Oil and gas equipment failure database construction method based on multiple algorithms

Abstract

The application discloses a kind of oil and gas equipment fault database construction methods based on multiple algorithms, method includes (i) the basic information of oil and gas equipment is collected;(ii) data management and configuration;(iii) oil and gas equipment fault record preprocessing;(iv) oil and gas equipment fault library analysis method and path;(v) oil and gas equipment fault library version update etc. Step.The application establishes oil and gas equipment hierarchy and fault mode, and the preprocessing and self-matching process of fault record;Innovative application OREDA method, OREDA-based bayesian method, multiple distribution life data analysis method Multiple algorithms, according to the failure characteristics and quantity of oil and gas equipment Actual selection of the best analysis method path, realize the result analysis and comparison between different methods;Through oil and gas equipment fault database construction and accumulation, realize higher scientificity and accuracy in equipment management, bring greater value and help for equipment reliability management and operation optimization.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of oil and gas equipment failure analysis, and particularly relates to an oil and gas equipment failure database construction method based on multiple algorithms. BACKGROUND

[0002] For the reliability research and database of oil and gas equipment, foreign related companies and institutions have developed greatly in recent years. Through continuous collection and arrangement of data, and application and analysis of some algorithms, a relatively complete database has been formed. For example, the OREDA database of DNV, the process equipment reliability database PERD of CCPS, etc. These data are key basic data and technologies for foreign related research institutions to carry out the whole life cycle management of oil and gas equipment.

[0003] For example, OREDA of DNV makes more introduction and application to the equipment classification method of the oil and gas industry and the calculation method of the failure rate; for the algorithm of the normal operation and maintenance stage of oil and gas equipment, OREDA obtains the failure rate of different failure modes of oil and gas equipment and the upper and lower limit values under the 90% confidence interval by using the exponential distribution method, and gives the standard deviation; at the same time, the method is studied and applied to the single sample and multiple samples, and finally a relatively complete failure rate library of oil and gas equipment is obtained.

[0004] Although the failure library of foreign oil and gas equipment is quite rich and complete, the foreign database is based on foreign equipment and facility failure data. There are obvious differences in the types of equipment and facilities, the operation conditions of equipment and facilities and the management mode of equipment and facilities between the foreign database and the domestic one. If the domestic oil and gas equipment and facility safety technology evaluation directly learns from the foreign database, the credibility of the evaluation result is relatively low; and the foreign database mainly calculates the failure rate by using the exponential distribution method, which is single. The calculated failure rate is a constant value. However, the failure rate of oil and gas equipment should change with time, especially for the equipment in the later stage of the "bathtub curve". In addition, there are few failure data of domestic oil and gas equipment, which cannot directly calculate the failure rate. Most of them use expert experience evaluation method to evaluate the reliability of oil and gas equipment, without systematically considering the enterprise's own equipment failure database obtained by using the Bayesian method based on the foreign failure library (such as OREDA), to provide reference for the reliability research of oil and gas equipment.

[0005] At present, the failure database analysis method of domestic oil and gas equipment is basically scattered. There is no clear analysis algorithm path and scientific and complete technical method based on failure data. There is no standardized process, and there is no related professional software system to realize the landing and application of equipment failure database; not to mention the specific feasible technical method of equipment reliability management and operation optimization based on the application of equipment failure database. SUMMARY

[0006] The application is proposed in view of the lack of clear, scientific and complete technical methods and standardized processes in the construction of domestic oil and gas equipment failure database, and aims to provide an oil and gas equipment failure database construction method based on multiple algorithms.

[0007] The application is realized by the following technical solutions:

[0008] An oil and gas equipment failure database construction method based on multiple algorithms comprises the following steps:

[0009] (i) Collecting basic information of oil and gas equipment

[0010] (ii) Data management and configuration

[0011] According to the collected basic information of oil and gas equipment, the information is established in the system, and the system is configured according to the business requirements of users;

[0012] (iii) Oil and gas equipment failure record preprocessing

[0013] The actual failure data of oil and gas equipment is established in the system, and the system is preprocessed to obtain the preprocessed equipment failure data of the system;

[0014] (iv) Oil and gas equipment failure library analysis

[0015] According to the actual failure characteristics and quantity of oil and gas equipment, the analysis path and method are selected to obtain the failure library data analysis result of oil and gas equipment, and the result is stored and imported into the oil and gas equipment failure database;

[0016] (v) Oil and gas equipment failure library version update

[0017] When there is a new field failure record of oil and gas equipment, steps (ii) and (iii) are repeated to iteratively update the failure library data of the equipment.

[0018] In the above technical solution, the basic information includes the platform type, equipment type, hierarchical structure, failure mode and code, cumulative running time of the equipment, and failure record of the equipment running site.

[0019] In the above technical solution, the failure record of the equipment running site and the cumulative running time of the equipment in step (iii) are established in the system by automatic or manual means.

[0020] In the above technical solution, the actual failure data of oil and gas equipment includes the failure record of the equipment running site and the cumulative running time of the equipment; the preprocessed equipment failure data of the system includes the standard equipment, failure mode classification, and equipment failure interval time; and the failure library data analysis result includes the equipment type to be analyzed, the failure rate and its upper and lower limits under different failure modes, and the standard deviation index.

[0021] In the technical solution, the algorithm in step (iv) includes any one or more of OREDA method, Bayesian method based on OREDA or multi-distribution life data analysis method.

[0022] In the technical solution, when the analysis path and algorithm are selected in step (iv), if there is no device failure record, the algorithm in OREDA is applied to perform analysis preferentially, the analyzed device type, failure rate and upper and lower limits thereof under different failure modes and standard deviation index are obtained and automatically stored, and the life data analysis method combined with the characteristics of the device type and failure mode is provided for comparative analysis.

[0023] In the technical solution, when the analysis path and algorithm are selected in step (iv), if the number of device failure records is small (0 < number of records < 5), the Bayesian method based on OREDA prior is applied to perform analysis preferentially, when the Bayesian algorithm analysis path is applied, the determination conditions and different paths of two cases that the prior distribution is Gamma distribution or not are determined, the path of Gamma distribution or the path of not Gamma distribution is selected according to the data of the same type of device and failure mode in the OREDA database, the failure database data analysis result of the oil and gas equipment is obtained and automatically saved, the life data analysis method combined with the characteristics of the device type and failure mode is provided for analysis and comparison, the corresponding result and graph are obtained and stored, and the corresponding data of the failure database is automatically obtained through system processing.

[0024] In the technical solution, when the analysis path and algorithm are selected in step (iv), if the number of device failure records is large (the number of device failure records is greater than or equal to 5), the multi-distribution life data analysis method is applied to analyze the data of two levels of device and failure mode, the analysis sub-path is selected and the data in the scene is analyzed, the failure database data analysis result of the oil and gas equipment is obtained and saved through system processing, and the reliability index and curve of the oil and gas equipment are provided.

[0025] Compared with the prior art, the present application has at least the following beneficial technical effects:

[0026] The application constructs a brand-new oil and gas equipment failure database construction method based on multiple algorithms, establishes an oil and gas equipment hierarchical structure and a failure mode, and a failure record preprocessing and self-matching process; the application innovatively applies multiple algorithms of OREDA method, OREDA-based Bayesian method, and multiple distribution life data analysis methods, and according to actual failure characteristics and quantity of oil and gas equipment, the best analysis method path is automatically recommended by the system, and other algorithms are selected manually for analysis, so that automatic optimization analysis of the oil and gas equipment failure database for each scene can be realized, and the results of different methods can be analyzed and compared; through oil and gas equipment failure database construction and accumulation, enterprises can realize higher scientificity and accuracy in equipment management, bring greater value and help to equipment reliability management and operation optimization, and provide more accurate data support for equipment selection, equipment safety level evaluation, and the like of a new platform or device. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 is a flowchart of the method of the application;

[0028] Figure 2 is a flowchart of oil and gas equipment failure database analysis in the method of the application.

[0029] For those skilled in the art, other related drawings can be obtained according to the above drawings without creative labor. DETAILED DESCRIPTION

[0030] In order for those skilled in the art to better understand the technical solutions of the application, the technical solutions of the application will be further described below by specific embodiments in combination with the drawings of the specification.

[0031] Embodiment 1

[0032] A multiple-algorithm-based oil and gas equipment failure database construction method, as shown in Figure 1 , includes the following steps:

[0033] (i) Collecting basic information of oil and gas equipment

[0034] The basic information includes a platform type where the oil and gas equipment is located, an equipment type, a hierarchical structure, a failure mode and code, a cumulative running time of the equipment, and a failure record of an equipment running site;

[0035] (ii) Data management and configuration

[0036] According to the collected basic information of the oil and gas equipment, the oil and gas equipment is established in the system, and the system is configured according to business needs of a user;

[0037] Defining the device type and hierarchy, and establishing a standard and flexible hierarchy according to the actual situation of the enterprise, realizing flexible configuration of the device type and automatic preprocessing of the fault record through the software system;

[0038] Analyzing and standardizing the fault mode of the device, distinguishing from the original fault mode and code shared by all devices, and setting a separate fault mode category for each type of device;

[0039] (ⅲ) Oil and gas equipment fault record preprocessing

[0040] The fault record of the device running site and the cumulative running time of the device are automatically or manually established in the system, and the system is preprocessed to obtain the preprocessed device fault data of the system;

[0041] The fault record data of the device is preprocessed to meet the requirements of the device type and hierarchy, the fault mode and code of the device set in the aforementioned data pipeline and configuration, and is processed into the format required by the algorithm analysis according to the key fields of the fault record form (such as automatic sorting of the fault record, interval calculation of the calendar time, self-prompting of abnormal data, etc.), and is classified according to the device and the fault mode, realizing the self-matching of the fault record of the device category and the fault mode;

[0042] (ⅳ) Oil and gas equipment fault library analysis

[0043] According to the actual failure characteristics and quantity of the oil and gas equipment, the results analysis and comparison between different methods are realized by automatically selecting the analysis path and method and supporting manual selection of other methods for analysis, and the fault library data analysis result of the oil and gas equipment is obtained and stored and imported into the oil and gas equipment fault database;

[0044] The actual failure characteristics of the device are the preprocessed data of step (ⅲ), and the quantity refers to the number of device fault records;

[0045] The fault library data analysis result includes the device type to be analyzed and the failure rate and its upper and lower limits under different fault modes of the device type, as well as the standard deviation index.

[0046] The algorithm in step (ⅳ) includes any one or more of the OREDA method, the Bayesian method based on OREDA, or the life data analysis method of multiple distributions.

[0047] The selection process of the analysis algorithm and path is shown in Figure 2 The specific method is as follows:

[0048] (a) For the case of no equipment failure record, the algorithm in OREDA is applied to analyze preferentially, and the analyzed equipment type and its failure rate under different failure modes and upper and lower limits and standard deviation index are obtained and automatically stored. Meanwhile, the life data analysis method combined with the characteristics of equipment type and failure mode is provided for comparative analysis, such as the Weibull distribution of single parameter, that is, the reasonable β parameter is selected according to the characteristics of equipment type or failure mode, and the corresponding data of the failure database is obtained;

[0049] (b) For the case of few equipment failure records (0 < record number < 5), the Bayesian method based on OREDA prior is applied to analyze preferentially. When the Bayesian algorithm is applied to analyze the path, the determination conditions and different paths of prior distribution as Gamma distribution (mean ≠ n / τ) or not as Gamma distribution (mean = n / τ) are clarified, and the corresponding path is selected according to the same type of equipment and failure mode data in the matched OREDA database, and the failure rate library parameters of oil and gas equipment (i.e. the analyzed equipment type and its failure rate under different failure modes and upper and lower limits and standard deviation index) are obtained and saved. Meanwhile, the life data analysis method combined with the characteristics of equipment type and failure mode is provided for analysis and comparison, such as the Weibull 2 parameter method, or the corresponding distribution such as the Weibull distribution of single parameter is manually selected, and the corresponding results and graphics are obtained and stored, and the corresponding data of the failure database is automatically obtained through system processing;

[0050] (c) For the case of sufficient equipment failure records (record number ≥ 5), a variety of distribution life data analysis methods are applied to analyze the data of two levels of equipment and failure mode, and the best matching selection and analysis of distribution model are carried out through correlation coefficient, LK value, ρ value, etc. That is, the system will automatically select and have a clear analysis sub-path to analyze the data of this scene, and through automatic processing of the system, the failure library parameters of oil and gas equipment (i.e. the analyzed equipment type and its failure rate under different failure modes and upper and lower limits and standard deviation index) are obtained and saved. The distribution type and its corresponding parameters are shown in the following table;

[0051] Distribution type and its corresponding parameters

[0052]

[0053]

[0054] At the same time, the oil and gas equipment reliability index and curve are provided, taking Weibull2 parameters as an example, shape parameter β, characteristic life parameter η, mean time between failures, B10 life, reliability and unreliability at a certain time, conditional reliability, and upper and lower confidence intervals of the above indexes are provided, and probability distribution graph, reliability and unreliability, probability density function, failure rate, contour map and self-defined markers are also provided; at the same time, for the failure record, manual selection is also supported, and Bayesian method is used for analysis and result comparison.

[0055] (v) Oil and gas equipment failure library version update

[0056] When new field failure records of oil and gas equipment are generated, steps (ii) and (iii) are repeated to iteratively update the failure library data of the equipment.

[0057] The oil and gas equipment hierarchical structure and failure mode, pre-processing and self-matching process of failure records, and various algorithms, paths and analysis processes of the new equipment failure database are preset in the software system based on the oil and gas equipment failure database construction method based on various algorithms; at the same time, the software system should also have the function of flexible matching through the data dictionary, realizing the flexible adjustment of the statistics and overview of the established oil and gas equipment type or failure mode data; the system can preprocess the imported oil and gas equipment failure data, automatically match the corresponding equipment type, and realize the automatic classification of the equipment and its subordinate failure mode data; according to the actual situation or scene of the equipment and failure mode, the system automatically recommends the optimal algorithm for analysis and result comparison storage, and other analysis methods can be applied for comparison and verification; after the analyzed data pass through the corresponding review process, they are stored in the oil and gas equipment failure library system, and data update iteration and version management functions are provided, and the data can be applied in subsequent equipment safety evaluation, maintenance optimization, new device selection, etc. Scene application; the system provides personnel permission management, log management, scalability, and other functions such as multi-scene development and application with other systems.

[0058] The application constructs a brand-new, more practical and more suitable oil and gas equipment failure database establishment method based on various algorithms, establishes flexible and standardized oil and gas equipment hierarchical structure, failure mode and coding model, and failure record preprocessing and self-matching process, clearly defines the processing flow of equipment failure record data, and realizes automatic processing of equipment source failure record to analyzable; through the various algorithms, paths and analysis process definitions of the brand-new equipment failure database, the analysis process and path for all scenarios are defined, so that the construction of the oil and gas equipment failure library has clear and feasible technical means; and in view of the situation that the failure rate will change in the actual operation and maintenance process of the oil and gas equipment, a new algorithm is introduced to establish and apply the equipment failure database; and the development of the software carrier improves the effectiveness and efficiency of the application and management of the above analysis method.

[0059] The application helps to form the failure database of an enterprise or even an industry, can accurately and quickly analyze the failure causes and characteristics of oil and gas equipment, provide treatment schemes, greatly improve the equipment operation and maintenance management efficiency, and can quantitatively show the reliability of the equipment through the failure rate or reliability index, improve the scientificity of the oil and gas equipment failure measurement, and construct a standardized and structured data and various algorithm automatic matching analysis which can be flexibly configured, realize the scientificity and continuous optimization of equipment reliability management.

[0060] Through the construction and accumulation of the oil and gas equipment failure database, the enterprise can realize higher scientificity and accuracy in equipment management, bring greater value and help to equipment reliability management and operation optimization, and provide more accurate data support for equipment selection, equipment safety level evaluation and the like of new platforms or devices.

[0061] The method of the application is a set of analysis method with originality and feasibility, simple operation and landing, can provide rich and accurate data support for equipment integrity and reliability management, and has great practical value.

[0062] The applicant declares that the above is only a specific embodiment of the application, but the protection scope of the application is not limited thereto, and those skilled in the art should understand that any changes or replacements within the technical scope disclosed by the application can be easily thought of by those skilled in the art, and all fall within the protection scope and disclosure scope of the application.

Claims

1. A method for building an oil and gas equipment failure database based on multiple algorithms, the method comprising: The method comprises the following steps: ​ (i) collecting basic information of oil and gas equipment The basic information includes the type of platform where the oil and gas equipment is located, the type of equipment, the hierarchical structure, the failure mode and code, the cumulative running time of the equipment and the failure record of the equipment running site; (ii) data management and configuration According to the collected basic information of the oil and gas equipment, the basic information is established in the system, and the configuration is carried out according to the business needs of the user; (iii) oil and gas equipment failure record preprocessing The actual scene data is established in the system, and the system is preprocessed to meet the requirements of the equipment type and hierarchical structure, the failure mode and code of the equipment in the aforementioned data management and configuration, and the key field of the failure record form is processed into the format required by the algorithm analysis, and the data classification is completed according to the equipment and the failure mode, the equipment category and the failure mode of the failure record are automatically matched, and the standard data that can be automatically analyzed by the system is obtained; (iv) oil and gas equipment failure library analysis According to the actual failure record characteristics of the oil and gas equipment, the analysis path and algorithm are automatically matched to obtain the failure library data analysis result of the corresponding oil and gas equipment, and the failure library data analysis result is stored and imported into the oil and gas equipment failure database; The algorithm includes any one or more of OREDA method, Bayesian method based on OREDA or multi-distribution life data analysis method; The oil and gas equipment failure library data analysis result includes the equipment type to be analyzed and the failure rate and its upper and lower limits under different failure modes, as well as the standard deviation index; If there is no equipment failure record, the algorithm in OREDA is selected for analysis, and the analyzed equipment type and its failure rate and its upper and lower limits under different failure modes, as well as the standard deviation index are automatically stored, and the life data analysis method combined with the characteristics of the equipment type and the failure mode is provided for comparative analysis; If the number of equipment failure records is 0 < number of records < 5, the Bayesian method based on OREDA prior is selected for analysis, and the Bayesian algorithm analysis path is applied, the judgment conditions and different paths of the prior distribution being Gamma distribution or not being Gamma distribution are determined, the path of Gamma distribution or the path of not being Gamma distribution is selected according to the same type of equipment and failure mode data in the OREDA database, the failure library data analysis result of the oil and gas equipment is obtained and automatically saved; And the life data analysis method combined with the characteristics of the equipment type and the failure mode is provided for analysis and comparison, the corresponding results and graphics are obtained and stored, and the corresponding data of the failure database is automatically obtained through system processing; If the number of equipment failure records is ≥5, the multi-distribution life data analysis method is selected, the data of the two levels of equipment and failure mode are selected, and the analysis sub-path is clear to analyze the data of the scene, the failure library data analysis result of the oil and gas equipment is obtained and saved through system processing, and the reliability index and curve of the oil and gas equipment are provided; (v) oil and gas equipment failure library version update When there is a new field failure record of the oil and gas equipment, steps (ii), (iii) and (iv) are repeated to iteratively update the failure library data of the equipment.

2. The method of claim 1, wherein: If no failure record, step (ⅳ) select OREDA algorithm for analysis.

3. The method for constructing a database of oil and gas equipment failures based on a plurality of algorithms of claim 1, wherein: If the number of failure records is 0 < number of records < 5, step (ⅳ) select Bayesian method based on OREDA prior for analysis, and apply Bayesian algorithm to analyze the path, according to the same type of equipment and failure mode data in OREDA database, determine the conditions and different paths of prior distribution as Gamma distribution or not as Gamma distribution, and select the corresponding path and algorithm for analysis.

4. The method for constructing a database of oil and gas equipment failures based on a plurality of algorithms of claim 1, wherein: If the number of failure records is ≥ 5, step (ⅳ) select multiple distribution life data analysis method.

Images