Oil and gas production alarm model construction method and device based on working condition base point
By calibrating the operating condition baseline, an oil and gas production early warning model was constructed, which solved the problems of low early warning efficiency and low reliability in oil and gas production scenarios, achieved high-precision early warning and timely response, reduced false alarms, and improved resource utilization.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-26
- Publication Date
- 2026-03-27
AI Technical Summary
Existing safety supervision methods for oil and gas production scenarios suffer from low early warning efficiency, low reliability, and incomplete early warning, making them unsuitable for oil and gas production scenarios with multiple stages and operating conditions.
By calibrating the operating condition baseline, an oil and gas production early warning model based on the operating condition baseline is constructed. Historical and real-time datasets are used for classification, and early warning models for different types are established. The early warning model is then used for timely response.
It has improved the targeting and accuracy of early warnings, reduced false alarms and invalid alarms, made rational use of resources, reduced the workload of regulatory personnel, and ensured production safety.
Smart Images
Figure CN121743953A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of oil and gas safety management technology, and specifically relates to a method and device for constructing an oil and gas production alarm model based on operating condition benchmarks. Background Technology
[0002] The purpose of oil and gas extraction is to achieve higher production volumes of oil and natural gas. Oil and gas production is influenced by a variety of factors, with hazards originating from various sources. From the perspective of construction technology and oil and gas extraction operating procedures, the main influencing factor is the professional and technical level and knowledge of the construction personnel. From the perspective of production safety, the main influencing factors are equipment performance and production efficiency. Improper operation may lead to safety accidents such as fires, explosions, and poisoning.
[0003] Currently, safety management in oil and gas production scenarios mainly involves two approaches:
[0004] The first method is human supervision, which involves appointing safety supervisors who use handheld monitoring equipment (such as gas concentration detectors) to monitor abnormal situations in oil and gas production environments. When an anomaly is detected, they issue warnings and direct the evacuation of workers. However, this method relies not only on the supervisors' experience and skill but also on their sense of responsibility, and is prone to problems such as low warning efficiency and reliability due to untimely or inadequate supervision.
[0005] The second approach is equipment monitoring, which involves installing sensors such as temperature sensors and other concentration sensors in the oil and gas production environment. These sensors collect data and upload it to an early warning center, where the center makes judgments and analyses. Compared to the first approach, this method can improve early warning efficiency and reliability to some extent; however, it suffers from incomplete warnings and is not suitable for oil and gas production scenarios with multiple stages and operating conditions. Summary of the Invention
[0006] To address the problems in the background art, this invention proposes a method and apparatus for constructing an oil and gas production alarm model based on operating condition benchmarks.
[0007] To achieve the above objectives, the present invention adopts the following technical solution:
[0008] A method for constructing an oil and gas production early warning model based on operating condition baselines, the method comprising:
[0009] Calibrate the first operating condition baseline;
[0010] Obtain historical datasets of oil and gas wells;
[0011] The historical dataset is classified based on preset classification criteria to obtain multiple sets of sub-historical datasets of different types; wherein, one type corresponds to one set of sub-historical datasets.
[0012] Based on the first operating condition baseline, a first early warning model is constructed according to one set of sub-historical datasets from the multiple sets of sub-historical datasets;
[0013] Iterate through all sub-historical datasets until you obtain the first early warning model for all types.
[0014] Furthermore, the preset classification conditions include: sampling location, temperature, and concentration of harmful gases.
[0015] Furthermore, the calibration of the first operating condition baseline includes:
[0016] Constructing multi-level feature parameters;
[0017] Select feature parameters;
[0018] Define the range of the characteristic parameters.
[0019] Furthermore, the method also includes:
[0020] Obtain real-time datasets of oil and gas wells;
[0021] The real-time dataset is classified based on preset classification conditions to obtain multiple sets of sub-real-time datasets of different types; wherein, one type corresponds to one set of sub-real-time datasets.
[0022] The first early warning model is used to analyze the sub-real-time dataset until early warning analysis results for all types of real-time datasets are obtained.
[0023] Furthermore, the method also includes:
[0024] The first operating condition base point is updated based on the early warning analysis results to obtain the second operating condition base point;
[0025] Based on the second operating condition baseline, the first early warning model is updated to obtain the second early warning model.
[0026] Furthermore, if the real-time dataset includes a newly added type of sub-real-time dataset, the method further includes:
[0027] Based on the second operating condition baseline, a first early warning model corresponding to the newly added type sub-real-time dataset is constructed according to the newly added type sub-real-time dataset.
[0028] Furthermore, the method also includes:
[0029] Based on the historical data, update the first operating condition baseline and calibrate the third operating condition baseline.
[0030] Based on the third operating condition baseline, the first early warning model is updated to obtain the third early warning model.
[0031] Furthermore, the method also includes:
[0032] Based on the first early warning model, construct the first early warning elimination model corresponding to the type;
[0033] Based on the early warning analysis results, determine whether the corresponding type of sub-real-time dataset meets the early warning conditions;
[0034] If the corresponding type of sub-real-time dataset meets the warning conditions, then determine whether the warning analysis result meets the preset warning cancellation conditions;
[0035] If the early warning analysis result meets the preset alarm cancellation conditions, then the first alarm cancellation model of the type corresponding to the early warning analysis result is matched;
[0036] Based on the first alarm cancellation model corresponding to the type of the early warning analysis results, an alarm cancellation operation is performed.
[0037] Furthermore, the preset alarm cancellation conditions include: the type of the sub-real-time dataset corresponding to the early warning analysis result and the triggering source of the sub-real-time dataset.
[0038] Secondly, this invention proposes an oil and gas production alarm model construction device based on operating condition benchmarks, comprising:
[0039] The operating condition baseline calibration module is used to calibrate the first operating condition baseline.
[0040] The historical dataset acquisition module is used to acquire historical datasets of oil and gas wells;
[0041] The historical dataset classification module is used to classify the historical dataset based on preset classification conditions to obtain multiple sets of sub-historical datasets of different types; wherein, one type corresponds to one set of sub-historical datasets.
[0042] The single-type early warning model construction module is used to construct a first early warning model based on the first operating condition baseline and according to one set of sub-historical datasets from multiple sets of sub-historical datasets;
[0043] The multi-type early warning model construction module is used to traverse all sub-historical datasets until the first early warning model corresponding to all types is obtained.
[0044] The beneficial effects of this invention are:
[0045] The method of this invention provides a benchmark or reference standard for the operation of processes or equipment by calibrating the operating condition baseline. Any data that deviates from this operating condition baseline may indicate a problem, which helps the early warning model to accurately identify abnormal situations in the production process. Furthermore, by establishing corresponding early warning models for different types of data, the pertinence and accuracy of early warnings can be significantly improved. This effectively solves the problems of frequent false alarms, invalid alarms, and excessive alarm volume caused by network fluctuations, instrument failures, and data quality issues, as well as the workload of monitoring personnel and technicians.
[0046] The method of this invention, by establishing an alarm cancellation model, can make reasonable use of existing resources, which is conducive to improving the utilization rate of existing resources, and also shortens the time for manual verification and confirmation.
[0047] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures pointed out in the description and the drawings. Attached Figure Description
[0048] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0049] Figure 1 The flowchart of the oil and gas production alarm model construction method based on operating condition baselines of the present invention is shown;
[0050] Figure 2 The diagram shows the framework of the oil and gas production alarm model construction system based on operating condition benchmarks according to the present invention. Detailed Implementation
[0051] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0052] Refer to the instruction manual appendix Figure 1 As shown, a method for constructing an oil and gas production alarm model based on operating condition benchmarks includes the following steps:
[0053] S1. Calibrate the first operating condition baseline;
[0054] In step S1, the operating condition baseline serves two purposes: firstly, it is a quantitative indicator representing the healthy state of various monitoring and analysis aspects such as process equipment, production dynamics, and production objects; secondly, it is a reasonable criterion for reverting from abnormal operating conditions to normal operating conditions. Therefore, calibrating the first operating condition baseline is fundamental to constructing the early warning model. It provides a standard operating state reference, serves as a benchmark for the subsequent construction of the early warning model, and ensures that the early warning model can accurately identify the deviation between actual operation and ideal state, thereby issuing timely warnings.
[0055] When assessing the operating conditions of processes or equipment related to well sites, production and injection metering stations, oil transfer and processing stations, collectable data (operating parameters, operating status, manually marked status, etc.) can be used for evaluation. Different processes and equipment need to establish targeted normal operating condition baseline definition standards based on their operating characteristics.
[0056] Specifically, the first operating condition baseline is determined, including:
[0057] S11. Construct multi-level feature parameters: formed by the convergence of sub-process or sub-equipment operating conditions;
[0058] S12. Select characteristic parameters: Multiple parameters are allowed, each parameter corresponds to a reasonable range, and the working condition is considered normal only when all parameters are within the reasonable range.
[0059] S13. The range of calibration characteristic parameters: The initial value is manually calibrated, and the alarm threshold can be referenced, taking into account the differences in process conditions.
[0060] For example, the first operating condition baseline is calibrated using a beam pumping unit well:
[0061] The baseline for the first operating condition of an oil well after well opening:
[0062] Current load [3500, 7465];
[0063] Load span [1980, 2800];
[0064] Back pressure [1.4, 2.4];
[0065] The pressure range is [1.7, 2.8].
[0066] S2. Obtain historical datasets of oil and gas wells;
[0067] In step S2, the historical dataset contains operational records of oil and gas wells under different times and conditions, which is a valuable resource for analyzing the operational patterns of oil and gas wells and identifying potential problems. By collecting historical data, it can be ensured that the early warning model can learn various operational modes and abnormal characteristics of oil and gas wells, thereby improving the accuracy of early warnings.
[0068] S3. Classify the historical dataset based on preset classification conditions to obtain multiple sets of different types of sub-historical datasets; where one type corresponds to one set of sub-historical datasets.
[0069] In step S3, the preset classification conditions are used to divide the complex and diverse historical dataset into several subsets with the same or similar characteristics. The preset classification conditions can be based on the collection location or on the impact on key factors that require abnormal early warning (such as temperature, concentration of harmful gases, etc.).
[0070] S4. Based on the first operating condition baseline, construct the first early warning model according to one set of sub-historical datasets from multiple sets of sub-historical datasets;
[0071] In step S4, an early warning model is constructed using historical datasets of the corresponding type, enabling the model to learn the normal operating mode and abnormal characteristics under this operating condition type; when the actual operating data deviates from the range predicted by the model, the early warning mechanism is triggered.
[0072] S5. Traverse all sub-historical datasets until the first early warning model corresponding to all types is obtained;
[0073] For the several established primary early warning models, which can provide early warnings for oil and gas wells, in order to achieve real-time monitoring and risk warning of oil and gas wells, it is first necessary to acquire real-time production data. Specifically, this includes the following steps:
[0074] Obtain real-time datasets of oil and gas wells;
[0075] The real-time dataset is classified based on preset classification criteria to obtain multiple sets of sub-real-time datasets of different types; one type corresponds to one set of sub-real-time datasets.
[0076] The first early warning model is used to analyze the sub-real-time datasets until early warning analysis results for all types of real-time datasets are obtained.
[0077] Since different types of data reflect different production conditions or circumstances, a corresponding early warning model is built for each type of data, which can more accurately capture abnormal situations in operation.
[0078] In some embodiments, during the process of performing early warning analysis and realizing early warning of real-time oil and gas production process based on the trained first early warning model, the model parameters of the target early warning model can be updated according to the data types newly collected in the real-time dataset collected during the early warning process or the data values collected for existing data types. For example, the reasonable range of feature parameters can be updated, the second operating condition base point can be calibrated, and then the first early warning model can be updated based on the updated second operating condition base point.
[0079] The real-time datasets collected during the early warning process may contain data of the same type as existing datasets, or may include newly added dataset types. If new data types are collected during the early warning process, a first early warning model corresponding to the newly added sub-real-time dataset is constructed based on the second operating condition baseline and the newly added sub-real-time dataset type.
[0080] In some embodiments, dynamic statistics are performed based on historical data to calculate and update the reasonable range of characteristic parameters, update the first operating condition baseline, calibrate the third operating condition baseline, and update the first early warning model based on the third operating condition baseline to obtain the third early warning model. The dynamic statistical methods include: the extreme value method, the average value method, and the normal distribution.
[0081] This invention constructs an early warning model and calibrates the operating condition baseline, enabling the early warning model to accurately identify the deviation between the actual operating conditions and the ideal operating conditions, thereby issuing timely early warnings.
[0082] This invention provides a benchmark or reference standard for the operation of processes or equipment by calibrating operating condition baselines. Any data deviating from this operating condition baseline may indicate a problem, helping the early warning model to accurately identify abnormal situations in the production process. Furthermore, by establishing corresponding early warning models for different types of data, the targeting and accuracy of early warnings can be significantly improved. This effectively solves the problems of frequent false alarms, invalid alarms, and excessive alarm volume caused by network fluctuations, instrument failures, and data quality issues, as well as the workload of monitoring and technical personnel.
[0083] However, relying solely on early warning signals is insufficient; timely and effective measures to extinguish them are essential to building a more comprehensive and efficient production safety assurance system. Extinguishing alarms is not only a proactive response to warning signals but also a crucial step in ensuring production safety and minimizing potential losses.
[0084] Specifically, based on the first early warning model, a first alarm cancellation model corresponding to the type is constructed;
[0085] Based on the early warning analysis results, determine whether the corresponding type of sub-real-time dataset meets the early warning conditions;
[0086] If the corresponding type of sub-real-time dataset meets the early warning conditions, then it is determined whether the early warning analysis result meets the preset early warning cancellation conditions. The preset early warning cancellation conditions include: the type of the sub-real-time dataset corresponding to the early warning analysis result and the triggering source of the sub-real-time dataset. For example, if the influencing parameter is a temperature parameter, and its triggering source is that the operating temperature of the oil and gas extraction equipment is too high, and the parameter value corresponding to the influencing parameter is within a controllable range, then it meets the preset early warning cancellation conditions.
[0087] If the early warning analysis results meet the preset alarm cancellation conditions, then the first alarm cancellation model of the corresponding type of early warning analysis results will be matched.
[0088] Based on the first alarm cancellation model corresponding to the early warning analysis results, the alarm cancellation operation is executed.
[0089] Specifically, the alarm suppression operation is as follows: In the current production scenario, there are multiple existing resource conditions that can weaken or eliminate the impact parameters corresponding to the early warning analysis results. The alarm suppression model will first identify and list all available alarm suppression resources, forming multiple sets of optional alarm suppression resource schemes;
[0090] For each set of selectable alarm-reduction resources, the alarm-reduction model will estimate the alarm-reduction result and alarm-reduction cost based on relevant information from the early warning analysis results (such as specific parameter values, trigger sources, and generation locations). The alarm-reduction result includes the degree of alarm-reduction, and the alarm-reduction cost includes the consumption of alarm-reduction resources and / or the degree of delay to oil and gas production.
[0091] After evaluating the alarm suppression results and costs of all available alarm suppression resources, the alarm suppression model will select the most suitable alarm suppression resource to perform the alarm suppression operation. For example, it will select an alarm suppression resource with a high degree of alarm suppression, low resource consumption, and low disruption to oil and gas production. The alarm suppression resource may include human resources and / or equipment resources.
[0092] This invention establishes an alarm suppression model, which enables the rational use of existing resources and helps to improve the utilization rate of existing resources.
[0093] Refer to the instruction manual appendix Figure 2 As shown, based on the same inventive concept, this invention also proposes a device for constructing an oil and gas production alarm model based on operating condition benchmarks, comprising:
[0094] Operating condition baseline calibration module 110 is used to calibrate the first operating condition baseline;
[0095] The historical dataset acquisition module 120 is used to acquire historical datasets of oil and gas wells; the historical datasets include multiple sets of sub-historical datasets of different types.
[0096] The historical dataset classification module 130 is used to classify historical datasets based on preset classification conditions to obtain multiple sets of sub-historical datasets corresponding to multiple types; wherein, one type corresponds to one set of sub-historical datasets.
[0097] The single-type early warning model construction module 140 is used to construct a first early warning model corresponding to a type based on the first operating condition base point and according to one set of sub-historical datasets from multiple sets of sub-historical datasets;
[0098] The multi-type early warning model construction module 150 is used to traverse all sub-datasets in multiple sets of sub-historical datasets until multiple first early warning models corresponding to multiple types are obtained.
[0099] Based on the same inventive concept, this disclosure also provides a computer-readable storage medium storing one or more programs, which, when executed, can realize the aforementioned method for constructing an oil and gas production early warning model based on operating condition benchmarks.
[0100] Based on the same inventive concept, this disclosure also provides an electronic device, including a processor, a communication interface, a memory, and a communication bus. The processor, communication interface, and memory communicate with each other via the communication bus. The processor is used to execute a program stored in the aforementioned computer-readable storage medium.
[0101] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.
[0102] Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for constructing an oil and gas production early warning model based on operating condition baselines, characterized in that, The method includes: Calibrate the first operating condition baseline; Obtain historical datasets of oil and gas wells; The historical dataset is classified based on preset classification criteria to obtain multiple sets of sub-historical datasets of different types; wherein, one type corresponds to one set of sub-historical datasets. Based on the first operating condition baseline, a first early warning model is constructed according to one set of sub-historical datasets from the multiple sets of sub-historical datasets; Iterate through all sub-historical datasets until you obtain the first early warning model for all types.
2. The method for constructing an oil and gas production alarm model based on operating condition benchmarks according to claim 1, characterized in that, The preset classification conditions include: sampling location, temperature, and concentration of harmful gases.
3. The method for constructing an oil and gas production alarm model based on operating condition benchmarks according to claim 1, characterized in that, The calibration of the first operating condition baseline includes: Constructing multi-level feature parameters; Select feature parameters; Define the range of the characteristic parameters.
4. The method for constructing an oil and gas production alarm model based on operating condition benchmarks according to claim 1, characterized in that, The method further includes: Obtain real-time datasets of oil and gas wells; The real-time dataset is classified based on preset classification conditions to obtain multiple sets of sub-real-time datasets of different types; wherein, one type corresponds to one set of sub-real-time datasets. The first early warning model is used to analyze the sub-real-time dataset until early warning analysis results for all types of real-time datasets are obtained.
5. The method for constructing an oil and gas production early warning model based on operating condition benchmarks according to claim 4, characterized in that, The method further includes: The first operating condition base point is updated based on the early warning analysis results to obtain the second operating condition base point; Based on the second operating condition baseline, the first early warning model is updated to obtain the second early warning model.
6. The method for constructing an oil and gas production early warning model based on operating condition benchmarks according to claim 5, characterized in that, If the real-time dataset includes a newly added type of sub-real-time dataset, then the method further includes: Based on the second operating condition baseline, a first early warning model corresponding to the newly added type sub-real-time dataset is constructed according to the newly added type sub-real-time dataset.
7. The method for constructing an oil and gas production early warning model based on operating condition benchmarks according to claim 1, characterized in that, The method further includes: Based on the historical data, update the first operating condition baseline and calibrate the third operating condition baseline. Based on the third operating condition baseline, the first early warning model is updated to obtain the third early warning model.
8. The method for constructing an oil and gas production early warning model based on operating condition benchmarks according to claim 4, characterized in that, The method further includes: Based on the first early warning model, construct the first early warning elimination model corresponding to the type; Based on the early warning analysis results, determine whether the corresponding type of sub-real-time dataset meets the early warning conditions; If the corresponding type of sub-real-time dataset meets the warning conditions, then determine whether the warning analysis result meets the preset warning cancellation conditions; If the early warning analysis result meets the preset alarm cancellation conditions, then the first alarm cancellation model of the type corresponding to the early warning analysis result is matched; Based on the first alarm cancellation model corresponding to the type of the early warning analysis results, an alarm cancellation operation is performed.
9. The method for constructing an oil and gas production early warning model based on operating condition benchmarks according to claim 8, characterized in that, The preset alarm cancellation conditions include: the type of the sub-real-time dataset corresponding to the early warning analysis result and the triggering source of the sub-real-time dataset.
10. A device for constructing an oil and gas production alarm model based on operating condition benchmarks, characterized in that, include: The operating condition baseline calibration module is used to calibrate the first operating condition baseline. The historical dataset acquisition module is used to acquire historical datasets of oil and gas wells; The historical dataset classification module is used to classify the historical dataset based on preset classification conditions to obtain multiple sets of sub-historical datasets of different types; wherein, one type corresponds to one set of sub-historical datasets. The single-type early warning model construction module is used to construct a first early warning model based on the first operating condition baseline and according to one set of sub-historical datasets from multiple sets of sub-historical datasets; The multi-type early warning model construction module is used to traverse all sub-historical datasets until the first early warning model corresponding to all types is obtained.