JP2012014619A - Computer system and risk diagnosis method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To estimate an accident occurrence position by calculating an accident occurrence probability of a pipeline in which an accident occurrence factor such as a defect is not specified.SOLUTION: A computer system includes: an operation history database for storing operation information of facilities; an accident history database for storing accident information that is information on accidents occurred in the facilities in the past; and a risk diagnosis server for estimating occurrence of accidents in the facilities. In the case where input of information on an occurrence position of a new accident and an occurrence factor of the new accident is received, the risk diagnosis server calculates a similarity between the information on the occurrence factor of the new accident and information on the occurrence factors of accidents that occurred in the past, calculates a similar accident occurrence probability of accidents to be incurred by factors similar to the occurrence factor of the new accident on the basis of the calculated similarity, specifies an estimated accident occurrence location where the occurrence of any accident caused by the factor similar to the occurrence factor of the new accident is estimated, and calculates an estimated accident occurrence probability in the estimated accident occurrence location on the basis of the similar accident occurrence probability.

Description

  The present invention relates to a computer system and a risk diagnosis method for managing a pipeline such as oil or natural gas. In particular, the present invention relates to a computer system and a risk management method for identifying a pipeline area where an accident is predicted to occur.

  In pipelines that transport oil or natural gas ranging from several hundred kilometers to several thousand kilometers, pipe defects occur due to changes in soil properties or weather. For example, abnormalities such as corrosion of pipes may occur due to changes in soil properties or weather, and serious accidents may occur at the abnormal locations. Therefore, it is important to investigate the state of defects in the pipeline and identify the dangerous section of the pipeline.

  The following methods are known as methods for acquiring the state of the pipeline.

  A sensor robot called a pig is inserted into the pipeline to investigate defects such as corrosion occurring in the pipeline. There is a method for determining how many years the pipe will break by comparing the state of the defect with the result of the investigation, calculating the speed and acceleration of defect growth, and determining the expansion of corrosion (for example, Patent Document 1).

  In the prior art, an ERF (Estimated Repeat Factor) is calculated based on information acquired by the pig. Normally, pipe replacement is recommended when the value of ERF is 1.0 to 1.2 or more.

  In addition, there is also known a method called hydrostatic test in which water is injected into a pipe to search for the possibility of pipe collapse (for example, see Non-Patent Document 1).

  The hydrostatic test is a method of injecting water when constructing a pipeline, applying a water pressure higher than the allowable pressure of the pipe, calculating the stress load applied to the pipe, and calculating the possibility of failure from the allowable pressure curve.

  Also, a method for constructing a reliability model based on historical data on pipeline damage and expert judgment is known (see, for example, Non-Patent Document 2). Non-Patent Document 2 shows a reliability model of an embedded pipeline. In particular, a concept called an accident tree is presented, and a range of damage is shown.

JP 2009-98762 A

Kiefner, John. F., Role of Hydrostatic Testing on Pipeline Integrity Assessment North Pipeline Integrity Workshop June 12, 2001 Cooke, R.M., et al Reliability Model For Underground Gas Pipelines, Probabilistic Safety Assessment and Management, Elsivier 2002

  A robot such as a pig acquires information on defects such as corrosion and dent that occur in a pipe using a sensor (ultrasonic wave, magnetic flux measurement, etc.) provided in the robot.

  However, since the method described in the cited document 1 depends on the accuracy of the sensor provided in the pig, it may not be possible to find a minute crack or the like.

  When stress corrosion cracking (SDC), which is a crack such as a laceration, has occurred, micro and longitudinal cracks of about several mm to cm may occur in groups. These fine cracks may not be captured because the sensor system provided by the pig is not sufficient. Actually, since the signal acquired by the sensor is analyzed by a person, the defect information such as the fine crack as described above may be missed.

  The cracks described above can lead to an explosion. Therefore, even when information is not acquired, it is necessary to calculate the possibility of occurrence of an explosion accident, a leakage accident, or the like.

  In addition, when using a method such as Hydro Static Test described in Non-Patent Document 1 in order to identify the location where an accident has occurred in a pipeline, it is an effective means for identifying the location where an accident has occurred, but it is also costly and frequently Difficult to do.

  Further, the method described in Non-Patent Document 2 is premised on the presence of a defect at the location where the accident occurred. However, accidents occur even in places where the existence of defects is unknown, such as the above-described minute crack group.

  Furthermore, depending on the seriousness of the defect, it may be difficult to predict the occurrence of an accident easily.

  The present invention has been made in view of the above-described problems, and an object thereof is to provide a method for calculating an accident occurrence probability and enabling risk analysis.

  A typical example of the present invention is as follows. That is, with reference to the operation history database that stores the operation information of the facility, the accident history database that stores the accident information that is information of the accident that has occurred in the past in the facility, the operation history database and the accident history database, A computer system comprising a risk diagnosis server that estimates the occurrence of an accident in the facility, the risk diagnosis server comprising a processor, a memory connected to the processor, and a network interface connected to the processor. The operation history database stores the operation information in which the position of the facility and information related to the facility are associated, and the accident history database includes the occurrence position of the accident that occurred in the past, and The above-mentioned matters that are associated with information on the causes of accidents that occurred in the past Information is stored, and when the risk diagnosis server receives input of information on the occurrence position of a new accident that is a newly occurred accident and information on the cause of the occurrence of the new accident, the risk diagnosis server refers to the accident history database, and A first similarity between information on the cause of occurrence of a new accident and information on the cause of occurrence of an accident that occurred in the past is calculated, and each occurrence factor of the new accident is calculated based on the calculated first similarity. The probability of occurrence of a similar accident of an accident caused by a factor similar to the cause of occurrence of the new accident is calculated, and the occurrence of an accident due to a factor similar to the cause of occurrence of the new accident is estimated with reference to the operation history database. The location where the accident occurred is identified, and the estimated probability of occurrence of the accident at the estimated location of the accident is calculated based on the probability of occurrence of similar accidents calculated for each new accident occurrence factor. And wherein the Rukoto.

  Even if information about the occurrence of an accident at a given facility cannot be obtained directly because the limit of the sensor is exceeded, an accident occurrence estimation site having an accident occurrence factor similar to the accident that occurred is estimated The probability of accident occurrence at the estimated location can be calculated. As a result, maintenance management in which the occurrence of an accident in the facility is predicted in advance becomes possible.

It is a block diagram explaining the structural example of the computer system in the 1st Embodiment of this invention. It is a block diagram explaining the hardware constitutions of the risk diagnosis system of the 1st Embodiment of this invention. It is a block diagram explaining the software structure of the risk diagnosis system of the 1st Embodiment of this invention. It is a flowchart explaining the outline | summary of the process which the risk diagnosis system of the 1st Embodiment of this invention calculates accident occurrence probability. It is explanatory drawing which shows the causal relationship of the accident occurrence factor in the 1st Embodiment of this invention. It is explanatory drawing which shows the causal relationship of the accident occurrence factor in the 1st Embodiment of this invention. It is explanatory drawing which shows the probability distribution of the pipe deterioration probability P (Defect) in the 1st Embodiment of this invention. It is explanatory drawing which shows the probability distribution of the pipe deterioration probability P (Configuration) in the 1st Embodiment of this invention. It is explanatory drawing which shows the probability distribution of the abnormality occurrence probability P (Soil) of the soil in the 1st Embodiment of this invention. It is explanatory drawing which shows the probability distribution of accident occurrence probability P (Pressure | Defect) in the 1st Embodiment of this invention. It is explanatory drawing which shows the probability distribution of the accident occurrence probability P (Pressure | Configuration) in the 1st Embodiment of this invention. It is explanatory drawing which shows the probability distribution of the failure probability P (CP) of CP installation in the 1st Embodiment of this invention. It is explanatory drawing which shows the probability distribution of the deterioration probability P (Coating) of the coating in the 1st Embodiment of this invention. It is explanatory drawing which shows the probability distribution of the failure probability P (CP | Soil) of CP installation in the 1st Embodiment of this invention. It is explanatory drawing which shows the probability distribution of the deterioration probability P (Coating | Soil) of the coating in the 1st Embodiment of this invention. It is explanatory drawing which shows the probability distribution of the failure probability P (CPMulf) of CP installation in the 1st Embodiment of this invention. It is explanatory drawing which shows the probability distribution of the deterioration probability P (CoatingMulf) of the coating in the 1st Embodiment of this invention. It is explanatory drawing which shows the relationship between the time passage in the valve equipment in the 1st Embodiment of this invention, or the welding part of a pipe, and accident occurrence probability. It is a flowchart explaining the detail of the process which the risk diagnosis system of the 1st Embodiment of this invention performs. It is a flowchart explaining the detail of the process which the risk diagnosis system of the 1st Embodiment of this invention performs. It is a flowchart explaining the detail of the process which the risk diagnosis system of the 1st Embodiment of this invention performs. It is a flowchart explaining the detail of the process which the risk diagnosis system of the 1st Embodiment of this invention performs. It is a flowchart explaining the detail of the process which the risk diagnosis system of the 1st Embodiment of this invention performs. It is explanatory drawing which shows an example of the pipeline area which has a factor similar to accident data in the 1st Embodiment of this invention. It is a figure explaining the application example of the risk diagnosis system of the 1st Embodiment of this invention. It is explanatory drawing which shows an example of the map data displayed on the map display part of the 1st Embodiment of this invention. It is explanatory drawing which shows an example of the result data displayed on the map display part of the 1st Embodiment of this invention. It is a flowchart explaining the update process of the weighting coefficient of the 1st Embodiment of this invention. It is explanatory drawing which shows an example of accident DB of the 1st Embodiment of this invention. It is explanatory drawing which shows an example of soil DB of the 1st Embodiment of this invention. It is explanatory drawing which shows an example of pipeline coordinate DB of the 1st Embodiment of this invention. It is explanatory drawing which shows an example of defect DB of the 1st Embodiment of this invention. It is explanatory drawing which shows an example of electrochemical DB of the 1st Embodiment of this invention. It is explanatory drawing which shows an example of the operation log | history DB of the 1st Embodiment of this invention. It is explanatory drawing which shows an example of probability distribution DB of the 1st Embodiment of this invention. It is explanatory drawing which shows an example of factor causal relationship DB of the 1st Embodiment of this invention.

  In the present invention, the probability of occurrence of an accident or failure similar to the accident or failure is estimated from information about the accident or failure that has occurred. Specifically, the position where the accident or failure is predicted to occur is specified by calculating the probability of occurrence of the accident or failure based on the data of the failure or accident that has occurred.

[First Embodiment]
Hereinafter, a computer system that analyzes the situation of an accident that has occurred and calculates the probability that a similar accident will occur will be described. The same applies to the failure.

  FIG. 1 is a block diagram illustrating a configuration example of a computer system according to the first embodiment of this invention. FIG. 1 shows a configuration example of a computer system in a natural gas pipeline.

  The computer system includes a pipeline 2201, compressors 2202 and 2203, cathode protection systems 2204 and 2205, a data center system 2213, a portable data input terminal 2214, and compressor stations 2215 and 2216.

  Natural gas is transported to the pipeline 2201 by the compressors 2202 and 2203.

  The cathode protection systems 2204 and 2205 obtain potential data, that is, cathode protection (CP) data in real time. The acquired CP data is transmitted to the risk diagnosis systems 2206 and 2207 via the gateway devices 2210 and 2211. Further, the acquired CP data is stored in the data center system 2213. Hereinafter, the cathode protection system is also referred to as CP equipment.

  The portable data input terminal 2214 is a device that inputs the location of the explosion at the site and information on the soil. The portable data input terminal 2214 acquires defect data measured by the pigging inspection, and transmits the acquired defect data to the compressor stations 2215 and 2216 by radio waves such as wireless communication.

  In an environment where wireless communication is not possible, a method of inputting data to the compressor stations 2215 and 2216 using a storage medium may be used.

  The compressor stations 2215 and 2216 are facilities that manage the pipeline 2201.

  The compressor station 2215 includes a risk diagnosis system 2206, a SCADA 2208, a gateway device 2210, and an offline data input system 2212. The compressor station 2216 includes a risk diagnosis system 2207, SCADA 2209, and a gateway device 2211.

  The risk diagnosis systems 2206 and 2207 are systems for calculating the accident occurrence probability of the pipeline 2201. The processing results of the risk diagnosis systems 2206 and 2207 are transmitted via the network 2217 to the data center system 2213.

  The configuration of the risk diagnosis system 2206 will be described later with reference to FIGS.

  The SCADA 2208 and 2209 collect data such as an acceptance pressure, a delivery pressure, a fluid acceptance temperature and a delivery temperature in the compressor 2203.

  The gateway devices 2210 and 2211 collect real-time data related to the pipeline 2201, such as in the case of real-time CP data. The gateway devices 2210 and 2211 transmit various data via the network 2217.

  The offline data input system 2212 is an interface device for capturing offline data acquired offline, such as inspection data and soil data acquired by a pig.

  The defect data of the pipeline 2201 is usually collected as offline data by pigging. The CP data may be collected in any data format of online data or offline data.

  The defect data of the pipeline 2201 is acquired using an ultrasonic wave or a magnetic sensor like a smart pig. The shape of the pipeline 2201 is acquired by a pig equipped with a gyro sensor.

  The data center system 2213 stores the accident occurrence probability calculated by the risk diagnosis systems 2206 and 2207. Further, the data center system 2213 stores shape data, defect data, CP potential data, and operation history data of the entire pipeline 2201.

  In the present embodiment, compressor stations 2215 and 2216 are arranged for each predetermined section. The compressor stations 2215 and 2216 manage a predetermined section of the pipeline. Thereby, even if it is the long pipeline 2201, it can divide | segment for every management area and can diagnose accident occurrence, Therefore A risk diagnosis can be performed efficiently in a short time. Further, by collecting the processing results of the compressor stations 2215 and 2216 in the data center system 2213, the overall situation can be easily grasped.

  In the case of an oil pipeline, the compressor 2202 is a turbine or a pump. The compressor stations 2215 and 2216 serve as pump stations.

  FIG. 2 is a block diagram illustrating a hardware configuration of the risk diagnosis system 2206 according to the first embodiment of this invention.

  The risk diagnosis system 2206 includes a processor 3101, a memory 3102, a nonvolatile storage medium 3103, and a network interface 3104.

  The processor 3101 executes a program developed on the memory 3102.

  The memory 3102 stores a program executed by the processor 3101 and information necessary for executing the program.

  The non-volatile storage medium 3103 stores programs and data developed on the memory 3102. As the nonvolatile storage medium 3103, for example, an HDD can be considered.

  The network interface 3104 is an interface for connecting to a network.

  FIG. 3 is a block diagram illustrating a software configuration of the risk diagnosis system 2206 according to the first embodiment of this invention.

  The risk diagnosis system 2206 includes an accident data input unit 101, an analysis instruction unit 102, an accident occurrence probability analysis unit 103, an accident data search unit 107, a map display unit 108, a soil data analysis unit 109, a shape analysis unit 111, and a defect analysis unit 113. A CP analysis unit 115, an operation history analysis unit 117, a weight coefficient update unit 119, a list output unit 120, and a map output unit 121.

  The risk diagnosis system 2206 includes an accident DB 122, a soil DB 123, a pipeline coordinate DB 124, a defect DB 125, an electrochemical DB 126, an operation history DB 127, a probability distribution DB 128, and a factor-causal relationship DB 129.

  The accident data input unit 101 inputs the accident data 130 to the accident occurrence probability analysis unit 103. The accident data 130 to be input includes the soil pH in the pipeline accident occurrence range, the coating state, and the like.

  The analysis instruction unit 102 instructs to search for a portion where an accident may occur.

  The accident occurrence probability analysis unit 103 calculates an accident occurrence probability in a predetermined range of the pipeline.

  Specifically, the accident occurrence probability analysis unit 103 calculates the similarity between the pipeline range in which no accident has occurred and the pipeline range in which the accident has occurred. In addition, the accident occurrence probability analysis unit 103 calculates an accident occurrence probability for each factor using pipeline operation data such as defect distribution, pipeline coordinates, soil pH, CP potential, pressure fluctuation, and temperature fluctuation.

  The accident occurrence probability analysis unit 103 includes a similarity analysis unit 104, a probability analysis unit 105, and a composite probability calculation unit 106.

  The similarity analysis unit 104 determines similarity between accident data related to accidents that have occurred in the past and accident data 130 related to the latest accident, and integrates data related to the accident by statistical processing.

  The probability analysis unit 105 calculates the accident occurrence probability using the parameter of the factor that caused the accident (accident occurrence factor parameter).

  The composite probability calculation unit 106 selects a pipeline section where there are a plurality of factors related to the occurrence of a pipe failure and a pipe accident. In addition, the composite probability calculation unit 106 acquires a weighting factor from the factor-causal relationship DB 129, and calculates an accident occurrence probability of an accident involving a plurality of accident occurrence factors using the acquired weighting factor.

  The accident data search unit 107 searches the accident DB 122 for accident data related to accidents that occurred in the past. The accident data search unit 107 stores the newly input accident data 130 in the accident DB 122.

  The map display unit 108 displays the result of risk diagnosis. A list representing the magnitude of risk, a map representing a portion with a large risk, etc. are displayed on the map display unit 108.

  The soil data analysis unit 109 searches the soil DB 123 for soil data such as the soil pH value of the pipeline in the specified range (pipeline start distance, end distance). The soil data analysis unit 109 outputs the search result to the similarity analysis unit 104 and the probability analysis unit 105.

  The shape analysis unit 111 acquires shape data related to the shape of the pipeline in the specified range from the pipeline coordinate DB 124. Further, the shape analysis unit 111 calculates a curvature value representing the bending of the pipe using the acquired shape data of the pipeline in the specified range. This is because the wrinkle shape generated at the bent portion of the pipe causes an accident.

  The defect analysis unit 113 acquires pipeline defect data in the specified section from the defect DB 125. In addition, the defect analysis unit 113 calculates a safety index such as ERF using the acquired defect information.

  The CP analysis unit 115 acquires CP potential data related to a cathode protection potential from the electrochemical DB 126. Further, the CP analysis unit 115 analyzes the abnormality of the potential (ON potential) using the acquired data on the cathode protection potential.

  The operation history analysis unit 117 acquires operation data indicating the operation history of the pipeline from the operation history DB 127. Further, the operation history analysis unit 117 adds up pressure fluctuations and temperature fluctuations.

  The weight coefficient updating unit 119 searches for a weight coefficient used when calculating a composite probability that is an accident occurrence probability related to a plurality of accident occurrence factors, and updates the weight coefficient based on the accident data 130.

  The weighting factor updating unit 119 searches for a flow relating to the relationship and dependency of the related accident occurrence factor when searching for the weighting factor, and further searches for a weighting factor based on the searched flow.

  The list output unit 120 outputs data for displaying the results of risk diagnosis as a list. Specifically, the list output unit 120 changes the order (sorting) based on the magnitude of the accident occurrence probability value, and creates a list of pipeline sections in descending order of the accident occurrence probability. In the present embodiment, the created list is tabular data.

  The map output unit 121 outputs data for displaying the result of risk diagnosis as a map. For example, when the accident occurrence probability value in a predetermined section is larger than a set threshold value, map data for displaying the section is output.

  The accident DB 122 stores information on the situation of pipelines of accidents that occurred in the past, that is, accident data at the time of occurrence of accidents of accidents that occurred in the past. For example, the accident DB 122 stores pipeline names, accident occurrence positions, pipe coating states, soil properties, and accident situations. Here, the situation of the accident is information indicating leakage, explosion, pipe crack occurrence, coating crack occurrence, and the like.

  The accident DB 122 stores accident data for the entire pipeline. The accident occurrence position is represented by a distance from a predetermined reference point. Hereinafter, a predetermined reference point is referred to as a base point.

  Details of the accident DB 122 will be described later with reference to FIG.

  The soil DB 123 stores soil data such as soil chemical characteristics such as soil pH in a predetermined section of the pipeline. For example, the soil DB 123 stores a distance from the base point and soil data corresponding to the distance from the base point. The soil DB 123 stores soil data for the entire pipeline range. In addition, about the detail of soil DB123, the detail is later mentioned using FIG.

  The pipeline coordinate DB 124 stores shape data related to the shape of the pipeline. The shape data may be two-dimensional or three-dimensional data. The pipeline coordinate DB 124 stores shape data of the entire pipeline range. Details of the pipeline coordinate DB 124 will be described later with reference to FIG.

  The defect DB 125 stores defect data which are parameters (occurrence position, length, width, depth, etc.) related to defects such as pipeline corrosion and cracks or dents. For example, the defect DB 125 stores a distance from the base point and defect data corresponding to the distance from the base point. The defect DB 125 stores defect data for the entire pipeline range. Details of the defect DB 125 will be described later with reference to FIG.

  The electrochemical DB 126 stores CP potential data such as CP potential. For example, the electrochemical DB 126 stores a distance from the base point and CP potential data corresponding to the distance from the base point. The electrochemical DB 126 stores CP potential data for the entire pipeline range. Details of the electrochemical DB 12 will be described later with reference to FIG.

  The operation history DB 127 stores operation data related to operation history of pipeline transportation (transport pressure, pipe outer wall temperature). For example, the operation history DB 127 stores a distance from the base point and an operation history corresponding to the distance from the base point. The operation history DB 127 stores operation data for the entire pipeline. Details of the operation history DB 127 will be described later with reference to FIG.

  The probability distribution DB 128 stores a probability distribution for calculating the accident occurrence probability for each accident occurrence factor. Details of the probability distribution DB 128 will be described later with reference to FIG.

  The causal relationship DB 129 stores the causal relationships of a plurality of accident occurrence factors. The factor causal relationship DB 129 also stores a weighting factor necessary for calculating the composite probability. Details of the cause-and-effect relationship DB 129 will be described later with reference to FIG.

  The accident data 130 is data relating to the investigation result of the accident that has occurred. The result data 131 is data displayed on the map display unit 108.

  Each of the DBs 122 to 129 described above is stored in a storage area of the nonvolatile recording medium 3103. In the present embodiment, the risk diagnosis system 2206 includes the DBs 122 to 129, but may be stored in another external storage system (not shown) or the like.

  Next, data stored in each database will be described.

  FIG. 28 is an explanatory diagram illustrating an example of the accident DB 122 according to the first embodiment of this invention.

  As shown in FIG. 28, in the accident DB 122 in this embodiment, description format data 2301 is stored as accident data.

  In the description format data 2301, the details of the accident are described between a line 2302 indicating the start of data and a line 2310 indicating the end of data. The description format data 2301 describes information acquired at the time of the accident.

  Specifically, an identification number 2303, an accident occurrence time 2304, a start distance 2305, an end distance 2306, a pH measurement value 2307, a coating state 2308, and an accident type 2309 are included.

  The identification number 2303 is a number for identifying accident data. It is an identifier given when stored in the accident DB 122. The accident occurrence time 2304 is the time when the accident occurred.

  The start distance 2305 represents the start point of the pipeline where the accident occurred. The end distance 2306 represents the end point of the pipeline where the accident occurred. Note that the start distance 2305 and the end distance 2306 may coincide.

  The measured pH value 2307 is the pH value of the soil at the accident occurrence point. The coating state 2308 is a coating state of a pipeline where an accident has occurred. From the coating state 2308, it can be inferred that the accident that occurred is due to an explosion.

  The accident type 2309 represents the type of accident that has occurred. In the example illustrated in FIG. 28, “Leak” representing leakage is stored.

  FIG. 29 is an explanatory diagram illustrating an example of the soil DB 123 according to the first embodiment of this invention.

  Soil data stored in the soil DB 123 is managed collectively for each survey period. In the present embodiment, the soil data 2402 corresponding to the survey time is managed by the metadata 2401.

  The metadata 2401 includes a pipeline name 2403, a measurement time 2404, and a file name 2405.

  The pipeline name 2403 is the name of the pipeline. The measurement time 2404 is a time when soil data is measured. A file name 2405 is a name of a file in which specific soil data 2402 is stored.

  In the present embodiment, the file name 2405 and the soil data 2402 are associated with each other, and the soil data 2402 can be acquired by referring to the file name 2405.

  The soil data 2402 includes a start distance 2406, an end distance 2407, a start distance error range 2408, and an end distance error range 2409.

  The start distance 2406 is the start distance of the pipeline. The end distance 2407 is the end distance of the pipeline. The start distance error range 2408 is an error in the start distance of the pipeline. The end distance error range 2409 is an error of the end distance of the pipeline.

  The soil data 2402 includes pH 2410, specific resistance, participation decay potential, and hydrogen sulfide presence index 2412 as measurement data.

  FIG. 30 is an explanatory diagram illustrating an example of the pipeline coordinate DB 124 according to the first embodiment of this invention.

  The pipeline coordinate DB 124 stores pipeline shape data in a data format 2501 as shown in FIG.

  The data format 2501 is variable-length data, and includes a data header 2502 and data parts 2503 and 2504.

  The data header 2502 includes a pipeline number and the number of figures. The number of figures indicates the number of figures constituting the pipeline.

  The data portion 2503 includes a coordinate number, a layer number, and a solid coordinate 2505. The number of coordinates represents the number of solid coordinates 2505. Data 1 (2503) and data 2 (2504) are stored in the number corresponding to the number of coordinates. The layer number is a number for identifying each figure. A solid coordinate 2505 is a solid coordinate (X, Y, Z).

  FIG. 31 is an explanatory diagram illustrating an example of the defect DB 125 according to the first embodiment of this invention.

  The defect data stored in the defect DB 125 is managed collectively for each inspection period. In the present embodiment, defect data 2602 corresponding to the inspection time is managed by the metadata 2601.

  The metadata 2601 includes a pipeline name 2603, an inspection time 2604, and a file name 2605.

  The pipeline name 2603 is the name of the pipeline. The inspection time 2604 is a time when the defect is inspected. A file name 2605 is a name of a file in which specific defect data 2602 is stored.

  In the present embodiment, the file name 2605 and the defect data 2602 are associated with each other, and the defect data 2602 can be acquired by referring to the file name 2605.

  The defect data 2602 includes a defect start distance 2606, a pipe defect position 2607, a defect length 2608, a defect circumferential start position 2609, a maximum depth 2610, a defect depth rate 2611, and a defect type 2612.

  The defect start distance 2606 is the position of the pipeline with the defect. Pipe defect location 2607 represents whether the defect exists inside or outside the pipe. Defect length 2608 represents the length of the defect. The defect start position 2609 in the circumferential direction represents the start position of the defect in the circumferential direction as time, taking the cross section of the pipe as a clock.

  Maximum depth 2610 represents the depth of the defect. The defect depth ratio 2611 is the ratio of the nominal pipe wall thickness to the maximum defect depth. The defect type 2612 is a type of defect such as metal loss or dent.

  FIG. 32 is an explanatory diagram illustrating an example of the electrochemical DB 126 according to the first embodiment of this invention.

  The CP electrical data stored in the electrochemical DB 126 is managed collectively for each measurement period. In the present embodiment, CP electrical data 2702 corresponding to the measurement time is managed by the metadata 2701.

  The metadata 2701 includes a pipeline name 2703, a measurement time 2704, and a file name 2705.

  The pipeline name 2703 is the name of the pipeline. The measurement time 2704 is the time when the potential is measured. A file name 2605 is a name of a file in which specific CP electrical data 2702 is stored.

  In the present embodiment, the file name 2705 and the CP electrical data 2702 are associated with each other, and the CP electrical data 2702 can be acquired by referring to the file name 2705.

  The CP electrical data 2702 includes a measurement point 2706 indicating a distance to the measurement point, a measurement point number 2707, a cathode potential measurement value (ON potential) 2708, a cathode potential measurement value (OFF potential) 2709, an anode potential, and an anode current. .

  FIG. 33 is an explanatory diagram illustrating an example of the operation history DB 127 according to the first embodiment of this invention.

  Data stored in the operation history DB 127 is managed collectively for each operation date and time. In this embodiment, operation data 2802 corresponding to the operation date and time is managed by metadata 2801.

  The metadata 2801 includes a pipeline name 2803, a date and time 2804, a compressor number 2805, and a file name 2806. The compressor number 2805 is an identifier for identifying the compressors 2202 and 2203.

  In the present embodiment, the file name 2806 and the operation data 2802 are associated with each other, and the operation data 2802 can be acquired by referring to the file name 2806.

  The operational data 2802 includes time 2807, delivery pressure 2808, delivery fluid temperature 2809, acceptance pressure 2810, and acceptance fluid temperature 2811. Time 2807 represents the time when the operational data was acquired.

  FIG. 34 is an explanatory diagram illustrating an example of the probability distribution DB 128 according to the first embodiment of this invention.

  The probability distribution DB 128 includes metadata 2901, a probability distribution library program 2902, and tabular data 2903.

  The probability distribution library program 2902 stores formulas (15) to (25) described later.

  When represented by an expression, the probability distribution library program 2902 is used, and when data cannot be approximated using an expression as shown in FIG. 10, tabular data is used.

  The metadata 2901 includes an index 2904, a probability distribution format 2905, and a number 2906.

  The index 2904 is an identifier for identifying information included in the metadata 2901. The probability distribution is searched by referring to the index 2904.

  The probability information format 2905 is information for identifying whether the probability distribution data is a program (Prog) or tabular data (Table).

  A number 2906 is a program or tabular data number. By referring to the number 2906, the corresponding probability distribution library program 2902 or the corresponding tabular data 2903 can be searched.

  The tabular data 2903 includes horizontal axis data (Horizontal) 2907 indicating dependent variables such as soil pH in FIG. 14 and vertical axis data (Vertical) 2908 indicating probability values. Furthermore, when a parametric variable is required, a column corresponding to the parametric variable is added to the tabular data 2903.

  FIG. 35 is an explanatory diagram illustrating an example of the cause-and-effect relationship DB 129 according to the first embodiment of this invention.

  The factor causal relationship DB 129 stores weight coefficient correspondence data 3001. The weighting coefficient correspondence data 3001 includes an ID 3002 and a value 3003.

  The ID 3002 is an identifier for identifying data stored in the weighting coefficient correspondence data 3001. A value 3003 is a value of a weighting coefficient.

  A data search method in each database will be described.

  As described above, each database in the present embodiment stores data related to pipeline management, and data such as soil, defects, and CP potential are managed in association with the distance from the base point.

  For example, the soil DB 123 stores data related to the soil properties of the section specified by the pipeline start distance and end distance. In addition, the defect DB 125 has the size (length, width, depth) of corrosion existing in the designated pipeline range (distance), and the electrochemical DB 126 has the designated pipeline range (distance). The value of CP potential is stored.

  That is, in the present embodiment, data stored in each database is searched using the distance from the base point as a search key.

  For example, the corrosion data of the pipeline is retrieved from the defect DB 125 based on the specified pipeline distance (start distance and end distance). As a method for realizing the above-described search, when defect data is stored in a relational database (RDB), a method using the SQL language that is a search language can be considered.

  Pipeline coordinates are composed of a three-dimensional pipeline shape coordinate sequence. Therefore, when the pipeline distance is designated, the coordinates within the designated distance are searched from the pipeline coordinate DB 124.

  FIG. 4 is a flowchart for explaining an overview of processing in which the risk diagnosis system 2206 according to the first embodiment of this invention calculates the accident occurrence probability.

  The risk diagnosis system 2206 analyzes the input accident data 130 (step 201). Specifically, the risk diagnosis system 2206 acquires soil data acquired at the accident occurrence point, pipeline shape data, defect data, CP potential data, and operation data from the input accident data 130. To do.

  Next, the risk diagnosis system 2206 refers to the accident DB 122 based on the analysis result of the accident data 130, and searches for accident data similar to the input accident data 130 (step 202).

  For example, when the accident data 130 is data related to a leakage accident, data related to the leakage accident is searched.

  The search accuracy can be improved by classifying into large-scale leak accidents and small-scale leak accidents. For example, leakage accidents can be classified by the amount of leakage or the extent of damage to the pipeline.

  The risk diagnosis system 2206 determines whether there is accident data similar to the input accident data 130 (step 203). That is, it is determined whether or not accident data of accidents that occurred in the past similar to the input accident data 130 is stored in the accident DB 122. Hereinafter, accident data of accidents that occurred in the past similar to the input accident data 130 are also referred to as similar accident data.

  If it is determined that there is no similar accident data, the risk diagnosis system 2206 ends the process.

  When it is determined that there is similar accident data, the risk diagnosis system 2206 determines, for each accident occurrence, whether there is similarity between the accident occurrence factor of the input accident data 130 and the accident occurrence factor of the similar accident data. . Further, when it is determined that there is a similar accident occurrence factor, the risk diagnosis system 2206 statistically integrates the similar accident occurrence factors and detects a factor parameter. Then, an analysis is performed for each factor parameter (step 204).

  The risk diagnosis system 2206 determines whether or not analysis has been performed for all factor parameters (step 205).

  If it is determined that analysis has been performed for all factor parameters, the risk diagnosis system 2206 proceeds to step 208.

  When it is determined that analysis has not been performed for all factor parameters, the risk diagnosis system 2206 searches the pipeline coordinate DB 124 for shape data similar to the pipeline shape data in the input accident data 130 ( Step 206). Here, the similarity means that the error of the factor parameter is small.

  Next, the risk diagnosis system 2206 calculates an accident occurrence probability based on each accident occurrence factor (step 207), and returns to step 205.

  The risk diagnosis system 2206 calculates the accident occurrence probability in consideration of all accident occurrence factors (step 208).

  When there are a plurality of accident occurrence factors in the pipeline section to be processed, the risk diagnosis system 2206 calculates an accident occurrence probability based on each accident occurrence factor.

  The risk diagnosis system 2206 displays a pipeline section in which the accident occurrence probability value calculated in step 208 is higher than a preset threshold value (step 209), and ends the process.

  Hereinafter, a specific method for calculating the accident occurrence probability from various data related to the pipeline will be described.

  Pipeline accidents include ruptures, explosions, and leaks. If there are micro cracks or pipe thinning, an accident will occur due to the progress of cracks and thinning over time.

  When an accident occurs, the situation of the accident that occurred is analyzed, and pipeline sections having similar accident occurrence factors are searched. However, when the existence of a serious defect is not known, it is difficult to analyze the cause of the accident, and it is difficult to determine the similarity using only one accident factor parameter.

  In the present invention, first, the risk diagnosis system 2206 searches for accidents that occurred in the past having a situation similar to a newly occurring accident. The risk diagnosis system 2206 analyzes a newly occurring accident based on the search result. Next, the risk diagnosis system 2206 detects a pipeline section having a situation similar to a newly occurred accident. Furthermore, the risk diagnosis system 2206 calculates the probability of occurrence of an accident in the pipeline section based on the analysis result of the newly occurring accident.

  The risk diagnosis system 2206 includes (1) defect distribution, (2) pipe bending, (3) CP potential, (4) soil hydrogen ion index (pH), (5) number of pressure fluctuations, and (6) temperature fluctuation. The situation at the accident occurrence point is grasped from the factor parameter. The situation of the accident occurrence point mentioned above uses the following parameter parameters, for example.

  (1) For the defect distribution, the ERF for each predetermined section of the pipeline is used as a factor parameter.

  (2) For pipe bending, the horizontal and vertical curvatures of the pipe are used as factor parameters.

  (3) For the CP potential, the time during which the base potential state is maintained is used as a factor parameter in a place where the ON potential of the CP is lower than the threshold value of −0.85 V.

  (4) For the soil hydrogen ion index (pH), the pH value of the soil is used as a factor parameter.

  (5) Regarding the number of pressure fluctuations, the number of fluctuations in the transportation pressure of the pipeline is used as a factor parameter. This is necessary because the transportation pressure in the pipeline fluctuates from moment to moment, and fluctuations in the transportation pressure cause accidents.

  (6) The temperature fluctuation uses the number of fluctuations in the temperature of the pipeline itself, which varies depending on the environment and the temperature of the fluid to be transported, as a factor parameter. This is necessary because the temperature of the pipeline varies depending on the environment or the temperature of the fluid to be transported, and the temperature variation causes an accident.

  Note that not only pipeline accidents that depend on one factor parameter may occur. Therefore, it is necessary to set conditions that take into account the relevance of each factor parameter. In the present embodiment, a conditional probability based on the condition is calculated as an accident occurrence probability.

  In this embodiment, the causal relationship of accident factors is defined in advance. The risk diagnosis system 2206 analyzes the accident according to the definition.

  FIG. 5A and FIG. 5B are explanatory diagrams illustrating the causal relationship of accident occurrence factors in the first embodiment of the present invention.

As shown in FIGS. 5A and 5B, the following ten factors are defined for the occurrence of an accident related to pipe damage in the present embodiment.
Factor 1: Pipe deterioration due to a defect → Pressure fluctuation → Pipe damage factor 2: Pipe deterioration due to a defect → Temperature fluctuation → Pipe damage factor 3: Pipe deterioration due to a defect → Pipe damage factor 4: Pipe bending → Pressure fluctuation → Pipe Damage factor 5: pipe bending → temperature fluctuation → pipe damage factor 6: pipe bending → pipe damage factor 7: soil pH abnormal value → CP equipment abnormality → pipe damage factor 8: soil pH abnormal value → coating deterioration → pipe damage factor 9: CP equipment abnormality → pipe damage factor 10: coating deterioration → pipe damage In this embodiment, the soil pH abnormal value is defined as abnormal when the pH value is lower than 7.0, that is, when it is acidic. .

  Factors 3 and 6 are pipe damage caused by stress concentration and bending moment concentration. Factors 9 and 10 are pipe damage caused by CP equipment abnormality due to failure such as equipment deterioration unrelated to the soil and coating deterioration caused by damage applied during construction.

  In the present embodiment, a causal relationship as shown in FIGS. 5A and 5B is defined, but it is also possible to add factors by adding a causal relationship.

  In the present embodiment, the accident analysis is executed based on the causal relationship shown in FIGS. 5A and 5B. In addition, the location of the accident is identified in consideration of the similarity with accidents that occurred in the past.

  For the accident analysis in the present embodiment, the following conditional probabilities are used.

  The conditional probabilities associated with pipe degradation (Factor 1, Factor 2) and pipe bending (Factor 4, Factor 5) due to defects such as corrosion are defined as follows:

  If there is a defect in the pipe, stress is applied to the defective part due to pressure fluctuation or temperature fluctuation and the pipe is damaged. In addition, the bent portion of the pipe may be in a wrinkle state (plastic deformation of the pipe due to the bending), and the pipe is damaged when pressure fluctuation or temperature fluctuation overlaps with the portion.

  In such a case, P1 (Damage) is calculated by the equation (1), where P1 (Damage) is the damage probability of the pipe due to defects and bending.

  Here, P (Defect) is the probability (factor 3) of pipe deterioration caused by a defect. P (Configuration) is a probability (factor 6) of pipe deterioration caused by bending of the pipe. Further, P (Pressure) is a pipe deterioration probability caused by pressure fluctuation, and is calculated by Expression (2). P (Temperature) is a pipe deterioration probability caused by temperature fluctuation, and is calculated by the equation (3).

  P (Pressure | Defect) is an accident occurrence probability (factor 1) caused by defects and pressure fluctuations. P (Pressure | Configuration) is an accident occurrence probability (factor 4) due to bending of the pipe and pressure fluctuation.

  P (Temperature | Defect) is an accident occurrence probability (factor 2) caused by defects and temperature fluctuations. P (Temperature | Configuration) is an accident occurrence probability (factor 5) caused by bending of the pipe and temperature fluctuation.

  Further, C1, C2, C3, C4, D1, D2, E1, and E2 are weighting coefficients, and satisfy Expression (4), Expression (5), and Expression (6).

  Conditional probabilities related to soil pH (Factor 7 and Factor 8) are defined as follows:

  When the acidity of the soil increases (when the pH value is 7 or less), the CP equipment becomes dysfunctional, and iron ions are dissolved from the potential difference between the pipe and the soil, so that the pipe is damaged.

  Also, when the soil acidity increases, the coating degrades and the pipe is damaged because the soil and the pipe are in direct contact.

  In such a case, if the damage probability of the pipe due to the soil abnormality is P2 (Damage), P2 (Damage) is calculated by Equation (7).

  P (CPMulf) is the CP equipment failure probability (factor 9) due to factors other than soil. P (CoatingMulf) is the deterioration probability of the coating due to factors other than soil, such as damage caused by a third party.

  P (CP) is a failure probability of the CP equipment, and is calculated by Expression (8). P (Coating) is the deterioration probability of the coating, and is calculated by equation (9).

  Here, P (Soil) is the probability of occurrence of abnormal soil due to abnormal soil pH value. P (CP | Soil) is a failure probability of the CP facility caused by the soil, that is, due to the factor 7. P (Coating | Soil) is a deterioration probability of the coating caused by the coating deterioration, that is, caused by the factor 8.

  F1, F2, G1, G2, H1, and H2 are weighting factors, and satisfy Expression (10), Expression (11), and Expression (12).

  Using each of the probabilities described above, the accident occurrence probability P (Damage) considering all factors is calculated by the equation (13).

  Here, the coefficients J1 and J2 satisfy Expression (14).

  In the present embodiment, the ID 3002 of the weight coefficient correspondence relationship data 3001 has the following correspondence.

  J1 ID 3002 is "1", J2 ID 3002 is "2", C1 ID 3002 is "3", C2 ID 3002 is "4", C3 ID 3002 is "5", C4 ID 3002 is "6", D1 ID3002 is "7", D2 ID3002 is "8", E1 ID3002 is "9", E2 ID3002 is "10", F1 ID3002 is "11", F2 ID3002 is "12", G1 ID3002 is “13”, G2 ID 3002 corresponds to “14”, H1 ID 3002 corresponds to “15”, and H2 ID 3002 corresponds to “16”.

  Next, each probability calculation method will be described.

  A plurality of models are known as methods for calculating probabilities such as P (Defect) and P (CP) described above. In the present embodiment, each probability is calculated using one of a plurality of models. Hereinafter, an example of the model will be described.

[1] Pipe Degradation Caused by Defect Factor FIG. 6 is an explanatory diagram showing a probability distribution of pipe degradation probability P (Defect) in the first embodiment of the present invention.

  The pipe deterioration probability P (Defect) has a probability distribution as shown by a curve 401 in FIG. In FIG. 6, the vertical axis represents the pipe deterioration probability P (Defect), and the horizontal axis represents the ERF (safety index). A curve 401 shown in FIG. 6 is expressed as in Expression (15).

Here, G ₁ is a shape parameter, and S ₁ is a scale parameter.

[2] Pipe degradation caused by bending of pipes FIG. 7 is an explanatory diagram showing a probability distribution of pipe degradation probability P (Configuration) in the first embodiment of the present invention.

  The pipe deterioration probability P (Configuration) has a probability distribution as shown by the curves 501 and 502 in FIG. In FIG. 7, the vertical axis represents the pipe deterioration probability P (Configuration), and the horizontal axis represents the pipe curvature. A curve 501 and a curve 502 illustrated in FIG. 7 are expressed as Expression (16).

Here, R is a curvature and R ₀ is an allowable limit. When R is equal to or less than R ₀ , the pipe deterioration probability P (Configuration) is defined as “0”.

  As shown in Expression (16), the shape parameter and the scale parameter are time-dependent parameters. Therefore, the time changes as shown by the curves 501 and 502 are made. This change corresponds to the deterioration of the pipe over time. The probability distribution may be divided for the horizontal direction and the vertical direction of the pipe.

[3] Danger due to soil pH FIG. 8 is an explanatory diagram showing the probability distribution of the abnormal occurrence probability P (Soil) of the soil in the first embodiment of the present invention.

  The pH value of the soil changes from season to season, and the degree of danger of the soil changes depending on the pH value. Therefore, the soil risk probability P (Soil) has a probability distribution as shown by the curves 601 and 602 in FIG. In FIG. 8, the vertical axis represents soil abnormality occurrence probability P (Soil), and the horizontal axis represents time (for example, month). A curve 601 and a curve 602 shown in FIG. 8 are expressed as in Expression (17).

  As shown in FIG. 8, the soil abnormality occurrence probability P (Soil) has a shape similar to the normal distribution.

[4] Pipe Degradation Generated Due to Pressure Fluctuation FIG. 9 is an explanatory diagram showing a probability distribution of accident occurrence probability P (Pressure | Defect) in the first embodiment of the present invention. FIG. 10 is an explanatory diagram showing a probability distribution of an accident occurrence probability P (Pressure | Configuration) in the first embodiment of the present invention.

  Accidents occur due to defects and pressure fluctuations, as well as pipeline bending and pressure fluctuations.

The accident occurrence probability P (Pressure | Defect) has a probability distribution as shown by a curve 702 and a curve 703 in FIG. In FIG. 9, the vertical axis represents the accident occurrence probability P (Pressure | Defect), and the horizontal axis represents the number of fluctuations N−N ₀ .

The accident occurrence probability P (Pressure | Configuration) has a probability distribution as shown by a curve 801 and a curve 802 in FIG. In FIG. 10, the vertical axis represents the accident occurrence probability P (Pressure | Configuration), and the horizontal axis represents the number of fluctuations N−N ₀ .

  A curve 701 and a curve 702 illustrated in FIG. 9 are expressed as in Expression (18). Further, the curve 801 and the curve 802 shown in FIG. 10 are expressed as in Expression (19).

Here, N ₀ represents the number of fluctuations at the allowable limit, and R ₀ represents the curvature at the allowable limit.

[5] Pipe deterioration caused by temperature fluctuation The accident occurrence probability P (Temperature | Defect) is the same as the probability distribution shown in FIG. 9, and the accident occurrence probability P (Temperature | Configuration) is the same as the probability distribution shown in FIG. Therefore, the description is omitted.

[6] CP Facility Failure FIG. 11 is an explanatory diagram showing a probability distribution of the failure probability P (CP) of the CP facility in the first embodiment of the present invention.

  With regard to the CP equipment, when the ON potential becomes −0.85 V with reference to copper-copper sulfate, the CP equipment fails. When time elapses after the potential as described above (CP equipment failure) occurs, the pipe is damaged.

The failure probability P (CP) has a probability distribution as shown by a curve 901 in FIG. In FIG. 11, the vertical axis represents failure probability P (CP), and the horizontal axis represents time t−t ₀ . A curve 901 shown in FIG. 11 is expressed as in Expression (20).

Here, t ₀ represents an allowable limit time.

[7] Coating Degradation FIG. 12 is an explanatory diagram showing the probability distribution of the coating degradation probability P (Coating) in the first embodiment of the present invention.

  If there is damage when the pipeline is coated, the coding will deteriorate due to the increased damage.

The coating deterioration probability P (Coating) has a probability distribution as shown by a curve 1001 in FIG. In FIG. 12, the vertical axis represents the coating deterioration probability P (Coating), and the horizontal axis represents time t−t ₀ .

  A curve 1001 shown in FIG. 12 is expressed as in Expression (21).

Here, t ₀ represents an allowable limit time.

[8] Failure of CP equipment due to soil pH FIG. 13 is an explanatory diagram showing the probability distribution of the failure probability P (CP | Soil) of the CP equipment in the first embodiment of the present invention.

  When the pH value of the soil is smaller than 7, an abnormality of the CP potential occurs due to an abnormality of the soil pH, which causes a failure of the CP equipment.

  The failure probability P (CP | Soil) of the CP facility has a probability distribution as shown by a curve 1101 in FIG. In FIG. 13, the vertical axis represents the failure probability P (CP | Soil), and the horizontal axis represents the soil pH value. A curve 1101 shown in FIG. 13 is expressed as in Expression (22).

[9] Coating Degradation due to Soil pH FIG. 14 is an explanatory diagram showing the probability distribution of the coating degradation probability P (Coating | Soil) in the first embodiment of the present invention.

  Soil abnormalities can cause abnormalities in pipe coating materials.

  The coating deterioration probability P (Coating | Soil) has a probability distribution as shown by a curve 1201 in FIG. In FIG. 14, the vertical axis represents the coating deterioration probability P (Coating | Soil), and the horizontal axis represents the soil pH value. A curve 1201 shown in FIG. 14 is expressed as in Expression (23).

[10] Pipe Damage Caused by Failure of CP Facility FIG. 15 is an explanatory diagram showing a probability distribution of a failure probability P (CPMulf) of the CP facility according to the first embodiment of the present invention.

  Failure of CP equipment due to factors other than soil will also damage the pipe.

The failure probability P (CPMulf) of the CP facility has a probability distribution as shown by a curve 1301 in FIG. In FIG. 15, the vertical axis represents the failure probability P (CPMulf) of the CP facility, and the horizontal axis represents time t−t ₀ . A curve 1301 shown in FIG. 15 is expressed as in Expression (24).

  As shown in FIG. 15, the failure of the CP facility does not immediately lead to damage of the pipe, but depends on the time elapsed since the occurrence of the failure of the CP facility.

[11] Pipe Damage Due to Coating Deterioration FIG. 16 is an explanatory diagram showing a probability distribution of a coating deterioration probability P (CoatingMulf) in the first embodiment of the present invention.

  Pipe deterioration is also caused by coating deterioration due to factors other than soil.

The coating deterioration probability P (CoatingMulf) has a probability distribution as shown by a curve 1401 in FIG. In FIG. 16, the vertical axis represents the coating deterioration probability P (CoatingMulf), and the horizontal axis represents time t−t ₀ . A curve 1401 shown in FIG. 16 is expressed as in Expression (25).

  As shown in FIG. 16, coating degradation does not immediately lead to pipe damage, but depends on the time elapsed since the occurrence of coding degradation.

  In addition to the above, when a leakage accident or the like is taken into consideration, the welded portion of the valve or pipe can be a cause of the accident. In fact, the welded part of the valve or pipe deteriorates with time.

  FIG. 17 is an explanatory diagram showing the relationship between the passage of time and the probability of occurrence of an accident in the valve equipment or the welded portion of the pipe in the first embodiment of the present invention.

  It can be assumed that the probability distribution in the welded part of the valve installation or pipe follows a bathtub-type curve as shown by curve 701 or curve 702 in FIG.

  As shown in FIG. 17, the failure rate is high due to a failure at the beginning of the valve installation, but the failure occurrence probability once decreases with the passage of time and then increases.

  The probability distributions shown in FIGS. 6 to 17 are stored in the probability distribution DB 128. The risk diagnosis system 2206 searches for a probability distribution based on the causal relationship of each factor shown in FIGS. 5A and 5B, and calculates an accident occurrence probability P (Damage) shown in Expression (13) using the probability distribution. .

  Next, specific processing of the risk diagnosis system 2206 will be described.

  18 to 22 are flowcharts illustrating details of processing executed by the risk diagnosis system 2206 according to the first embodiment of this invention.

  The risk diagnosis system 2206 searches for similar accident data that has occurred in the past based on the input failure and accident data. Further, the risk diagnosis system 2206 calculates the probability of occurrence of an accident in consideration of the common factors in the accident data, and searches for a pipeline range where a similar failure or accident is predicted based on the calculation result. To do.

  In the following description, a case where an accident occurs will be described, but the same processing is performed even when a failure occurs.

  When the risk diagnosis system 2206 receives a process start command from the analysis instruction unit 102, the risk diagnosis system 2206 acquires the accident data 130 from the accident data input unit 101 (step 1601). The acquired accident data 130 is input to the accident occurrence probability analysis unit 103.

  The accident data 130 includes the name of the location where the accident occurred, the name of the pipeline, the pipeline range (length) where the accident occurred, the soil pH at the location of the accident, information related to the time when the accident occurred, and the coating status Etc. are included.

  The pipeline range where the accident occurred represents a predetermined section of the pipeline including the accident occurrence point, and specifically represents a section specified by the start distance and end distance of the pipeline.

  The data related to the time when the accident occurred includes information such as year, month, and day.

  The coating state includes information obtained by quantifying the coating state. For example, “0” indicates that the coating state is normal, “1” indicates that the coating is deteriorated, and “2” indicates that the coating is damaged.

  The data included in the accident data 130 is an example, and other information may be included.

  The accident occurrence probability analysis unit 103 instructs the defect analysis unit 113 to search for defect data (step 1602).

  The defect analysis unit 113 that has received an instruction from the accident occurrence probability analysis unit 103 refers to the defect DB 125 and searches for defect data in the pipeline range where the accident occurred. Specifically, the defect analysis unit 113 searches for defect data in the range shown in Expression (26).

Here, L is the distance from the base point to the accident point of the pipeline. Further, L−α is the start distance, and L + α is the end distance. Α is a predetermined value.

  The defect analysis unit 113 analyzes the safety of the pipeline based on the retrieved defect data (step 1603).

  Specifically, the defect analysis unit 113 divides a pipeline range in which an accident has occurred, that is, a pipeline in the range shown in Expression (26) into fixed sections, and analyzes the distribution of defects existing in each section. It is not necessary to analyze the section where the pipe has been replaced or repaired.

  In the present embodiment, a risk value called ERF is used to analyze the safety of the pipeline. The ERF has a standard for each country, and in this embodiment, the standard of the country to which the present invention is applied is used. For example, a method of calculating the ERF using the formula (27) or the formula (28) is conceivable.

Here, d is the maximum corrosion depth [mm], t is the pipe wall thickness [mm], P is the allowable pressure [MPa], P _MAOP is the allowable pressure [MPa], F is the safety coefficient, and A is the Folia coefficient. To express.

  If there is a defect in the pipe, the pipe thickness t is reduced, which is different from the actual pipe thickness. Therefore, the ERF value when there is a defect is larger than the ERF value when there is no defect.

  In addition, pipe thinning may proceed over time, and it is necessary to predict changes in pipe thickness when an accident occurs. There are several ways to predict pipe thinning. For example, when defect data at different times is acquired at the same place, it can be predicted by substituting the acquired data into Equation (29).

Here, H _new is the latest defect data, and H _old is the previously acquired defect data. Δt is a time difference at which H _new and H _old are acquired. The defect data includes the length, width, and depth of the defect.

  When two pieces of defect data can be acquired, the defect change rate V as shown in the equation (29) can be calculated. In addition, when three or more pieces of defect data can be acquired, the acceleration of defect change can be calculated.

  On the other hand, when the above-described defect data over time cannot be obtained, a method of calculating the rate of defect change using equation (30) is conceivable.

  The defect analysis unit 113 refers to the defect DB 125 based on the calculated ERF value, and whether or not a defect having a predetermined threshold value (for example, ERF value is 1.0) or more exists in the range shown in Expression (31). Is determined (step 1604). The defect analysis unit 113 notifies the accident occurrence probability analysis unit 103 of the determination result. As the predetermined threshold, a value determined for each country to which the present invention is applied is used.

Here, m is a predetermined value and satisfies m> 0 and m> α.

  When it is determined that there is a defect in the range shown in Expression (31), the accident occurrence probability analysis unit 103 proceeds to Step 1607.

  When it is determined that there is no defect in the range shown in Expression (31), the accident occurrence probability analysis unit 103 calculates the similarity between the accident corresponding to the accident data 130 and the accident that has occurred in the past (step 1606). ). Specifically, the similarity analysis unit 104 calculates the similarity with the accident corresponding to the accident data 130 using the ERF value of the accident that has occurred in the past. For example, the following method is used for calculating the similarity.

The similarity analysis unit 104 calculates an ERF value for each interval, and obtains a maximum value ERF _new from the calculated ERF values.

  Next, the similarity analysis unit 104 instructs the accident data search unit 107 to search for accidents that have occurred in the past from the accident DB 122.

The similarity analysis unit 104 refers to the data of accidents that occurred in the past based on the search result, and obtains the maximum ERF value ERF _old using the method described above for accidents that occurred in the past. The similarity analysis unit 104 calculates S by substituting the calculated ERF _new and ERF _old into equation (32).

Here, when the value of m in equation (31) is decreased and the certain section range becomes smaller, the degree of coincidence decreases. Therefore, m in equation (31) is varied and S in each range is calculated.

  The similarity analysis unit 104 calculates the defect distribution similarity λ by substituting S calculated by the equation (32) into the equation (33).

Each item is multiplied by a coefficient C _i indicating importance as a weighting coefficient. Here, i corresponds to the variable section range.

  When the overlap of dangerous defect distributions is large, the similarity λ is a value close to “1.0”. In addition, when the overlap of dangerous defect distributions is small, the similarity λ is a value close to “0”.

Further, when the formula (34) is satisfied with respect to the predetermined values λ _th and ε, it can be determined that the accident corresponding to the accident data 130 is similar to the accidents that occurred in the past.

Note that λ _th may be an average value of λ calculated using accidents that have occurred in the past.

  The accident occurrence probability analysis unit 103 calculates an accident pipe deterioration probability P (Defect) corresponding to the accident data 130 (step 1607). Specifically, the probability analysis unit 105 calculates an accident pipe deterioration probability P (Defect) corresponding to the accident data 130.

  When it is determined in step 1604 that a dangerous defect exists, the probability analysis unit 105 searches the probability distribution curve shown in FIG. 6 with reference to the probability distribution DB 128, and the ERF value calculated in step 1603 and the searched probability. Using the distribution curve, the pipe deterioration probability P (Defect) of the accident corresponding to the accident data 130 is calculated.

  On the other hand, if it is determined in step 1604 that there is no dangerous defect, the probability analysis unit 105 cannot calculate the pipe deterioration probability P (Defect) using the ERF value. Accordingly, the probability analysis unit 105 uses the pipe deterioration probability P (Defect) of the accident that has occurred in the past and the calculated similarity λ, and the pipe deterioration probability P of the accident corresponding to the accident data 130 according to the equation (35). (Defect) is calculated.

  Next, the accident occurrence probability analysis unit 103 calculates an accident occurrence probability caused by the pipeline shape of the accident corresponding to the accident data 130, that is, a pipe deterioration probability P (Configuration). The larger the horizontal and vertical curvatures of the pipeline shape (the smaller the radius of curvature), the higher the probability of occurrence of an accident due to the influence of plastic deformation at the bent part.

  Therefore, the accident occurrence probability analysis unit 103 searches the pipeline shape data at the point where the accident corresponding to the accident data 130 occurs (step 1608).

  Specifically, the accident occurrence probability analysis unit 103 specifies the pipeline start distance and end distance included in the accident data 130 input from the accident data input unit 101, and shapes the acquisition of pipeline shape data. The analysis unit 112 is instructed.

  The shape analysis unit 112 that has received the command refers to the pipeline coordinate DB 124, searches for shape data indicated by three-dimensional coordinates indicating the pipeline shape at the accident occurrence point, and analyzes the searched shape data for an accident occurrence probability analysis. To the unit 103. In this embodiment, (X, Y) and (d, Z) are acquired as three-dimensional coordinates. Here, d represents the distance from the base point.

The accident occurrence probability analysis unit 103 calculates a curvature using the acquired shape data ((X, Y) and (d, Z)) (step 1609). Here, assuming that the curvature in the horizontal direction is R ₁ , the curvature R ₁ can be calculated using Expressions (36) to (39).

Here, k is a predetermined value. + K indicates k coordinates ahead of the currently referenced coordinates, and -k indicates k coordinates after the currently referenced coordinates.

Further, if the curvature of the vertical direction is R _2, the curvature R ₂ can be calculated by changing the X of formula (36) to (38) in d, the Y Z, the R ₁ to R _2.

  Since the similarity of the pipeline shape cannot be compared only by comparing the curvatures within a certain range, the accident occurrence probability analysis unit 103 thins out the coordinates and calculates the similarity regarding the pipeline shape in a wide area.

  The accident occurrence probability analysis unit 103 instructs the accident data search unit 107 to search for pipeline shape data for accidents that have occurred in the past (step 1610).

  The accident data search unit 107 that has received the instruction refers to the accident DB 122 based on the pipeline range (length) in which the accident included in the accident data 130, and obtains the shape data of the pipeline in the accident that has occurred in the past. Search for.

  The accident occurrence probability analysis unit 103 calculates the similarity between the pipeline shape in the accident that occurred in the past searched in step 1610 and the pipeline shape in the input accident data 130 (step 1611). Specifically, the similarity analysis unit 104 calculates the pipeline shape similarity. The similarity is calculated using, for example, the following method.

  The similarity analysis unit 104 calculates S using Expression (40).

Here, S indicates the similarity to the pipeline range acquired in step 1610. i = 1, 2, t ₁ represents the curvature in the horizontal direction, and R ₂ represents the curvature in the vertical direction. Also, t _{i old} represents the curvature of the pipeline shape in accidents that occurred in the past.

  Using the calculated S, the similarity analysis unit 104 calculates the pipeline shape similarity λ using Equation (41).

Here, j corresponds to the case where the coordinate thinning method is changed.

  The accident occurrence probability analysis unit 103 calculates an accident pipe deterioration probability P (Configuration) corresponding to the accident data 130 (step 1612). Specifically, the probability analysis unit 105 calculates an accident pipe deterioration probability P (Configuration) corresponding to the accident data 130.

  The probability analysis unit 105 searches the probability distribution curve shown in FIG. 7 with reference to the probability distribution DB 128, and uses the calculated pipe curvature and the searched probability distribution curve to determine the pipe deterioration of accidents that have occurred in the past. Probability P (Configuration) is calculated.

  Specifically, the probability analysis unit 105 uses the calculated curvature as an input to obtain the pipe deterioration probability P (Configuration) of accidents that have occurred in the past, using the curve shown in FIG. Here, as shown in FIG. 7, a curve corresponding to the service life of the pipe is defined according to the range, such as 25 to 40 years. Therefore, when the time T is input, a curve in a time range including T is selected, and further, the pipe deterioration probability P (Configuration) of accidents that have occurred in the past is calculated by inputting the curvature.

  The probability analysis unit 105 further calculates the pipe deterioration probability P (Configuration) of the accident corresponding to the accident data 130 as shown in the equation (42).

  By multiplying by λ, it is possible to calculate the probability considering the similarity with accidents that have occurred in the past. When λ is small, the similarity with accidents that have occurred in the past is not high, so the pipe deterioration probability P (Configuration) is a small value.

  Next, the accident occurrence probability analysis unit 103 has an accident occurrence probability due to an abnormality in the soil pH value of the accident corresponding to the accident data 130, that is, an abnormal occurrence probability P (Soil) of the soil in the accident corresponding to the accident data 130. Is calculated.

  First, the accident occurrence probability analysis unit 103 instructs the accident data search unit 107 to search for soil data related to accidents that have occurred in the past (step 1613).

  The accident data search unit 107 that has received the command stores the soil data included in the accident data 130 in the soil DB 123 and searches the soil DB 123 for soil data related to accidents that occurred in the past. The accident data search unit 107 searches the soil data based on the pipeline range (length) where the accident included in the accident data 130 occurs. The retrieved soil data is transmitted to the accident occurrence probability analysis unit 103.

  The soil data to be used may include soil pH, water content, ionization rate, and the like. In this embodiment, a case where pH is used will be described. In addition, application to water content or ionization rate is also possible easily.

  The accident occurrence probability analysis unit 103 uses the soil pH in the accident that occurred in the past searched in step 1613 to calculate the similarity of the soil pH between the accident corresponding to the accident data 130 and the accident that occurred in the past ( Step 1614). Specifically, the similarity analysis unit 104 calculates the similarity of soil pH between an accident corresponding to the accident data 130 and an accident that occurred in the past. The soil pH similarity λ is calculated using, for example, the following formulas (43) and (44).

Here, C is an arbitrary coefficient.

  The higher the similarity is, the closer λ becomes to “1.0”. In addition, when soil data is not acquired, the weighting coefficient of the probability of occurrence of an accident related to soil is “0” according to the value of S.

  The accident occurrence probability analysis unit 103 calculates the abnormality occurrence probability P (Soil) of the soil of the accident corresponding to the accident data 130 (step 1615). Specifically, the probability analysis unit 105 calculates an abnormality occurrence probability P (Soil) of the soil of the accident corresponding to the accident data 130.

  The probability analysis unit 105 searches the probability distribution as shown in FIG. 8 with reference to the probability distribution DB 128, and calculates the abnormal occurrence probability P (Soil) of the soil of the accident that has occurred in the past using the probability distribution. .

  Specifically, the probability analysis unit 105 uses the data relating to the time when the accident occurred included in the accident data 130 and the retrieved probability distribution, and the abnormal occurrence probability P ( Soil) is calculated. Note that the soil abnormality occurrence probability P (Soil) is calculated using a probability distribution curve corresponding to the area where the pipeline is laid.

  Further, as shown in Expression (45), the probability analysis unit 105 multiplies the accident occurrence probability P (Soil) of the soil of the accident that has occurred in the past by the similarity λ calculated in Step 1614, thereby calculating the accident data. An abnormal occurrence probability P (Soil) of the accident soil corresponding to 130 is calculated.

  Next, the accident occurrence probability analysis unit 103 determines whether or not the CP potential is abnormal (step 1616).

  The accident occurrence probability analyzing unit 103 designates the range shown in the equation (31) and determines whether or not there is a portion where the CP potential is lower than −0.85 V in the CP analyzing unit 115. The CP potential is based on a copper-copper sulfate reference. If the CP potential is lower than −0.85 V, it is determined that the CP potential is abnormal.

  When it is determined that there is no abnormality in the CP potential, the probability analysis unit 105 sets P (CPMulf) = 0, and the accident occurrence probability analysis unit 103 proceeds to step 1619.

  The location where the CP potential is determined to be abnormal, the accident occurrence probability analysis unit 103 searches for accidents that occurred in the past that have similar CP potentials, and the accident corresponding to the accident data 130 and the past searched The degree of similarity with the accident that occurred is calculated (step 1617).

  Specifically, the accident occurrence probability analysis unit 103 first obtains the time when the CP potential is lower than −0.85 V in the CP analysis unit 115 in a place where the CP potential is lower than −0.85 V. Command. When the CP equipment is repaired, the time when the CP potential becomes lower than −0.85 V after the repair is acquired.

The CP analysis unit 115 that has received the command refers to the electrochemical DB 126 and acquires the time when the CP potential is lower than −0.85V. Here, let the acquired time be t ₀ .

  Next, the accident occurrence probability analysis unit 103 instructs the accident data search unit 107 to search for accidents that occurred in the past that have similar CP potentials. The accident data search unit 107 that has received the command has occurred in the past with a similar CP potential based on the distance (L) of the accident occurrence point corresponding to the accident data 130 and the CP potential included in the accident data. Search for accidents.

  The similarity analysis unit 104 sets the similarity λ to “1” when there is accident data that has a similar CP potential in the past, and when there is no accident data that has a similar CP potential in the past. The similarity λ is set to “0”. Note that the similarity λ may be determined to be a continuous amount according to the CP potential value.

  The accident occurrence probability analysis unit 103 calculates the failure probability P (CPMulf) of the CP equipment of the accident corresponding to the accident data 130 (step 1618).

Specifically, the probability analysis unit 105 searches the probability distribution curve shown in FIG. 15 with reference to the probability distribution DB 128. The probability analysis unit 105 calculates the failure probability P (CPMulf) of the CP facility using the retrieved probability distribution curve and the acquired time t ₀ and the time t when the accident included in the accident data 130 occurs. To do. Further, the probability analysis unit 105 substitutes λ calculated in Step 1617 into the equation (46) to calculate the failure probability P (CPMulf) of the CP facility of the accident corresponding to the accident data 130.

  The accident occurrence probability analysis unit 103 refers to the coating state included in the accident data 130 and determines whether or not there is coating deterioration (step 1619).

  For example, when the state of the coating is expressed using a numerical value, if the numerical value is a numerical value indicating deterioration or damage, it is determined that there is coating deterioration.

  If it is determined that there is no coating deterioration, the accident occurrence probability analysis unit 103 proceeds to step 1622.

  If it is determined that there is coating deterioration, the accident occurrence probability analysis unit 103 instructs the accident data search unit 107 to search for information on the coating state in an accident that has occurred in the past (step 1620).

  The accident data search unit 107 that has received the command searches the accident DB 122 for the pipeline range of an accident that has occurred due to coating deterioration.

  The accident occurrence probability analysis unit 103 calculates an accident coating deterioration probability P (CoatingMulf) corresponding to the accident data 130 (step 1621).

  First, the similarity analysis unit 104 calculates a similarity. For example, let M be the number of accidents that occurred in the past due to coating deterioration, and among the accidents that occurred in the past, the number of accidents that occurred in the past with the same numerical value representing the coating state in the accident data 130 is N Then, the similarity analysis unit 104 can calculate the similarity λ by calculating M / N.

  The probability analysis unit 105 estimates the coating deterioration time. For example, when the pipe is replaced or repaired, the time difference up to the present is calculated from the time when the replacement or repair is performed.

  Then, the probability analysis unit 105 searches for a probability distribution curve as indicated by 1001 in FIG. The probability analysis unit 105 calculates a coating deterioration probability P (CoatingMulf) using the estimated coating deterioration time and the retrieved probability distribution curve. Further, the probability analysis unit 105 calculates an accident coating deterioration probability P (CoatingMulf) corresponding to the accident data 130 by multiplying the calculated coating deterioration probability P (CoatingMulf) by the similarity λ.

  The accident occurrence probability analysis unit 103 calculates an accident occurrence probability P (Pressure | Defect) due to defects and pressure fluctuations (step 1622). That is, the probability of occurrence of an accident due to factor 1 (deterioration of pipe due to defect → pressure fluctuation → pipe damage) is calculated.

  First, the probability analysis unit 105 calculates the number of changes. Here, the limit number of pressure fluctuations is used as the number of fluctuations, and can be obtained by the equation (47).

Here, I is a safety factor constant, C is a constant, N is the number of pressure fluctuations until breakage, q is a constant of 2.0 to 3.0, and S is a stress value. As the stress value S, a geometric average of the stress value in the axial direction and the stress value in the circumferential direction is used.

  Next, the accident occurrence probability analysis unit 103 instructs the operation history analysis unit 117 to search for a transport pressure record. The operation history analysis unit 117 that has received the command refers to the operation history DB 127 to search for a record relating to the transport pressure up to now, and transmits the search result to the accident occurrence probability analysis unit 103.

The accident occurrence probability analysis unit 103 uses the received data relating to the transport pressure up to _now to calculate the pressure fluctuation number N _{now in} which the pressure fluctuation fluctuates within a certain range.

The probability analysis unit 105 calculates N _diff by substituting N calculated by Equation (47) and N _now calculated into Equation (48).

The probability analysis unit 105 searches a probability distribution curve as shown in FIG. The probability analysis unit 105 calculates the accident occurrence probability P (Pressure | Defect) due to the factor 1 using the calculated N _diff and the retrieved probability distribution curve.

  As shown in FIG. 9, the accident occurrence probability P (Pressure | Defect) due to factor 1 is large at the beginning of pipeline installation, decreases when time passes, and increases again due to material deterioration after a certain period of time. .

  The accident occurrence probability analysis unit 103 calculates a conditional accident occurrence probability related to the pressure fluctuation (step 1623).

  Specifically, the probability analysis unit 105 multiplies P (Defect) calculated in Step 1607 by P (Pressure | Defect) calculated in Step 1622, thereby generating a conditional accident related to pressure fluctuation. The probability (P (Pressure | Defect) × P (Defect)) is calculated.

  The accident occurrence probability analysis unit 103 calculates an accident occurrence probability P (Pressure | Configuration) due to pipe bending and pressure fluctuation (step 1624). That is, the accident occurrence probability due to factor 4 is calculated.

As the number of fluctuations, the limit number of pressure fluctuations is used as in step 1622. The probability analysis unit 105 calculates N _diff by calculating Expression (48) in the same manner as in Step 1622.

The probability analysis unit 105 searches a probability distribution curve as shown in FIG. The probability analysis unit 105 calculates an accident occurrence probability P (Pressure | Configuration) due to the factor 4 using the calculated N _diff and the searched probability distribution curve.

  The accident probability P (Pressure | Configuration) due to factor 4 has a large probability at the beginning of pipeline installation as shown in FIG. 10, decreases with time, and again after a certain period of time, due to material deterioration. To increase.

  The accident occurrence probability analysis unit 103 calculates an accident occurrence probability related to pressure fluctuation (step 1625).

  Specifically, the probability analysis unit 105 multiplies P (Configuration) calculated in Step 1612 by P (Pressure | Configuration) calculated in Step 1624, thereby causing an accident occurrence probability related to pressure fluctuation. (P (Pressure | Configuration) × P (Configuration)) is calculated.

  The accident occurrence probability analysis unit 103 calculates an accident occurrence probability P (Temperature | Defect) due to defects and temperature fluctuations (step 1626). That is, the accident occurrence probability due to factor 2 is calculated.

  As the number of fluctuations, the limit number of temperature fluctuations is used. The limit number of temperature fluctuations can be obtained by the same method as in step 1622.

The probability analysis unit 105 calculates N using Equation (47). In the case of temperature fluctuation, the constant C is changed. Next, the probability analysis unit 105 calculates N _diff using Equation (48).

The probability analysis unit 105 searches a probability distribution curve as shown in FIG. The probability analysis unit 105 calculates the accident occurrence probability P (Temperature | Defect) due to the defect and the temperature variation using the calculated N _diff and the retrieved probability distribution curve.

  The accident occurrence probability analysis unit 103 calculates a conditional accident occurrence probability related to temperature fluctuation (step 1627).

  Specifically, the probability analysis unit 105 multiplies P (Defect) calculated in Step 1607 by P (Temperature | Defect) calculated in Step 1626, thereby causing a conditional accident related to temperature fluctuation. The occurrence probability (P (Temperature | Defect) × P (Defect)) is calculated.

  The accident occurrence probability analysis unit 103 calculates an accident occurrence probability P (Temperature | Configuration) caused by bending of the pipe and temperature fluctuation (step 1628). That is, the accident occurrence probability caused by the factor 5 is calculated.

The probability analysis unit 105 calculates N using equation (47) as in step 1626. Next, the probability analysis unit 105 calculates N _diff using Equation (48).

The probability analysis unit 105 searches a probability distribution curve as shown in FIG. The probability analysis unit 105 uses the calculated N _diff and the retrieved probability distribution curve to calculate an accident occurrence probability P (Temperature | Configuration) due to pipe bending and temperature fluctuation.

  The accident occurrence probability analysis unit 103 calculates a conditional accident occurrence probability related to temperature fluctuation (step 1629).

  Specifically, the probability analysis unit 105 multiplies P (Configuration) calculated in Step 1612 by P (Temperature | Configuration) calculated in Step 1628, thereby causing a conditional accident related to temperature fluctuation. Occurrence probability (P (Temperature | Configuration) × P (Configuration)) is calculated.

  The accident occurrence probability analysis unit 103 determines whether or not there is an abnormality in the CP facility (step 1630).

  If it is determined that there is no abnormality in the CP facility, the accident occurrence probability analysis unit 103 proceeds to step 1633.

  When it is determined that there is an abnormality in the CP facility, the accident occurrence probability analysis unit 103 calculates a failure probability P (CP | Soil) of the CP facility in which the soil causes the accident (step 1631). That is, the accident occurrence probability due to the factor 7 is calculated.

  Specifically, the probability analysis unit 105 searches the probability distribution DB 128 for a probability distribution curve as shown in FIG. The probability analysis unit 105 calculates the accident occurrence probability by specifying the soil pH value using the time when the acquired CP potential is lower than −0.85 V and the retrieved probability distribution curve. . The calculated accident occurrence probability is P (CP | Soil).

  The probability analysis unit 105 multiplies P (Soil) calculated in Step 1615 by P (CP | Soil) calculated in Step 1631 (Step 1632).

  The accident occurrence probability analysis unit 103 determines whether there is coating deterioration (step 1633).

  If it is determined that there is no coating deterioration, the accident occurrence probability analysis unit 103 proceeds to step 1636.

  When it is determined that there is coating deterioration, the accident occurrence probability analysis unit 103 calculates a coating deterioration probability P (Coating | Soil) caused by the coating deterioration (step 1634). That is, the accident occurrence probability due to the factor 8 is calculated.

  Specifically, the probability analysis unit 105 searches the probability distribution DB 128 for a probability distribution curve as shown in FIG. The probability analysis unit 105 calculates the accident occurrence probability by designating the retrieved probability distribution curve and the soil pH value. The calculated accident occurrence probability is P (Coating | Soil).

  The probability analysis unit 105 multiplies P (Soil) calculated in Step 1615 by P (Coating | Soil) calculated in Step 1634 (Step 1635).

  Next, the accident occurrence probability analysis unit 103 extracts a pipeline section having a factor similar to the accident data 130 in the pipeline managed by the risk diagnosis system 2206 (step 1636).

  Hereinafter, a pipeline section having factors similar to the accident data 130 is referred to as an accident occurrence estimation section.

  The following methods can be considered as methods for extracting the estimated accident occurrence section. First, the accident occurrence probability analysis unit 103 refers to each of the DBs 123 to 127 and calculates the similarity for each factor of the factors 1 to 10 in the accident data 130. The accident occurrence probability analysis unit 103 extracts a section where the calculated similarity is a predetermined value or more as an accident occurrence section.

  In addition, the calculation method of the similarity of each factor uses the same method (for example, refer to step 1607) when calculating the similarity between the accident data 130 and the accident that has occurred in the past.

  FIG. 23 is an explanatory diagram illustrating an example of a pipeline section having factors similar to the accident data 130 according to the first embodiment of this invention.

  The pipeline 1701 represents the pipeline managed by the risk diagnosis system 2206 as a straight line shape.

  As shown in FIG. 23, an accident occurrence estimation section is extracted for each factor. The accident occurrence probability of each factor can be calculated according to the distance of the pipeline.

  In the present embodiment, the composite probability calculation unit 106 calculates an accident occurrence probability in a pipeline section having a factor similar to the accident data 130 for each factor. Further, the composite probability calculation unit 106 extracts a pipeline section having an accident occurrence probability equal to or higher than a preset threshold as an accident occurrence section.

  In the present embodiment, the accident occurrence estimation section is managed by the start distance and end distance of the pipeline.

  In the example shown in FIG. 23, the sections 1703 and 1704 are the factor 1 accident occurrence estimation sections. Sections 1705 and 1706 are accident occurrence estimation sections of factor 2. Sections 1707 to 1709 are accident occurrence estimation sections of factor 3. Sections 1710 and 1711 are accident occurrence estimation sections of factor 4. Sections 1712 and 1713 are accident occurrence estimation sections of factor 5. Sections 1714 to 1716 are factor 6 accident occurrence estimation sections. A section 1717 is an accident occurrence estimation section of factor 7. An interval 1718 is an accident occurrence estimation interval of factor 8. Sections 1719 and 1720 are accident occurrence estimated sections of factor 9. A section 1721 is an accident occurrence estimation section of the factor 10.

At this time, the cause of the accident overlaps in the following sections.
Overlapping section 1 (1722): section 1703, section 1705, section 1707, section 1721
Overlapping section 2 (1723): section 1708, section 1710, section 1714, section 1719
Overlapping section 3 (1724): section 1704, section 1713
Overlapping section 4 (1725): section 1706, section 1711, section 1715
That is, it is determined that the overlapping section 1 (1722), the overlapping section 2 (1723), the overlapping section 3 (1724), and the overlapping section 4 (1725) are pipeline sections having a plurality of factors.

  The accident occurrence probability analysis unit 103 extracts the start distance and the end distance of the pipeline sections where the factors overlap as overlapping sections.

  The accident occurrence probability analysis unit 103 determines whether there is a pipeline section having a plurality of factors (step 1637).

  If it is determined that there is no pipeline section having a plurality of factors, the accident occurrence probability analysis unit 103 proceeds to step 1642.

  When it is determined that there is a pipeline section having a plurality of factors, the accident occurrence probability analysis unit 103 searches for a weighting coefficient (step 1638).

  Specifically, the accident occurrence probability analysis unit 103 instructs the weighting factor update unit 119 to search for a weighting factor. The weighting factor updating unit 119 that has received the command acquires the weighting factor from the cause-and-effect relationship DB 129 and transmits the acquired weighting factor to the accident occurrence probability analyzing unit 103.

  In the present embodiment, the weighting factor updating unit 119 reads out the weighting factor of each factor using the ID 3002 as a search key, and transmits the read weighting factor of each factor to the accident occurrence probability analyzing unit 103.

  The retrieved weight coefficient is updated based on the analysis result of the accident data 130 in the next step 1639.

  For example, if the soil pH value at the accident occurrence point is high, the weighting coefficient related to the accident occurrence probability of the accident induced by the soil pH value increases, and there are many places where the pipe is bent at the accident occurrence point. The weighting coefficient related to the accident occurrence probability related to the pipeline shape becomes large.

  The accident occurrence probability analysis unit 103 reads the weighting factor of each factor from the factor-causal relationship DB 129, and updates the read weighting factor based on the analysis result of the accident data 130 (step 1639).

  The weighting factors in the present embodiment include C1 to C4, D1 to D2, E1 to E2, F1 to F2, G1 to G2, H1 to H2, and J1 to J2. Each weight coefficient satisfies the conditions shown in the equations (4) to (6), (10) to (12), and (14).

  In addition, when using the equations (1) to (3) and the equations (7) to (9), the pipe damage probability due to defects and bends is expressed as P1 (Damage) and the pipe damage probability due to soil abnormality. P2 (Damage) is expressed by the equations (49) and (50).

  Here, C1 to C4 are weighting factors related to the factors 1 to 6. D1 to D2 are weighting factors related to factor 1 and factor 4. E1 to E2 are weighting factors related to the factor 2 and the factor 5. F1 to F2 are weighting factors related to the factors 7 to 10, and G1 to G2 are weighting factors related to the factor 7 and the factor 9. H1 to H2 are weighting factors related to the factor 8 and the factor 10. J1 to J2 are weighting factors related to the factors 1 to 10.

  The weighting factor is updated by actually examining the pipe. Hereinafter, a specific example will be described. The case where the accident corresponding to the accident data 130 is caused by an accident due to a combination of an abnormality in the CP potential and a defect in the pipe will be described. It is also assumed that the service life of the pipe where the accident occurred has been exceeded.

  At this time, when it is determined that the corresponding years of the pipes in the pipeline in the overlapping section are N years later, the weighting factors related to the pressure fluctuation are D1 and D2, which satisfy the condition of Expression (5).

  Since the presence of a defect is recognized around the point where the accident corresponding to the accident data 130 occurs, the frequency of the accident generated due to the defect and the frequency of the accident generated due to the bending of the pipe are updated.

  That is, when the accident occurrence frequency caused by the defect is DN1, and the accident occurrence frequency caused by the bending of the pipe is DN2, the equation (51) and the equation (52) are updated.

  Further, when the bending of the pipe due to the defect is also related, DN2 becomes DN2 + 1, and DN1 and DN2 are updated by Expression (53) and Expression (54), respectively.

  For accidents caused by other factors as well, the weighting coefficient is updated using mathematical formulas. In this case, the frequencies of the factors related to C1-C4, E1-E2, F1-F2, G1-G2, H1-H2, J1-J2, CN1-CN4, EN1-EN2, FN1-FN4, GN1-GN2, respectively. , HN1 to HN2, and JN1 to JN2.

  E <b> 1 to E <b> 2 are weighting factors in the probability of occurrence of an accident due to defects, pipeline bending, and temperature fluctuations. G1 to G2 are weighting factors in the probability of occurrence of an accident due to soil pH and CP facility abnormality. H1 and H2 are weighting factors in the probability of occurrence of an accident due to soil pH and coating deterioration.

C1-C4, F1-F2, and J1-J2 are considered as follows.
C1: When the defect corresponding to the accident data 130 and the cause of the defect and pressure fluctuation, and the pipeline shape and pressure fluctuation cannot be ignored, the frequency CN1 is added.
C2: When the defect corresponding to the accident data 130 and the cause of the defect and the temperature variation, and the pipeline shape and the temperature variation cannot be ignored, the frequency CN2 is added.
C3: CN3 is added when the cause of the defect alone cannot be ignored for the accident corresponding to the accident data 130.
C4: CN4 is added when the pipeline bends due to the accident corresponding to the accident data 130 cannot be ignored.
F1: FN1 is added when the soil pH and the cause of CP equipment abnormality cannot be ignored for the accident corresponding to the accident data 130.
F2: FN2 is added when the soil pH and the cause of coating deterioration cannot be ignored for the accident corresponding to the accident data 130.
J1: JN1 is added when the factors related to defects and pipeline bending cannot be ignored.
J2: JN2 is added when factors related to soil, CP equipment abnormality and coating deterioration cannot be ignored.

  As the number of related factors increases, the change in the weighting coefficient decreases, but as the frequency of the factors increases, a significant value is obtained. In addition, even if no accident has actually occurred, when pipe cracking occurs, repair, pipe replacement, CP equipment inspection, CP potential measurement, coating replacement, etc. In addition, when an abnormality is found, accident data may be registered as an accident has occurred.

  The accident occurrence probability analysis unit 103 transmits the updated weight coefficient to the weight coefficient update unit 119. The weighting factor updating unit 119 stores the received weighting factor in the cause-and-effect relationship DB 129.

  The accident occurrence probability analysis unit 103 calculates a normalization coefficient (step 1640).

  This is because the accident occurrence probability may be greater than 1 when the weighting factor updated so as to satisfy the expressions (4) to (6), (10) to (12), and (14) is used. A normalization coefficient for normalizing the weighting coefficient is calculated.

  When the normalization coefficient is Nor, the composite probability calculation unit 106 calculates the normalization coefficient Nor using Expression (55).

Here, P _j is an accident occurrence probability calculated for each combination of factors 1 to 10.

  The accident occurrence probability analysis unit 103 calculates an accident occurrence probability considering a plurality of factors in the overlapping section (step 1641).

  Specifically, the accident occurrence probability analysis unit 103 is configured such that the composite probability calculation unit 106 uses the updated weighting coefficients and the calculated normalization coefficients as equations (1) to (3) and equations (7) to (9). ) And calculate the accident probability P (Damage) considering all factors.

  If it is determined in step 1637 that there is no pipeline section having a plurality of factors, the probability analysis unit 105 calculates the accident occurrence probability of the section not related to the weighting factor (step 1642), and the process proceeds to step 1643. That is, the accident occurrence probability for each factor is calculated.

  The accident occurrence probability analyzing unit 103 rearranges (sorts) the calculated accident occurrence probability values in descending order, and generates output information to be displayed on the map display unit 108 (step 1643). The generated information includes, for example, map data and tabular data.

  The map output unit 121 displays map data based on the generated output information (step 1644). Specifically, the position where the accident is predicted to occur on the map is displayed on the map display unit 108 in descending order of the accident occurrence probability.

  The map data displayed on the map display unit 108 will be described later with reference to FIG.

  The list output unit 120 displays tabular result data 131 on the map display unit 108 based on the generated output information (step 1645). Accordingly, it is possible to confirm a pipeline range where an accident similar to the accident data 130 occurs.

  The result data 131 displayed on the map display unit 108 will be described later with reference to FIG.

  By the above process, the accident occurrence probability analysis unit 103 ends the process.

  FIG. 25 is an explanatory diagram illustrating an example of map data displayed on the map display unit 108 according to the first embodiment of this invention.

  The map display unit 108 displays a pipeline shape 1901 and a section 1902 where an accident may occur.

  Pipeline shape 1901 represents the shape of a pipeline where an accident may occur. A section 1902 in which an accident may occur represents a section of a pipeline in which the accident occurrence probability is equal to or higher than a predetermined threshold.

  When an input / output device such as a mouse and a keyboard provided in the map display unit 108 is used to align a pointer 1904 displayed on the map display unit 108 with a section 1902 where an accident may occur, accident occurrence probability information 1903 is obtained. Is displayed.

  The accident occurrence probability information 1903 includes an accident occurrence probability P (Damage) considering all factors and an accident occurrence probability for each factor.

  In the example shown in FIG. 25, the accident occurrence probability P (Damage) considering all the factors is “0.58”, the accident occurrence probability due to the factor 1 is “0.87”, and the accident occurrence probability due to the factor 8 is “0”. .45 ".

  Various display contents of the accident occurrence probability information 1903 can be considered. For example, an accident occurrence probability P (Damage) considering all factors and a method of displaying the accident occurrence probabilities of all related factors can be considered. Alternatively, a method of displaying the accident occurrence probability P (Damage) in consideration of all the factors and the accident occurrence probability of a highly relevant factor may be used.

  For example, a factor having a high relevance may be a factor having a probability value equal to or higher than a predetermined value in the accident occurrence probability P (Damage), or a factor having a weighting factor equal to or higher than a predetermined value. This makes it possible to specify a dominant factor in the accident occurrence probability P (Damage).

  FIG. 26 is an explanatory diagram illustrating an example of the result data 131 displayed on the map display unit 108 according to the first embodiment of this invention.

  Result data 131 includes ID 2000, start distance 2001, end distance 2002, accident occurrence probability 2003, factor 1 (2004), factor 2 (2005), factor 3 (2006), factor 4 (2007), factor 5 (2008), Factor 6 (2009), Factor 7 (2010), Factor 8 (2011), Factor 9 (2012) and Factor 10 (2013) are included.

  ID 2000 is an identifier for identifying a pipeline section in which an accident may occur. The start distance 2001 is the start distance of the pipeline corresponding to ID2000. The end distance 2002 is the end distance of the pipeline corresponding to ID2000.

  The accident occurrence probability 2003 is an accident occurrence probability P (Damage) considering all factors.

  Factor 1 (2004) is the probability of an accident occurring due to factor 1. Factor 2 (2005) is the probability of an accident occurring due to factor 2. Factor 3 (2006) is the probability of an accident occurring due to factor 3. Factor 4 (2007) is the probability of an accident occurring due to factor 4. Factor 5 (2008) is the probability of an accident occurring due to factor 5. Factor 6 (2009) is an accident occurrence probability caused by factor 6. Factor 7 (2010) is the probability of an accident occurring due to factor 7. Factor 8 (2011) is the probability of an accident occurring due to factor 8. Factor 9 (2012) is the probability of an accident occurring due to factor 9. Factor 10 (2013) is the probability of an accident occurring due to factor 10.

  Hereinafter, a specific example will be described.

  FIG. 24 is a diagram illustrating an application example of the risk diagnosis system 2206 according to the first embodiment of this invention.

As an application scene of the present invention, for example, the following cases can be considered.
(1) As a result of investigating the pipe, there was no serious defect (for example, a defect with an ERF of 1.2 or more) on the pipeline, but a serious defect occurred at a remote location. If.
(2) When there is a bend of about 45 degrees in the pipeline.
(3) When the operation data was examined, the temperature fluctuation was small, but the service life of the pipeline was long, and the pressure fluctuation reached the limit value N.
(4) When the soil data is acquired, the soil pH shows 7 or less throughout the year, and the coating is found to be damaged.

  In the case as described above, the accident occurrence probability P (Damage) in consideration of each factor is calculated.

  The accident occurrence pipeline 1801 is an accident pipeline corresponding to the accident data 130.

  The soil pH 1803 is information on the soil pH around the accident occurrence pipeline 1801. The soil pH 1803 is pH = 2.0.

  A new accident occurrence section 1805 is a range of an accident caused by a pipeline crack.

  The pipeline shape 1806 is information related to the pipeline shape of the accident occurrence pipeline 1801. A hatched portion of the pipeline shape 1806 represents a bent portion of the pipe.

  The defect portion 1808 is information regarding the distribution of defects in the accident occurrence pipeline 1801. Here, in the defect portion 1808, the vertical axis represents the cross section of the pipe. In the example shown in FIG. 24, it is sectional drawing of a pipe at the time of making a perpendicular direction with respect to the ground into 12:00. The horizontal axis of the defective portion 1808 represents the pipe distance.

  The coating deterioration 1810 is information related to the pH range of the accident occurrence pipeline 1801.

  The similar pipeline 1802 is a pipeline range having factors similar to the accident corresponding to the accident data 130.

  The soil pH 1804 is information regarding the soil pH around the similar pipeline 1802. The soil pH 1804 is pH = 2.5.

  The pipeline shape 1807 is information on the pipeline shape of the similar pipeline 1802. The defect portion 1809 is information regarding the distribution of defects in the similar pipeline 1802. The coating deterioration 1811 is information regarding the coating deterioration of the similar pipeline 1802.

  In the example shown in FIG. 24, the similarity λ of the pipeline shape is calculated as “0.92” from the equation (41), and the similarity λ of the defect distribution is calculated as “0.75” from the equation (33). . In this embodiment, it is determined that there is similarity for any factor.

  In this case, the damage probability P1 (Damage) of the pipe due to the defect and the bending is calculated as P1 (Damage) = 0.87 using the equation (1).

  Further, the similarity λ of the soil pH is calculated as “0.80” from the equation (44). In this case, P2 (Damage) is calculated as P2 (Damage) = 0.45 using Equation (7) as the pipe damage probability due to soil abnormality.

  Further, when J1 = 0.74 and J2 = 0.26, the accident occurrence probability P (Damage) considering all the factors is calculated using Expression (13). At this time, the normalization coefficient is calculated as shown in Expression (56).

  As described above, P (Damage) = 0.58 is calculated by multiplying P (Damage) calculated using Expression (7) by the normalization coefficient calculated using Expression (56). .

  In addition, the value read from factor causal relationship DB129 is used for each weighting coefficient.

  As described above, the risk of occurrence of a similar accident can be estimated based on the accident occurrence probability calculated from the accident data 130.

  The weighting factor may be updated when a pipe survey is performed. At this time, the accident probability curve and the weighting coefficient are updated. For example, if a pipe crack is found during pipe inspection, if an abnormality is found during pipe repair or pipe replacement, CP equipment inspection, CP potential measurement, or coating replacement, You may register as accident data 130. In this case, a process of updating only the weighting coefficient may be executed separately from the estimation of the similar accident.

  FIG. 27 is a flowchart for describing weight coefficient update processing according to the first embodiment of this invention.

  The accident occurrence probability analysis unit 103 acquires the accident data 130 (step 2101). This process is the same as step 1601.

  The accident data 130 includes information such as soil data, pipeline shape, defect data, CP potential and coating damage.

  The accident occurrence probability analysis unit 103 updates the weighting coefficient based on the acquired accident data 130 (step 2102). The method for updating the weighting coefficient is the same as that in step 1639.

  The accident occurrence probability analysis unit 103 transmits the updated weight coefficient to the weight coefficient update unit 119.

  The weighting factor updating unit 119 that has received the updated weighting factor reflects the updating result of the related weighting factor based on the causal relationship as shown in FIG. 3 (step 2103), and ends the process.

[Modification]
In the modification, an accident such as a gas or oil leak will be described.

  Since the configuration of the computer system and the risk diagnosis system 2206 is the same as that of the first embodiment, the description thereof is omitted.

  In the case of an accident such as a gas or oil leak, the stress resulting from the transport pressure is concentrated on the defective portion of the pipeline, and the location where the stress is concentrated ruptures, causing an accident. Also, accidents may occur due to aging of valves or welding.

  Therefore, for accidents such as gas or oil leaks, it is necessary to take into account the factors related to equipment abnormalities as described above.

  Therefore, in the modified example, the accident occurrence probability is calculated in consideration of the factor of equipment abnormality in addition to the defect, the CP potential, and the operation history.

  The failure probability of the valve or welding is represented by a bathtub curve as shown in FIG. In such a case, the probability of occurrence of an accident is calculated by paying attention not only to the shape of the pipe but also to the position of the equipment.

Hereinafter, a case where the probability of occurrence of a gas leakage accident is obtained will be described. In the modified example, the following causal relationship is considered.
Factor 1: Valve deterioration-Pipe damage factor 2: Weld deterioration-Pipe damage factor 3: Defect occurrence-Pipe damage The probability of occurrence of an accident related to valve deterioration or welding deterioration is calculated using the probability distribution curve shown in FIG. The Further, the accident occurrence probability P (Leak) due to the defect can be calculated using the equation (57).

Here, K1, K2, and K3 are weighting factors, and satisfy Expression (58). In the present embodiment, K1 is associated with the ID3002 of “17”, K2 is associated with the ID3002 of “18”, and K3 is associated with the ID3002 of “19” and stored in the cause-and-effect relationship DB 129.

  Here, P (Valve) is a probability that the valve is deteriorated or damaged. As shown in FIG. 17, P (Valve) is expressed as a function of time. The probability value of damage is high immediately after the valve is installed, but the probability of damage decreases as time passes. However, as time passes further, deterioration proceeds due to factors such as wear when the fluid passes.

  P (Weld) is the probability that the weld will deteriorate. P (Leak | Valve) is the probability that leakage will occur where there is a valve. P (Leak | Weld) is the probability that leakage will occur where there is welding. P (Defect) is the probability that a leak will occur where there is a defect.

  Here, P (Leak | Valve), P (Leak | Weld), and P (Defect) can be calculated as follows.

  The accident occurrence probability analysis unit 103 refers to the accident DB 122 and searches for records of leakage accidents.

  Specifically, the accident occurrence probability analysis unit 103 obtains the number N of accident data whose accident type (Type) 2309 of the description format data 2301 as shown in FIG. 28 is “Leak”, and is further described as Cause. The numbers Q, R, and S of the row data described as “Valve” or “Weld” in the column are obtained. M indicates the number of times that leakage has occurred due to deterioration or damage of the valve, and L indicates the number of times that leakage has occurred due to deterioration or damage of the weld.

  Further, the accident occurrence probability analysis unit 103 substitutes the calculated N, Q, R, and S into the equations (59) to (61), so that P (Leak | Valve), P (Leak | Weld), and P (Defect) is calculated.

  In addition, about weighting factors K1, K2, and K3, accuracy can be improved by reflecting accident statistical information other than an owner's pipeline like the following formula | equation (64)-(66).

Here, T is the total number of leakage accident data (T> N), U is the number of leakage accidents due to valve deterioration (U> M), V is the number of leakage accidents due to welding (V> Q), and W is leakage due to defects. Number of accidents (W> R).

  If accident statistical information other than the owner's pipeline does not exist, K1, K2 and K3 calculated using the equations (62) to (64) may not be used.

  In this case, the weighting factor is retrieved from the factor causal relationship DB 129, and the accident probability P due to the defect shown in equation (57) is calculated using the retrieved weighting factor and the normalization factor calculated using equation (55). (Leak) is calculated.

Note that P _j in the equation (55) corresponds to three of P (Leak | Valve) × P (Valve), P (Leak | Weld) × P (Weld), and P (Defect).

  According to one aspect of the present invention, in a pipeline facility, there is an accident occurrence factor that is similar to the accident that occurred even when the information on the pipeline defect cannot be obtained directly because the sensor limit is exceeded. The occurrence of an accident can be estimated. As a result, maintenance management in which the occurrence of an accident in the pipeline is predicted in advance becomes possible.

  In addition, according to one aspect of the present invention, by analyzing an accident that occurs in a place where there is no clear defect or the like, the accident occurrence probability of the accident is calculated, and the occurrence of an accident similar to the accident is predicted. be able to.

  Although the present embodiment has been described by taking a pipeline that uses natural gas as hot water as an example, it can also be applied to a pipeline that transports a liquid such as petroleum or ethylene, or a gas such as hydrogen or carbon dioxide. .

  In the present embodiment, a pipeline has been described as an example, but the present invention can also be applied to maintenance management of facilities related to length, such as railroad rails or telephone lines.

DESCRIPTION OF SYMBOLS 101 Accident data input part 102 Analysis instruction | indication part 103 Accident occurrence probability analysis part 104 Similarity analysis part 105 Probability analysis part 106 Compound probability calculation part 107 Accident data search part 108 Map display part 109 Soil data analysis part 111 Shape analysis part 112 Shape analysis Unit 113 defect analysis unit 115 CP analysis unit 117 operation history analysis unit 119 weight coefficient update unit 120 list output unit 121 map output unit 122 accident DB
123 Soil DB
124 Pipeline coordinate DB
125 Defect DB
126 Electrochemical DB
127 Operation history DB
128 probability distribution DB
129 Cause-and-effect database
130 Accident data 131 Result data 2201 Pipeline 2202, 2203 Compressor 2204 Cathode protection system 2206, 2207 Risk diagnosis system 2208, 2209 SCADA
2210, 2211 Gateway device 2212 Offline data input system 2213 Data center system 2214 Portable data input terminal 2215, 2216 Compressor station Compressor station 2217 Network 3101 Processor 3102 Memory 3103 Non-volatile storage medium 3104 Network interface

Claims (12)

An operation history database for storing operation information of equipment, an accident history database for storing accident information that is information on accidents that have occurred in the past in the equipment, the operation history database, and the accident history database, with reference to the equipment A computer system comprising a risk diagnosis server for estimating the occurrence of an accident in
The risk diagnosis server includes a processor, a memory connected to the processor, and a network interface connected to the processor,
The operation history database stores the operation information in which the position of the facility and information related to the facility are associated with each other,
The accident history database stores the accident information in which the occurrence position of the accident that occurred in the past is associated with the information on the cause of the accident that occurred in the past,
The risk diagnosis server
When accepting the input of information on the occurrence position of a new accident that is a newly occurred accident and information on the cause of the occurrence of the new accident, the information on the cause of the occurrence of the new accident and the past information are referred to the accident history database. Calculate the first similarity with the information on the cause of the accident that occurred,
Based on the calculated first similarity, for each new accident occurrence factor, calculate a similar accident occurrence probability of an accident caused by a factor similar to the new accident occurrence factor,
Referring to the operation history database, identify an accident occurrence estimated location that is a position where the occurrence of an accident due to a factor similar to the new accident occurrence factor is estimated,
A computer system that calculates an accident occurrence probability at the accident occurrence estimated location based on a similar accident occurrence probability calculated for each cause of occurrence of the new accident. The accident history database stores a weighting factor for each accident occurrence factor,
Information related to the equipment included in the operation information is information on the cause of the accident in the equipment,
The risk diagnosis server
When specifying the accident occurrence estimated location, referring to the operation history database, for each position of the facility included in the operation information, information on the cause of the accident in the facility corresponding to the position of the facility; Calculating a second similarity with the information on the cause of the new accident,
When the calculated second similarity is equal to or greater than a predetermined value, the location of the facility corresponding to information on the occurrence factor of the accident in the facility is specified for each occurrence factor as the accident occurrence estimated location,
When calculating the probability of occurrence of an accident at the place where the accident is estimated, the weighting factor for each occurrence factor is read from the accident history database;
Update the weighting factor for each occurrence factor so that the weighting factor for the accident occurrence factor corresponding to the accident occurrence estimated location is increased,
The probability of occurrence of an accident at the estimated place of occurrence of an accident is calculated using the calculated probability of occurrence of a similar accident for each cause of occurrence of a new accident and the updated weighting factor for each cause of occurrence of the accident. 1. The computer system according to 1.   The risk diagnosis server displays information related to the occurrence factor in the new accident in which the estimated probability of occurrence of the accident in the estimated occurrence location of the accident and the weighting factor of the occurrence factor of the new accident are equal to or greater than a predetermined value. The computer system according to claim 2, wherein the data is generated.   The risk diagnosis server is configured to display data for displaying the estimated accident occurrence probability at the calculated accident occurrence estimated location and information related to the cause of the new accident with the similar accident occurrence probability being a predetermined value or more. The computer system according to claim 2, wherein the computer system is generated. The operation history database is a database that stores the operation information related to the pipeline,
The operation information includes, for each position where the pipeline is installed, the shape of the pipeline, the defect distribution of pipes in the pipeline, the cathode protection potential in the cathode protection system, and environmental information around the pipeline And facility information, and transportation information of the pipeline,
The accident history database is a database that stores the accident information related to the pipeline,
The accident information includes, for each accident occurrence point, the shape of the pipeline, the distribution of pipe defects in the pipeline, the cathode protection potential in the cathode protection system, and environmental information around the pipeline. Including equipment information,
The risk diagnosis server
The shape of the pipeline at the point of occurrence of the new accident, the distribution of pipe defects in the pipeline, the cathode protection potential in the cathode protection system, and the surrounding environment information and facility information where the pipeline is installed, The shape of the pipeline at the point of occurrence of an accident that occurred in the past, the distribution of pipe defects in the pipeline, the cathode protection potential in the cathode protection system, and the surrounding environment information and facility information where the pipeline is installed And calculating the first similarity based on
The shape of the pipeline at the point of occurrence of the new accident, the distribution of pipe defects in the pipeline, the cathode protection potential in the cathode protection system, and the surrounding environment information and facility information where the pipeline is installed, Based on the shape of the pipeline at the position where the pipeline is installed, the defect distribution of the pipe in the pipeline, the cathode protection potential in the cathode protection system, the surrounding environment information and facility information where the pipeline is installed, The computer system according to claim 2, wherein the second similarity is calculated. The computer system further includes a probability distribution database that stores a probability distribution for calculating a probability of occurrence of an accident for each occurrence factor,
The cause of the accident is that the pipeline shape is bent, the pipe is defective, the cathode protection potential abnormal voltage in the cathode protection system and the cathode protection system equipment failure, the pipe coating abnormality, the pipeline is installed Anomalies in soil hydrogen ion index, and deterioration of the pipe due to pressure fluctuation and temperature fluctuation in transportation using the pipeline,
The risk diagnosis server
Referring to the probability distribution database, the pipeline shape is bent, the pipe is defective, the cathode protection potential abnormal voltage in the cathode protection system and the cathode protection system equipment failure, the pipe coating abnormality, the pipeline is Calculate the probability of accident occurrence of accidents that occurred in the past with respect to abnormalities of the hydrogen ion index of the installed soil, and all the causes of deterioration of the pipe and temperature fluctuation due to pressure fluctuations in transportation using the pipeline ,
Using the calculated accident occurrence probability of an accident that occurred in the past, and the calculated first similarity for each new accident occurrence factor, the similarity for each new accident occurrence factor 6. The computer system according to claim 5, wherein an accident occurrence probability is calculated. An operation history database for storing operation information of equipment, an accident history database for storing accident information that is information on accidents that have occurred in the past in the equipment, the operation history database, and the accident history database, with reference to the equipment A risk diagnosis method in a computer system comprising a risk diagnosis server for estimating the occurrence of an accident in
The risk diagnosis server includes a processor, a memory connected to the processor, and a network interface connected to the processor,
The operation history database stores the operation information in which the position of the facility and information related to the facility are associated with each other,
The accident history database stores the accident information in which the occurrence position of the accident that occurred in the past is associated with the information on the cause of the accident that occurred in the past,
The method
When the risk diagnosis server receives input of information on the occurrence position of a new accident that is a newly occurred accident and information on the cause of occurrence of the new accident, the cause of occurrence of the new accident is referred to the accident history database. A first step of calculating a first similarity between the information on the accident and the information on the cause of the accident that occurred in the past, and the risk diagnosis server is configured to calculate the new similarity based on the calculated first similarity. A second step of calculating a similar accident occurrence probability of an accident caused by a factor similar to the new accident occurrence factor for each accident occurrence factor;
A third step in which the risk diagnosis server refers to the operation history database and identifies an accident occurrence estimated location that is a position where an occurrence of an accident due to a factor similar to the occurrence factor of the new accident is estimated;
The risk diagnosis server includes a fourth step of calculating an accident occurrence estimation probability at the accident occurrence estimated location based on a similar accident occurrence probability calculated for each occurrence factor of the new accident. Risk diagnosis method. The accident history database stores a weighting factor for each accident occurrence factor,
Information related to the equipment included in the operation information is information on the cause of the accident in the equipment,
The third step includes
The risk diagnosis server refers to the operation history database, and for each position of the facility included in the operation information, information on the cause of the accident in the facility corresponding to the position of the facility, and the new accident A fifth step of calculating a second similarity with the information of the cause of occurrence;
If the calculated second similarity is greater than or equal to a predetermined value, the risk diagnosis server sets the location of the facility corresponding to the information on the cause of the accident in the facility as the accident occurrence estimated location. A sixth step of identifying each,
The fourth step includes
A seventh step in which the risk diagnosis server reads a weighting factor for each occurrence factor from the accident history database;
An eighth step in which the risk diagnosis server updates a weighting factor for each occurrence factor so that a weighting factor for the accident occurrence factor corresponding to the accident occurrence estimated place is increased;
The risk diagnosis server uses the calculated similar accident occurrence probability for each new accident occurrence factor and the updated weighting factor for each occurrence factor to calculate an accident occurrence estimation probability at the accident occurrence estimated location. The risk diagnosis method according to claim 7, comprising: 9 steps.   In the method, the risk diagnosis server further relates to a cause of occurrence of the new accident in which the calculated probability of occurrence of the accident at the estimated occurrence location of the accident and a weighting factor of the cause of occurrence of the new accident are equal to or greater than a predetermined value. The risk diagnosis method according to claim 8, further comprising a tenth step of generating data for displaying information to be displayed.   In the method, the risk diagnosis server further includes an accident occurrence estimated probability at the calculated accident occurrence estimated place, and information related to an occurrence factor of the new accident in which the similar accident occurrence probability is a predetermined value or more. The risk diagnosis method according to claim 8, further comprising an eleventh step of generating data for display. The operation history database is a database that stores the operation information related to the pipeline,
The operation information includes, for each position where the pipeline is installed, the shape of the pipeline, the defect distribution of pipes in the pipeline, the cathode protection potential in the cathode protection system, and environmental information around the pipeline And facility information, and transportation information of the pipeline,
The accident history database is a database that stores the accident information related to the pipeline,
The accident information includes, for each accident occurrence point, the shape of the pipeline, the distribution of pipe defects in the pipeline, the cathode protection potential in the cathode protection system, and environmental information around the pipeline. Including equipment information,
In the first step, the risk diagnosis server is configured such that the shape of the pipeline at the occurrence point of the new accident, the distribution of pipe defects in the pipeline, the cathode protection potential in the cathode protection system, and the pipeline Environmental information and facility information around the installed area, the shape of the pipeline at the point of occurrence of the accident that occurred in the past, pipe defect distribution in the pipeline, cathode protection potential in the cathode protection system, and the pipe A twelfth step of calculating the first similarity based on environmental information and facility information around the line;
In the fifth step, the risk diagnosis server is configured such that the shape of the pipeline at the point of occurrence of the new accident, the distribution of pipe defects in the pipeline, the cathode protection potential in the cathode protection system, and the pipeline Environmental information and facility information around the installation, the shape of the pipeline at the location where the pipeline is installed, pipe defect distribution in the pipeline, cathode protection potential in the cathode protection system, the pipeline is installed The risk diagnosis method according to claim 8, further comprising a thirteenth step of calculating the second similarity based on surrounding environmental information and facility information. The computer system further includes a probability distribution database that stores a probability distribution for calculating a probability of occurrence of an accident for each occurrence factor,
The cause of the accident is that the pipeline shape is bent, the pipe is defective, the cathode protection potential abnormal voltage in the cathode protection system and the cathode protection system equipment failure, the pipe coating abnormality, the pipeline is installed Anomalies in soil hydrogen ion index, and deterioration of the pipe due to pressure fluctuation and temperature fluctuation in transportation using the pipeline,
The second step includes
The risk diagnosis server refers to the probability distribution database, the pipeline shape is bent, the pipe is defective, the abnormal voltage of the cathode protection potential in the cathode protection system and the equipment failure of the cathode protection system, the coating of the pipe Abnormalities, abnormalities in the hydrogen ion index of the soil where the pipeline is installed, and deterioration of the pipe due to pressure fluctuations in transportation using the pipeline, and all occurrence factors of temperature fluctuations in the past. A fourteenth step of calculating an accident occurrence probability;
The risk diagnosis server uses the calculated accident occurrence probability of an accident that has occurred in the past, and the calculated first similarity for each new accident occurrence factor, and The risk diagnosis method according to claim 11, further comprising: a fifteenth step of calculating the similar accident occurrence probability for each occurrence factor.