed on the corrected scores to obtain the interaction coefficients between risk characteristics; Based on the interaction influence coefficient, weights are assigned to the current working condition data to obtain the scenario-adaptive weights of the risk characteristics; Based on the revised score and scenario adaptation weights, the risk characteristics are comprehensively quantified to obtain the preliminary risk indicators of the target oil well.

[0055] By thoroughly analyzing the failure mode records of the target oil well, we can comprehensively extract the core information directly related to failure risk. This information covers key dimensions such as the probability of failure, the degree of production loss that may result, the scope of damage to the oil well equipment, and the technical difficulty required for repair. We organize and summarize this extracted key information into a set of indicators that can clearly reflect the essential attributes of failure risk, and finally obtain the risk characteristics of the target oil well.

[0056] For each extracted risk characteristic, a dual assessment is conducted from two different dimensions. One dimension is the average impact of the risk characteristic in similar oil well failure cases in the past, and the other dimension is the corresponding level of the risk characteristic in the industry-recognized risk level standard. The assessment results of the two dimensions are combined, and a basic score is assigned to each risk characteristic using a unified scoring rule, thereby obtaining the basic score of the risk characteristic.

[0057] Based on the current actual operating data of the target oil well, the basic score of each risk characteristic is analyzed to determine its compatibility with the current operating conditions. If there are factors in the current operating conditions that aggravate or mitigate the impact of the risk characteristic, the basic score is adjusted accordingly based on the specific degree of influence of these factors. For example, if the current oil well production is high, the loss caused by a certain risk characteristic may be greater, so its basic score is appropriately increased; otherwise, it is decreased, and finally, the corrected score of the risk characteristic is obtained.

[0058] Cross-dimensional correlation analysis is performed on the corrected scores of all risk characteristics to systematically study the interaction and mutual influence between different risk characteristics. The extent to which a change in one risk characteristic will affect or drive other risk characteristics is determined. Based on the degree of mutual influence, a specific numerical coefficient is set for each pair of risk characteristics. This coefficient is the interaction coefficient between risk characteristics.

[0059] Based on the obtained interaction coefficients and the actual values ​​of various parameters in the current operating data of the target oil well, a weight is assigned to each risk feature. For risk features with large interaction coefficients and that play a dominant role in the overall failure risk under the current operating conditions, higher weights are assigned, while for risk features with smaller overall risk impacts, lower weights are assigned. This results in a risk feature weight allocation that fits the current actual situation, which is the situation-adaptive weight of the risk features.

[0060] The corrected score of each risk feature is multiplied by the corresponding scenario adaptation weight to obtain the weighted score of each risk feature. The weighted scores of all risk features are then summarized and integrated to obtain a quantitative value that can comprehensively reflect the overall failure risk level of the target oil well. This value is the preliminary risk index of the target oil well.

[0061] Dynamic weight allocation is performed on risk characteristics. The weights are determined using the analytic hierarchy process (AHP). First, a hierarchical structure of risk characteristics is constructed. Then, a judgment matrix is ​​constructed using the 1-9 scaling method. The eigenvectors are calculated and normalized to obtain the weights. Combined with the basic scores of the risk characteristics, the preliminary risk indicators of the target oil well are obtained. The basic score ranges from 0 to 10 points. 8-10 points correspond to a production loss ≥30% or severe equipment damage. 5-7 points correspond to a production loss of 15%-30% or moderate equipment damage. 2-4 points correspond to a production loss of 5%-15% or minor equipment damage. 0-1 points correspond to a production loss <5% and no equipment damage.

[0062] The initial risk indicators are adjusted for suitability. Referring to parameters such as current well production, formation pressure, and equipment service life, a deviation correction coefficient table is established: 1.2 for production exceeding the design value by 10%, and 0.8 for production falling below the design value by 10%; 1.1 for formation pressure exceeding the standard value by 5%, and 0.9 for formation pressure falling below the standard value by 5%; and 1.3 for equipment service life exceeding the design life by 50%, and 1.0 for equipment service life within the design life. The upper and lower limits of the suitability adjustment are ±30% of the base score, ultimately yielding a quantitative risk characterization of the target well.

[0063] Using the obtained risk quantification as the retrieval basis, a comprehensive search is conducted in the historical case database of the target oil well for past failure cases with similar risk quantification. Cases with similar risk levels, failure modes, and well conditions to the current target oil well are selected. Well workover strategies that have been used and proven effective in these cases are extracted and compiled to form a candidate strategy set for the target oil well.

[0064] A comprehensive analysis of each strategy in the candidate strategy set is conducted. Combining the current specific failure mode of the target oil well, risk quantification, and actual production needs, the candidate strategies are optimized, integrated, and dynamically arranged. The execution sequence, operation steps, required equipment and resources, key control points, and connection logic between each step are clarified. Redundant links in the strategy that are not compatible with the current working conditions are eliminated, and necessary adaptation adjustment measures are added. Finally, a complete, feasible, and highly targeted simulated well workover plan for the target oil well is formed.

[0065] The beneficial effects are that by extracting key indicators from multiple dimensions and dynamically assigning weights, the risk of well failure can be accurately quantified and assessed. Historical case retrieval based on quantitative characterization ensures the rationality and feasibility of candidate strategies, while dynamic process orchestration allows the workover plan to closely match the actual working conditions of the target well, effectively improving the adaptability and operability of the workover plan. This provides scientific and reliable guidance for the efficient implementation of subsequent workover operations and significantly reduces the blindness and risk costs of workover operations.

[0066] S6. Based on the simulated well workover scheme, logical deduction is performed on the key operation node data of the target oil well to obtain the simulation effect of the target oil well.

[0067] In this embodiment of the invention, the step of logically deducing the key operational node data of the target oil well based on the simulated well workover scheme to obtain the simulation effect of the target oil well includes: Key nodes of the simulated well workover scheme are extracted to obtain the discretized operation nodes of the target oil well; Path planning is performed on the discretized operation nodes to obtain the initial operation path of the target oil well; Resource consumption simulations were performed on the initial proposed path to obtain feasible operational paths for the target oil well; Based on feasible operation paths, the execution status of discrete operation nodes is reproduced by time-series simulation to obtain the simulation effect of the target oil well.

[0068] A comprehensive review of all operational steps involved in the simulated well workover scheme was conducted, and the core operational links that play a decisive role in the success or failure of the well workover operation were selected. These links include key aspects such as installation and commissioning of workover equipment, location of downhole fault points, running in of workover tools, fault handling, and well completion testing. Each core operational link was broken down as an independent operational unit, ultimately resulting in the discretized operational nodes of the target oil well.

[0069] Based on the sequential logical relationship and operational process requirements of the discrete operation nodes, the execution order of each operation node is clarified. At the same time, combined with the actual well conditions, downhole environment and technical specifications of well workover operations, an operational route is planned that can smoothly connect each operation node, avoid process conflicts and meet safety operation standards. This route clearly defines the start conditions, execution time and connection method with the next node of each operation node, thereby obtaining the initial operation path of the target oil well.

[0070] Based on the initial operation path, the various resources required for the execution of each discrete operation node are calculated one by one, including the usage time of workover equipment, fuel or electricity consumption, wear and tear of workover tools, manpower input and working time, and the consumption of various consumables. Through precise statistics and calculation, it is determined whether there is a shortage, waste or unreasonable allocation of resources in the initial operation path. Path segments with resource consumption exceeding expectations or unreasonable allocation are adjusted and optimized to ensure that the operation path is feasible in terms of resource guarantee, and finally a feasible operation path for the target oil well is obtained.

[0071] According to the operation sequence and time arrangement specified in the feasible operation path, the execution process of each discrete operation node is simulated step by step. The operation actions, equipment operating status, parameter changes and connection process between nodes are simulated in detail. Various data in the simulation process are recorded in real time, including operation progress, equipment operating parameters, failure rate, resource consumption dynamics, etc. These simulation data and processes are integrated in a time sequence to fully reproduce the entire process of well workover operation from start to finish. Finally, a simulation effect of the target oil well can be formed that can intuitively reflect the well workover operation effect, resource consumption and operation risk.

[0072] The beneficial effects are that key nodes are extracted to accurately identify core operational links, path planning ensures the rationality of the operation process, resource consumption simulation ensures the feasibility of the operation, and time-series simulation fully restores the entire operation process. The final simulation results can predict the actual situation of well workover operations in advance, providing a reliable basis for optimizing well workover plans, avoiding operational risks, and rationally allocating resources, effectively improving the success rate and efficiency of well workover operations.

[0073] like Figure 2 The diagram shown is a functional block diagram of an oil well failure and well workover process simulation system provided in an embodiment of the present invention.

[0074] The oil well failure and workover process simulation system 100 described in this invention can be installed in an electronic device. Depending on the functions implemented, the oil well failure and workover process simulation system 100 may include a multi-source data association and mining module 101, a knowledge-guided symptom identification module 102, a failure feature encoding module 103, a matching and judgment module 104, a scheme construction module 105, and a simulation effect generation module 106. The modules described in this invention can also be called units, referring to a series of computer program segments that can be executed by the processor of an electronic device and perform a fixed function, stored in the memory of the electronic device.

[0075] In this embodiment, the functions of each module / unit are as follows: The multi-source data association mining module 101 is used to perform collaborative association mining on the static parameter data and dynamic production history data of the target oil well to obtain multi-source association data of the target oil well. The knowledge-guided symptom identification module 102 is used to identify the symptom of the target oil well based on multi-source correlation data, and obtain a combination of potential failure symptom data of the target oil well. The failure feature encoding module 103 is used to perform feature clustering encoding on the combination of potential failure symptom data to obtain the oil well failure mode feature library of the target oil well. The matching and analysis module 104 is used to match and analyze the current operating data of the target oil well based on the oil well failure mode feature library to obtain the failure mode record of the target oil well. The scheme construction module 105 is used to perform quantitative adaptation assessment of the failure risk of the target oil well based on the failure mode record, obtain a quantitative characterization of the target oil well, and construct a strategy for the target oil well based on the quantitative characterization to obtain a simulated well workover scheme for the target oil well. The simulation effect generation module 106 is used to perform logical deduction on the key operation node data of the target oil well based on the simulated well repair scheme, so as to obtain the simulation effect of the target oil well.

[0076] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0077] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0078] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0079] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0080] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0081] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for simulating oil well failure and well workover process, characterized in that, The method includes: S1. Collaborative correlation mining is performed on the static parameter data and dynamic production history data of the target oil well to obtain multi-source correlation data of the target oil well; S2. Based on multi-source correlation data, knowledge-guided symptom identification is performed on the target oil well to obtain a combination of potential failure symptom data for the target oil well. S3. Perform feature clustering encoding on the combination of potential failure symptom data to obtain the oil well failure mode feature library of the target oil well; S4. Based on the oil well failure mode feature library, match and analyze the current operating data of the target oil well to obtain the failure mode record of the target oil well. S5. Based on the failure mode record, the failure risk of the target oil well is quantitatively adapted and evaluated to obtain the quantitative characterization of the target oil well. Based on the quantitative characterization, a strategy is constructed for the target oil well to obtain a simulated well workover plan for the target oil well. S6. Based on the simulated well workover scheme, logical deduction is performed on the key operation node data of the target oil well to obtain the simulation effect of the target oil well.

2. The method for simulating oil well failure and well workover process as described in claim 1, characterized in that, The method involves collaborative correlation mining of the static parameter data and dynamic production history data of the target oil well to obtain multi-source correlation data of the target oil well, including: Acquire static parameter data and dynamic production history data of the target oil well. Static parameter data includes well completion structure data and formation physical property data, while dynamic production history data includes production data, pressure data, and well workover operation data. By performing time-series alignment between static parameter data and dynamic production history data, the spatiotemporal aligned data of the target oil well is obtained; Feature coupling mining is performed on spatiotemporally aligned data to obtain multi-source correlation data of the target oil well.

3. The method for simulating oil well failure and well workover process as described in claim 1, characterized in that, The method of identifying potential failure symptom data of target oil wells based on multi-source correlation data, and obtaining a combination of potential failure symptom data of target oil wells, includes: Knowledge condensation is performed on historical failure cases of the target oil well to obtain the knowledge characteristics of the target oil well; Based on prior knowledge features, the key parameters in multi-source correlated data are mapped to their importance to obtain feature analysis guidance for the target oil well; Based on feature analysis guidance, anomaly detection is performed on the time-series evolution process in multi-source correlated data to obtain the primary anomaly pattern of the target oil well. The intensity of the primary anomaly patterns is calibrated to obtain a combination of potential failure symptom data for the target oil well.

4. The method for simulating oil well failure and well workover process as described in claim 3, characterized in that, The method of mapping the importance of key parameters in multi-source correlated data based on knowledge features to obtain feature analysis guidance for the target oil well includes: Logical deconstruction of knowledge features yields a dynamic parameter constraint framework for the target oil well; Based on a dynamic parameter constraint framework, the correlation of multi-source associated data is measured to obtain the parameter correlation degree of multi-source associated data. The variance contribution of the parameter correlation is evaluated to obtain the contribution weights of the multi-source correlation data; Core nodes are identified to determine the correlation degree of parameters, thereby obtaining the dominant parameters of multi-source correlated data. Based on the dominant parameters and contribution weights, online attention allocation is performed on multi-source correlated data to obtain characteristic analysis guidance for target oil wells.

5. The method for simulating oil well failure and well workover process as described in claim 1, characterized in that, The process of performing feature clustering and encoding on the combination of potential failure symptom data to obtain a well failure mode feature library for the target oil well includes: The potential failure symptom data combination is vectorized to obtain the standardized symptom vector of the target oil well; The dimensionality reduction mapping of the standardized symptom vector yields the dimensionality reduction feature representation of the target oil well; Based on the dimensionality reduction feature representation, the symptom similarity of the standardized symptom vectors is calculated. The formula for calculating the symptom similarity is as follows: ; in, For the first The and the first The similarity value of each symptom For the first The t-th eigenvalue of a standardized symptom vector For the first The t-th eigenvalue of a standardized symptom vector The number of feature dimensions for the standardized symptom vector; Based on symptom similarity, spatial clustering is performed on the dimensionality-reduced feature representation to obtain a preliminary cluster division of the target oil wells; Pattern induction was performed on the preliminary cluster classification to obtain the failure modes of the target oil well; Abstract failure modes are encapsulated through knowledge association to obtain a feature library of oil well failure modes for the target oil well.

6. The method for simulating oil well failure and well workover process as described in claim 1, characterized in that, The method involves matching and analyzing the current operating data of the target oil well based on the oil well failure mode feature library to obtain the failure mode record of the target oil well, including: Real-time monitoring of the current operating data of the target oil well; Based on the oil well failure mode feature library, pattern matching is performed on the current operating data to obtain candidate failure modes of the target oil well. Contextual semantic fusion is performed on candidate failure modes to obtain the comprehensive failure mode of the target oil well; The integrity of the comprehensive failure modes is verified to obtain the failure mode record of the target oil well.

7. The method for simulating oil well failure and well workover process as described in claim 1, characterized in that, The method involves quantitatively adapting and assessing the failure risk of the target oil well based on failure mode records, obtaining a quantitative characterization of the target oil well, and constructing a strategy based on this characterization to obtain a simulated well workover plan for the target oil well, including: Key indicators are extracted from failure mode records to obtain the risk characteristics of the target oil well; By dynamically assigning weights to risk characteristics, preliminary risk indicators for the target oil well are obtained. The initial risk indicators are adjusted for suitability to obtain a quantitative risk characterization of the target oil well. Based on risk quantification, the historical case database of the target oil well is searched and adapted to obtain candidate strategies for the target oil well. The candidate strategies are dynamically programmed to obtain a simulated well workover plan for the target oil well.

8. The method for simulating oil well failure and well workover process as described in claim 7, characterized in that, The dynamic weighting of risk characteristics to obtain preliminary risk indicators for the target oil well includes: A dual assessment of risk characteristics yields a basic score for those characteristics. Based on the current operating data of the target oil well, the basic score is corrected for deviation to obtain the corrected score of risk characteristics; Cross-dimensional correlation analysis was performed on the corrected scores to obtain the interaction coefficients between risk characteristics; Based on the interaction influence coefficient, weights are assigned to the current working condition data to obtain the scenario-adaptive weights of the risk characteristics; Based on the revised score and scenario adaptation weights, the risk characteristics are comprehensively quantified to obtain the preliminary risk indicators of the target oil well.

9. The method for simulating oil well failure and well workover process as described in claim 1, characterized in that, The simulation-based well workover scheme involves logically deducing the key operational node data of the target oil well to obtain the simulation effect of the target oil well, including: Key nodes of the simulated well workover scheme are extracted to obtain the discretized operation nodes of the target oil well; Path planning is performed on the discretized operation nodes to obtain the initial operation path of the target oil well; Resource consumption simulations were performed on the initial operation path to obtain a feasible operation path for the target oil well; Based on feasible operation paths, the execution status of discrete operation nodes is reproduced by time-series simulation to obtain the simulation effect of the target oil well.

10. A simulation system for oil well failure and well workover process, characterized in that, The system for implementing the oil well failure and well workover process simulation method of claim 1 includes: The multi-source data association mining module is used to collaboratively associate and mine the static parameter data and dynamic production history data of the target oil well to obtain the multi-source association data of the target oil well. The knowledge-guided symptom identification module is used to identify the symptom of a target oil well based on multi-source correlation data, and obtain a combination of potential failure symptom data of the target oil well. The failure feature encoding module is used to perform feature clustering encoding on the combination of potential failure symptom data to obtain the oil well failure mode feature library of the target oil well; The matching and analysis module is used to match and analyze the current operating data of the target oil well based on the oil well failure mode feature library to obtain the failure mode record of the target oil well. The scheme construction module is used to quantitatively adapt and assess the failure risk of the target oil well based on the failure mode record, obtain the quantitative characterization of the target oil well, and construct a strategy for the target oil well based on the quantitative characterization to obtain the simulated well workover scheme for the target oil well. The simulation effect generation module is used to logically deduce the key operational node data of the target oil well based on the simulated well workover scheme, and obtain the simulation effect of the target oil well.

Images