Inspection planning support device, inspection planning support method, storage medium, and inspection planning support program

Abstract

This invention provides a technology for selecting a tube to be inspected from among numerous tubes installed in a multi-tube heat exchanger. [Solution] In an inspection planning support device 2 that assists in planning an inspection of a multi-tube heat exchanger 10 equipped with a large number of tubes 102, the inspection tube determination unit 24 determines the tubes 102 to be inspected based on the results of statistical analysis of inspection data for heat exchangers 10 of the same type, which are based on information regarding the wall thickness of the tubes 102 and the type of heat exchanger 10 and the position of the tubes 102. The tubes 102 located in positions where there is a high risk of wall thickness reduction exceeding a preset threshold are determined to be inspected tubes.

Description

Technical Field

[0001] The present invention relates to a technique for formulating an inspection plan for tubes of a multi-tube heat exchanger.

Background Art

[0002] Plants for fluid treatment include petroleum refineries that perform distillation and desulfurization of crude oil and various intermediate products, chemical plants that produce petrochemical products, intermediate chemicals, polymers, etc., and natural gas plants that perform liquefaction of natural gas and separation and recovery of natural gas liquids. These plants are equipped with various heat exchangers for heating and cooling the fluids handled inside.

[0003] Among the various heat exchangers provided in a plant, a multi-tube heat exchanger (hereinafter also simply referred to as a "heat exchanger") has a large number of tubes that can reach dozens to over a thousand, and heat exchange is performed between the fluid flowing on the inner surface side of each tube and the fluid flowing on the outer surface side. Also, in some cases, heat exchange is performed using a few tubes, such as in a double-tube heat exchanger. These tubes are made of, for example, metal, but wall thickness reduction progresses due to erosion caused by the fluid flow and corrosion (corrosion) due to the electrochemical action between the metal and the fluid. As a result of the progress of wall thickness reduction, when an opening penetrating the tube wall is formed, a tube leak occurs where the high-pressure side fluid flows into the low-pressure side.

[0004] In this regard, for heat exchangers, tubes are removed and their wall thickness is measured during periodic maintenance of the plant. Based on the wall thickness inspection results, tubes with significant wall thinning can be identified and replaced or plugged to prevent tube leaks. On the other hand, plants may have dozens to hundreds or even more heat exchangers, and the total number of tubes in these heat exchangers is enormous. In this case, 100% inspection using eddy current testing is the basic method for non-magnetic tubes such as copper alloys. However, for magnetic tubes such as carbon steel and low-alloy steel, 100% inspection is not practical because technology for high-speed inspection has not been established.

[0005] Therefore, in heat exchangers where eddy current testing is difficult, it is common practice to extract some tubes and perform representative inspections. When selecting the tubes to be inspected, the inspector selected the tubes based on past wall thickness inspection results. However, the task of selecting tubes to be inspected by referring to past inspection results for a large number of tubes is burdensome for the inspector. Furthermore, differences in experience and skill among multiple inspectors can lead to inconsistencies in the appropriateness of tube selection, potentially resulting in differences in the effectiveness of preventing tube leaks.

[0006] Patent Document 1 describes a technique for calculating a critical index for formulating a preventive maintenance strategy for a heat exchanger equipped with tube bundles, based on the results of scoring six pre-evaluation elements such as production importance, self-importance, and expected life. Patent Document 2 describes a technique for dividing a heat exchanger into multiple zones and displaying an estimated value of the likelihood of leakage occurring in the entire heat exchanger on a screen, based on information such as the age of inspected tubes in a first zone, which includes inspected tubes and uninspected tubes. [Prior art documents] [Patent Documents]

[0007] [Patent Document 1] Chinese Patent Application Publication No. 115796825 [Patent Document 2] International Publication No. 2012 / 060744 [Overview of the Initiative] [Problems that the invention aims to solve]

[0008] Patent documents 1 and 2 do not disclose a technique for individually selecting tubes to be inspected when planning an inspection of a heat exchanger. Therefore, a technique for individually selecting tubes to be inspected when planning an inspection of a heat exchanger is desired.

[0009] This invention has been made in view of the above circumstances, and provides a technique for selecting a tube to be inspected from among a large number of tubes provided in a multi-tube heat exchanger. [Means for solving the problem]

[0010] This inspection planning support device is an inspection planning support device that assists in planning inspections of multi-tube heat exchangers equipped with a large number of tubes, The heat exchanger is characterized by having an inspection tube determination unit that, based on the results of statistical analysis of inspection data that includes the type common to the heat exchanger being inspected, among the inspection data which associates information on the wall thickness of each of the many tubes contained in the heat exchanger with the type of the heat exchanger and the position of the tubes arranged in the heat exchanger, determines from among the many tubes arranged in the heat exchanger being inspected that the tubes located in positions where there is a high risk of wall thickness reduction exceeding a preset threshold will be inspected, as inspection tubes to be inspected.

[0011] The aforementioned inspection planning support device may include the following: (a) The types are classified based on the type of heat exchanger and the phase state of the heat supply fluid and heat receiving fluid flowing on the inner or outer side of the tube. In this case, the type is selected from horizontal, vertical, or air fin cooler (AFC). (b) The inspection data includes information on the wall thickness measured from the inner and outer surfaces of the tube, and the statistical analysis is based on the inspection data for the inner and outer surfaces of the tube, respectively. (c) In the statistical analysis described above, the region in the heat exchanger where the numerous tubes are arranged is divided into multiple areas, each containing multiple tubes, For each of the divided areas, the greater the proportion of tubes with wall thinning greater than the threshold, the higher the risk is considered to be for the tubes placed in that area. In this case, the statistical analysis selects the area from among the divided areas where the proportion is greater than the threshold as the recommended area containing the test tubes.

[0012] (d) In the statistical analysis, the variability of the thickness of the tubes in the inspection data is evaluated for each type, and the inspection tube determination unit adjusts the number of inspection tubes based on the evaluation of the variability. In this case, the inspection tube determination unit increases the number of inspection tubes the greater the proportion of tubes in which the thickness of the tubes deviates in the direction of being greater than a preset range in the evaluation of the variability. Furthermore, the evaluation of the variability is performed using a box plot created from the inspection data with an upper limit set, and it is determined that the thickness of the tubes deviates in the direction of being greater than a preset range by the thickness of the tubes being an outlier greater than the upper limit. (e) The heat exchanger to be inspected is equipped with a path through which a heat supply fluid or heat receiving fluid passes on the inner surface of the tube, and when determining a predetermined number of inspection tubes for the heat exchanger to be inspected, the inspection tube determination unit determines, based on the inspection data, for each path, the tube with the greatest wall thinning or the smallest wall thickness from among the tubes belonging to that path as the inspection tube, and when the number of determined tubes is less than the number of inspections, it further determines the tubes located in high-risk positions as the inspection tubes, and the number of inspections is greater than the number of paths.

[0013] (f) The system includes a type classification unit that identifies which of the above types the heat exchanger to which the inspection data was measured belongs, based on statistics obtained from the information on wall thickness. In this case, the type classification unit performs type classification of the heat exchanger using a learning model created by pre-machine learning the correspondence between the statistics and the types. Furthermore, the type classification unit creates the learning model by machine learning the correspondence between the statistics selected from a group of statistics consisting of maximum value, minimum value, data range, mean, median, interquartile range, first quartile, third quartile, and standard deviation, and the types, using the random forest method, based on the information on wall thickness obtained for the inner or outer surface of the tube. [Effects of the Invention]

[0014] According to the present invention, it is possible to individually determine which tubes should be inspected from a multi-tube heat exchanger under inspection. [Brief explanation of the drawing]

[0015] [Figure 1] This is a block diagram of an inspection planning support device for a multi-tube heat exchanger according to an embodiment. [Figure 2A] This is a diagram illustrating a horizontally mounted heat exchanger. [Figure 2B] This is a diagram illustrating a vertically mounted heat exchanger. [Figure 2C] It is an explanatory diagram of an air fin cooler (AFC). [Figure 3] It is a classification diagram of the type of a multi-tube heat exchanger. [Figure 4] It is an example of a statistic used when creating a learning model for classifying multi-tube heat exchangers by type. [Figure 5] It is an example of displaying inspection data classified by type in a box-and-whisker plot with an upper limit value set. [Figure 6] It is an explanatory diagram showing an example of setting a score based on a statistical index. [Figure 7A] It is an example of area setting for a vertical or horizontal heat exchanger. [Figure 7B] It is an example of area setting for an AFC. [Figure 8] It is a distribution diagram showing the relationship between the area set for the heat exchanger and the data for which the thinning rate is an outlier. [Figure 9] It is a distribution diagram showing the relationship between the area set for the heat exchanger and the data for which the maximum value of the outlier for each horizontal heat exchanger exceeds a predetermined threshold value. [Figure 10A] It is a first flowchart related to an example of the determination operation of the tube to be inspected. [Figure 10B] It is a second flowchart related to an example of the determination operation of the tube to be inspected. [Figure 11] It is a diagram showing the arrangement position of the tube determined as the inspection target. [Figure 12] It is a block diagram showing a configuration example of an inspection plan support device using a cloud service.[[ID=�9]]

Embodiments for Carrying Out the Invention

[0016] First, the outline of the inspection plan support device 2 according to the embodiment will be described while referring to FIG. 1. FIG. 1 is a block diagram of the plant 1 of this example, and a schematic diagram of the plant 1 in which support for formulating an inspection plan for a multi-tube heat exchanger is performed using this.

[0017] Plant 1 is not particularly limited as long as it has the function of processing fluids, and examples include petroleum refining plants that perform distillation and desulfurization of crude oil and various intermediate products, chemical plants that produce petrochemical products, intermediate chemicals, polymers, etc., and natural gas plants that perform liquefaction of natural gas, separation and recovery of natural gas liquids, etc.

[0018] Plant 1 is equipped with a variety of equipment, including static equipment such as towers and tanks and various heat exchangers, dynamic equipment such as pumps, piping between these static and dynamic equipment, as well as various control (instrumentation) equipment and electrical equipment. Among these pieces of equipment, the inspection planning support device 2 has the function of assisting in the planning of inspections of the tubes installed in the multi-tube heat exchanger 10 (hereinafter also simply referred to as "inspection planning support").

[0019] Figures 2A and 2C are schematic diagrams showing examples of heat exchanger 10 types that are subject to inspection planning support by the inspection planning support device 2. Figure 2A shows an example configuration of a horizontal heat exchanger 10A, and Figure 2B shows an example configuration of a vertical heat exchanger 10B. These heat exchangers 10A and 10B have a configuration in which many tubes 102, for example several tens to a thousand, are arranged inside a shell 101 that forms the outer shell. As shown in Figure 2A, in the horizontal heat exchanger 10A, these tubes 102 are arranged to extend in a direction intersecting the direction of gravity. Also, as shown in Figure 2B, in the vertical heat exchanger 10B, the tubes 102 are arranged to extend along the direction of gravity.

[0020] Figures 2A and 2B show fixed-tube sheet type heat exchangers 10A and 10B, in which both ends of each tube 102 are held by a fixed tube sheet 103. However, the structure of heat exchangers 10A and 10B that can be used to provide inspection planning support using the inspection planning support device 2 is not limited to the fixed-tube sheet type. It may also be a U-tube type using U-shaped tubes 102, or other types of structures such as a floating-head type in which one end of the tube 102 is held by a fixed tube sheet 103 and the other end is held by a movable tube sheet. Heat exchangers 10A and 10B may, for example, be equipped with a path 106 through which the fluid flowing on the tube 102 side passes.

[0021] Furthermore, as illustrated in Figure 2A, in the heat exchanger 10A, the spaces on the inlet and outlet sides of the tubes 102 within the shell 101 may be partitioned by a partition plate 105 to form paths 106 containing multiple tubes 102 each (the horizontal heat exchanger 10A shown in Figure 2A is an example with two paths). (Note that when indicating the number of paths 106, the symbol "106" is omitted.) In this case, the fluid flowing through the tubes 102 passes through the tubes 102 constituting each path 106 in order. In addition, heat exchangers 10A and 10B may be provided with baffles 104 to guide the flow direction of the fluid flowing within the shell 101.

[0022] Furthermore, as illustrated in Figure 2B, in the heat exchanger 10B, a partition plate 105 is not required to be provided within the shell 101 (the vertical heat exchanger 10B shown in Figure 2B is an example with one path). In the above examples, an example was shown in which a partition plate 105 is provided in the horizontal heat exchanger 10A to form two paths 106, and an example was shown in which a partition plate 105 is not provided in the vertical heat exchanger 10B to form one path 106, but the examples are not limited to these. For example, a partition plate 105 may be provided in the vertical heat exchanger 10B to form two paths 106, or a partition plate 105 may not be provided in the horizontal heat exchanger 10A to form one path 106. For example, three or more paths 106 may be formed for both the horizontal heat exchanger 10A and the vertical heat exchanger 10B.

[0023] In the heat exchangers 10A and 10B having the configuration described above, heat exchange takes place between the fluid flowing on the shell 101 side and the fluid flowing on the tube 102 side. The fluids flowing on the shell 101 side and the tube 102 side may be gases or liquids. Furthermore, the fluids flowing on either the shell 101 side, the tube 102 side, or both may undergo phase changes such as condensation from gas to liquid or evaporation from liquid to gas.

[0024] Next, Figure 2C shows a multi-tube heat exchanger 10C configured as an AFC (Air Fin Cooler, also called ACHE (Air Cooled Heat Exchanger)) (hereinafter referred to as "AFC10C"). The AFC10C is configured as a cooler that cools the fluid flowing through the tubes 102 with air, which is the cooling fluid (corresponding to the "heat receiving fluid" described later). The AFC10C includes a tube bundle 114 that bundles together many tubes 102 through which the fluid to be cooled (corresponding to the "heat supply fluid" described later) flows, and a fan 110 for supplying the cooling fluid to the tube bundle 114. In the AFC10C, each tube 102 is arranged to extend in a direction intersecting the direction of gravity. From this perspective, the AFC10C can also be understood as a variation of the horizontal heat exchanger 10A (see Figure 3(b) described later).

[0025] The tube bundle 114 is open on both its upper and lower sides, allowing cooling fluid to flow from bottom to top through the gaps between adjacent tubes 102. The side perimeter of the frame that constitutes the tube bundle 114 is fixed to the upper surface of the frame. The lower surface of the tube bundle 114 corresponds to the intake port for the cooling fluid.

[0026] The fan 110 comprises a plurality of rotor blades 111 arranged to extend radially from a center of rotation. The base end of each rotor blade 111 is connected to the upper end of a rotating shaft 112 at the center of rotation, and the rotating shaft 112 is arranged to extend in the vertical direction. The lower side of the rotating shaft 112 passes through a tube bundle 114, and its lower end is connected to a drive unit 115 located below the tube bundle 114. As shown in Figure 2C, for example, the drive unit 115 may be configured by directly connecting the rotating shaft 112 to a rotary motor. Alternatively, the drive unit 115 may consist of a pulley mechanism (not shown) provided at the lower end of the rotating shaft 112 and a rotary motor that drives this pulley mechanism.

[0027] A duct 113 is provided in the region extending from the upper surface of the frame of the tube bundle 114, through the lateral position of the rotor blade 111, to the upper side thereof, forming an exhaust path for circulating the air that has passed through the tube bundle 114. The duct 113 has a cylindrical shape and opens upwards at a position above the rotor blade 111.

[0028] Although Figure 2C shows an example of a suction-type AFC10C with the rotor blades 111 positioned above the tube bundle 114, a push-type AFC10C with the rotor blades 111 positioned below the tube bundle 114 may also be used. In this case, the lower surface of the rotor blades 111 serves as the intake port for the cooling fluid.

[0029] Returning to the explanation of Figure 1, the inspection planning support device 2, which assists in the inspection planning of the various heat exchangers 10 described above, is configured, for example, by a computer. The inspection planning support device 2 may be installed in a central control room located within the premises of Plant 1. Alternatively, the inspection planning support device 2 may be installed in an office located remotely from the premises of Plant 1.

[0030] As shown in Figure 1, the inspection planning support device 2 includes an information acquisition unit 21 that acquires raw data for inspection data of each tube 102 in the heat exchanger 10 installed in the plant 1, a storage unit 22 that stores information such as raw data, a type classification unit 23 that performs type classification of the heat exchanger 10, and an inspection tube determination unit 24 that determines the inspection tubes that should be inspected in the heat exchanger 10. The inspection planning support device 2 also includes a data input / output control unit and an image processing unit (neither of which are shown). Using the functions of these input / output control unit and image processing unit, it is possible to receive and display information regarding the type classification of the heat exchanger 10, and to display images of information regarding the tubes 102 that have been determined to be inspected to the user, in communication with the user's input / output terminal 3. In the example shown in Figure 1, an example is shown in which the input / output terminal 3 is configured with a touch panel display that receives input from the user via a display screen.

[0031] The information acquisition unit 21 acquires data including information regarding the wall thickness of the tubes 102 in each heat exchanger 10. To distinguish it from the "inspection data" described later, this data is also called "raw data" for generating the inspection data. The information regarding wall thickness includes information that identifies whether the inspection was performed from the inner or outer surface of the tube 102, and data that allows for the identification of "corrosion inspection statistics" described later, such as the wall thickness of the tube 102. Furthermore, the raw data may associate information regarding the wall thickness of the tubes 102 with information that identifies the plant 1, information that identifies the heat exchanger 10 to be inspected, information that identifies the individual tubes 102 installed in the heat exchanger 10, and the inspection date and time. In addition, the raw data may include information indicating the arrangement position of the tubes 102 within the heat exchanger 10. Here, "inspection data" is data in which information regarding the wall thickness of each of the many tubes 102 contained in the heat exchanger 10 is associated with the type of heat exchanger 10 described later and the position of the tubes 102 installed in the heat exchanger 10.

[0032] The raw data may be obtained, for example, through an external inspection history management system for the heat exchanger 10. This management system stores a history of wall thickness information obtained in past inspections, associated with the various identification information described above. Alternatively, the raw data may be individually entered by the user. The storage unit 22 stores the raw data obtained via the information acquisition unit 21. This raw data may be managed as an inspection history database, associated with, for example, identification information and inspection date and time information.

[0033] Here, the inventors of this case, in order to solve the problem of individually selecting the tube 102 to be inspected from among the many tubes 102 provided in the heat exchanger 10, performed data analysis on information regarding the wall thickness of the tubes 102 in the many heat exchangers 10. As a result, they found that when (i) the type of heat exchanger 10, (ii) the phase state of the fluid flowing on the inner and outer sides of the tube 102, and (iii) the arrangement position of the tube 102 within the heat exchanger 10 are the same, the information regarding the wall thickness of tubes 102, even if they are located in different heat exchangers 10, often shows similar trends.

[0034] Therefore, the inspection planning support device 2 of this embodiment groups the wall thickness information in correspondence with the information in (i) to (iii), and performs statistical analysis on the grouped wall thickness information to evaluate, for example, the variation in the amount of wall thinning per unit period (wall thinning rate [mm / year]). Based on the results of this statistical analysis, the inspection planning support device 2 is configured to select an increasing number of tubes 102 with common placement positions in (iii) as inspection targets, the more tubes 102 have wall thinning that deviates from the predetermined range.

[0035] However, the "information regarding wall thickness" here is not necessarily limited to the "thinning rate," which is the amount of wall thickness lost per unit period. For example, the thinning rate per unit period, which is the ratio of the amount of wall thickness lost to the initial wall thickness, may be used. Alternatively, the wall thickness of tube 102 obtained from the inspection may be used itself.

[0036] Using the method outlined above, the inspection planning support device 2 selects the tubes 102 to be inspected in each heat exchanger 10. In this case, the type classification unit 23 associates the wall thickness information contained in the raw data stored in the memory unit 22 with the type of heat exchanger 10 (grouping as described above). Here, the type of heat exchanger 10 is a classification of the heat exchanger 10 based on (i) the model of the heat exchanger 10 and (ii) the phase state of the fluid flowing on the inner and outer sides of the tubes 102.

[0037] Figure 3 is a diagram listing examples of combinations of (i) and (ii) that define the type of heat exchanger 10. Since the content of (ii) differs between the horizontal heat exchanger 10A, the vertical heat exchanger 10B, and the AFC 10C, they are shown separately in Figures 3(a) and (b). In these diagrams, the leftmost column (column 1) shows (i) "Type of heat exchanger 10". Next, the second to fourth columns from the left (columns 2 to 4) show specific examples of (ii) "Phase state of the fluid flowing on the inner and outer sides of the tube 102".

[0038] The second column indicates whether the heat supply fluid on the high-temperature side or the heat receiving fluid on the low-temperature side is flowing through the inner or outer surface of tube 102. In all cases of the horizontal heat exchanger 10A, vertical heat exchanger 10B, and AFC 10C, "tube side" indicates that the fluid is flowing on the inner surface of tube 102. "Shell side" in the horizontal heat exchanger 10A and vertical heat exchanger 10B indicates that the fluid is flowing through the space inside the shell 101, i.e., on the outer surface of tube 102. "Atmospheric side" in the AFC 10C indicates that the fluid is flowing inside the tube bundle 114, i.e., on the outer surface of tube 102.

[0039] Furthermore, in descriptions such as "shell side / tube side," the forward side of the slash ( / ) indicates the space through which the heat supply fluid flows, and the backward side indicates the space through which the heat receiving fluid flows. In the AFC10C, the heat supply fluid is only the fluid to be cooled flowing inside the tube 102, and the heat receiving fluid is only the air flowing through the tube bundle 114. Therefore, there is only one combination of spaces through which the heat supply fluid and heat receiving fluid flow: "tube side / air side" (second column of Figure 3(b)).

[0040] The third column shows the phase state of the heat supply fluid in the space through which it flows. In notations such as "liquid-liquid," the part before the hyphen (-) indicates the phase state when the fluid flows into the space, and the part after the hyphen indicates the phase state when the fluid flows out of the space. Therefore, the notation "gas-liquid" indicates that a phase change is occurring in the space through which the heat supply fluid flows, where the gas condenses into a liquid.

[0041] Column 4 shows the phase state of the heat supply fluid in the space through which the heat receiving fluid flows. The meaning of notations such as "liquid-liquid" is the same as in Column 3. In AFC10C, the heat receiving fluid is only the atmosphere, which is a gas, and there is no phase change, so it is a single case of "gas-gas" from the time it flows into the tube bundle 114 until it flows out (Column 4 of Figure 3(b)).

[0042] As described above, by combining the cases shown in columns 1 to 4 of Figure 3(a), horizontal heat exchangers 10A and vertical heat exchangers 10B can be classified into 36 types. Furthermore, by combining the cases shown in columns 1 to 4 of Figure 3(b), AFC 10C can be classified into 3 types. Therefore, the "thickness information" stored in the memory unit 22 can be associated with a total of 39 types of heat exchangers 10.

[0043] In the 39 types described above, the lower part of Figure 3 also shows examples of using four-letter symbols to concisely represent the space and phase state through which the heat supply fluid and heat receiving fluid flow. "A" and "B" are symbols used to identify whether it is a heat supply fluid or a heat receiving fluid, and are shown in the first letter. "S" and "T" are symbols used to identify whether the space through which the fluid flows is the shell side, the atmosphere side, or the tube side, and are shown in the second letter. "L" and "V" indicate whether the layer state is liquid or gaseous, and the phase state at inflow and outflow are shown in the third and fourth letters, respectively.

[0044] According to the symbolic rules described above, the symbol "ASVV" in Example 1) shown in Figure 3 means that "the heat supply fluid flows on the shell side (or atmospheric side), is a gas at inflow and a gas at outflow." Similarly, the symbol "BTLV" in Example 2) shown in the same figure means that "the heat receiving fluid flows on the tube side, is a liquid at inflow and a gas at outflow." Therefore, as shown in Figures 5, 8, and 9 later, by writing "ASVV+BTLV," etc., the contents of columns 2 to 4 of Figure 3(a) and (b) in the heat exchanger 10 being explained can be succinctly represented.

[0045] The types of heat exchangers 10 described above are a unique classification set when performing the selection of inspection tubes using the inspection planning support device 2. Therefore, the wall thickness information contained in the raw data acquired from the information acquisition unit 21 is not necessarily pre-classified in correspondence with these types. On the other hand, the raw data stored in the storage unit 22 is enormous in number, and it may not be practical for each user to individually associate the wall thickness information contained in this raw data with the type of heat exchanger 10.

[0046] Therefore, the type classification unit 23 identifies which of the above types the heat exchanger 10, whose wall thickness information has been measured, belongs to, based on statistical quantities obtained from the wall thickness information. For example, the type classification unit 23 performs type classification of the heat exchanger 10 using a learning model that has been created in advance by machine learning the correspondence between the aforementioned statistical quantities and types.

[0047] Figure 4 shows examples of wall thickness information (left column) and statistics (right column) used in creating a learning model of the heat exchanger 10. The wall thickness information uses the thinning rate obtained by inspecting the inner and outer surfaces of the tube 102, respectively. As for statistics, an example is selecting at least one from a set of statistics consisting of the maximum value, minimum value, data range (difference between maximum and minimum values), mean, median, interquartile range, first quartile, third quartile, and standard deviation for each thinning rate on the inner and outer surfaces. The example shown in Figure 4 illustrates the case where all nine of these statistics are used for each of the inner and outer thinning rates (a total of 18 statistics).

[0048] Regarding machine learning methods, there are no particular limitations as long as they are capable of creating a learning model that can perform type classification of the heat exchanger 10 based on these statistics (training data). Specific examples include methods such as the random forest method and the boosting method.

[0049] For example, a learning model was created using the Random Forest method with the aforementioned statistical data (training data) obtained from 16,795 data points on the thinning rate of the inner and outer surfaces of 27 types of heat exchangers 10, totaling 995 units. In this case, type classification was performed using the learning model for 1,636 units, with 33,223 data points on the inner surface and 17,766 data points on the outer surface. As a result, type classification into 21 types of heat exchangers 10 (horizontal heat exchanger 10A, vertical heat exchanger 10B, AFC 10C) was performed, and the accuracy rate was 96.9%. However, when checking the accuracy rate for individual types, some had low accuracy rates. Considering such cases, after performing type classification by the learning model, the user may check and correct the classification results (information processing between the type classification unit 23 and the input / output terminal 3 in Figure 1).

[0050] The original data stored in the memory unit 22 is grouped by the type classification unit 23, which associates it with type classification information that has been corrected by the user as appropriate. The type classification unit 23 then outputs this information (information on wall thickness, type of heat exchanger 10, and information indicating the arrangement position of the tubes 102 within the heat exchanger 10) to the inspection tube determination unit 24 as inspection data.

[0051] It should be noted that it is not a mandatory requirement that the inspection planning support device 2 be equipped with a type classification unit 23 and automatically perform type classification of the heat exchanger 10. For example, the type classification information may be associated with an external management system for the inspection history of the heat exchanger 10. In this case, the information acquisition unit 21 will directly acquire the inspection data with the above configuration from an external source. Furthermore, if possible, it is not ruled out that the user may individually perform type classification on the raw data to generate the inspection data.

[0052] The inspection tube determination unit 24 selects the inspection tubes 102 to be inspected for a specific heat exchanger 10, which is the next target of inspection, based on the inspection data. In selecting the inspection tubes, (1) the number of tubes 102 to be inspected is determined based on the results of statistical analysis of inspection data common to the heat exchanger 10 and type to be inspected. Next, (2) the area where many tubes 102 are located is divided into multiple areas, and a risk assessment is performed for each area based on the results of statistical analysis. Tubes 102 located in high-risk locations are given priority, and the number of inspection tubes determined in (1) is also satisfied. Specific examples of processes (1) and (2) will be explained below with reference to Figures 5 to 9.

[0053] Figure 5 illustrates the results of a statistical analysis using box plots to determine the variation in the wall thickness of tubes 102 (information regarding the wall thickness of tubes 102) for two types of heat exchangers 10. The box plots shown in Figure 5 all show the degree of variation in the wall thickness of horizontal heat exchanger 10A, with the vertical axis representing the wall thickness [mm / year] value. When determining the number of tubes 102 to be inspected, it is important to understand the degree of variation in the direction of increasing wall thickness. Therefore, Figure 5 shows box plots for the region where the wall thickness is greater than the median value, which will be discussed later.

[0054] On the left side of Figure 5, a box plot of the wall thinning rate is shown for "Type 1," a horizontal heat exchanger 10A, which is "ATVV + BSLV (heat supply fluid flows on the tube 102 side, gas at both inflow and outflow + heat receiving fluid flows on the shell 101 side, liquid at inflow and gas at outflow)." On the right side of the same figure, a box plot of the wall thinning rate is shown for "Type 2," a horizontal heat exchanger 10A, which is "ATVV + BSLL (heat supply fluid flows on the tube 102 side, gas at both inflow and outflow + heat receiving fluid flows on the shell 101 side, liquid at both inflow and outflow and gas at outflow)."

[0055] The inspection tube determination unit 24 calculates statistical quantities such as the median of the thinning rate and the third quartile for each heat exchanger 10, based on data on the thinning rate of the inner and outer surfaces of the tubes 102 installed in the heat exchanger 10. From the statistical quantities obtained for each heat exchanger 10 in this way, a set of statistical quantities for heat exchangers 10 of the same type is obtained. Then, from this set of statistical quantities for each type of heat exchanger 10, the median values ​​of each statistical quantity (hereinafter also referred to as "type-specific medians" to avoid confusion with the median of the thinning rate) are identified, such as the median of the thinning rate and the median of the third quartile.

[0056] In the box plot shown in Figure 5, the median and the third quartile are the type-specific medians calculated in this way. The upper limit of the box plot is the value obtained by adding 1.5 times the interquartile range (IQR), which is the difference between the third quartile and the first quartile, to the third quartile (both the first and third quartiles are type-specific medians). The upper limit is an indicator used to determine if the thinning rate deviates from a predetermined range, and data with a thinning rate greater than this upper limit are considered "outliers."

[0057] In Type 1, the above statistical analysis was performed on a total of 60 horizontal heat exchangers 10A, and a box plot with an upper limit was created, as shown on the left side of Figure 5. As a result, the percentage of outliers (data exceeding the upper limit) among the total number of thinning rate data obtained from the 60 horizontal heat exchangers 10A was 1.9%. In Type 2, a similar statistical analysis was performed on a total of 27 horizontal heat exchangers 10A, and the outlier rate was 5.9%.

[0058] It should be noted that the statistical analysis methods used to create the box plot are not limited to the example above (a method of calculating the thinning rate statistics for each heat exchanger 10 and finding the "type-specific median," which is the median of the set of statistics for heat exchangers 10 of the same type). For example, statistics such as the median and third quartile could be calculated for each type of heat exchanger 10 using all the thinning rate data.

[0059] Based on the results of the statistical analysis of the contents described above, the inspection tube determination unit 24 scores a number of pre-set evaluation indices to evaluate the degree of variation in the thinning rate of the tubes 102 (in this disclosure, the evaluation of the degree of variation may also be referred to as "evaluation of variation"). Furthermore, the following calculation is performed on the scores of these evaluation indices, and the larger the calculated value, the more adjustment is made to increase the number of tubes 102 to be inspected. That is, based on the calculated score, the more proportion of tubes 102 whose thinning rate variation deviates from the pre-set range, the more the decision is made to increase the number of tubes to be inspected. The evaluation indices used to determine the number of tubes to be inspected are the median thinning rate (median by type), the upper limit, and the outlier rate of the thinning rate, as shown in Figure 5.

[0060] Figure 6 below shows examples of score settings for the evaluation indicators described above, as well as the calculation process. According to these examples, the score is set so that the score increases as the thinning rate itself (median M) increases, and as the degree of variation in the thinning rate (upper limit UV, outlier rate OR) increases. Furthermore, the calculated value is obtained by adding the product of the scores related to the degree of variation in the thinning rate to the score for the thinning rate itself. In the example shown in Figure 5, the calculated score for Type 1 is "0" and the calculated score for Type 2 is "3", leading to the conclusion that the number of tubes 102 to be inspected should be greater for Type 2 than for Type 1.

[0061] Furthermore, the method for calculating the score in determining the number of tubes 102 to be inspected is not limited to the example above. Other methods may be used as long as they can evaluate the degree of variation between the type of heat exchanger 10 and the degree of thinning of tubes 102 (the "thinning rate" in the example above). For example, the range of possible outlier rate OR values ​​described above could be divided into several categories, and the score could be set so that the score increases as the outlier rate OR value increases.

[0062] Next, we will explain the method for determining the number of tubes to be extracted from Tube 102 based on the score obtained using the method described above. For example, API (American Petroleum Institute) Recommended Practice 581 specifies a method for developing inspection and maintenance plans based on risk, taking into account the probability of equipment failure and the magnitude of the impact if a failure occurs. In this recommendation, the relationship between inspection effectiveness and uncertainty of inspection results is evaluated on a five-point scale from A to E. For example, if the statistical uncertainty of the inspection is small, the inspection effectiveness is evaluated as high, and if the inspection has high uncertainty, the inspection effectiveness is evaluated as low.

[0063] Applying this concept to the inspection of tubes 102 in the heat exchanger 10, the more tubes 102 are inspected, the lower the uncertainty of the inspection becomes, and if all tubes are inspected, the uncertainty becomes 0%. On the other hand, setting an excessive number of tubes to be inspected leads to unnecessary cost increases. Therefore, in this embodiment, the "outlier rate" (the ratio of data points that are outliers exceeding the upper limit out of the total number of thinning rate data obtained from each type of heat exchanger 10), as explained using Figure 5, was adopted as a criterion for measuring the effectiveness of the inspection. The ratio of the number of tubes inspected to the total number of tubes 102 installed in the heat exchanger 10 to be inspected is called the "sampling rate" (if the tube 102 is a U-shaped tube, one tube is counted as two straight tubes 102). At this time, the value "(sampling rate / outlier rate) × 100 [%]" was used as an index to evaluate the effectiveness of the inspection ("inspection effectiveness index").

[0064] Table 1 below shows the relationship between the value of the inspection effectiveness index and the five-level evaluation set in this embodiment. In this embodiment, the inspection planning support device 2 reduces the uncertainty of the inspection by performing the process (2) described later (risk assessment for each area based on the results of statistical analysis). Therefore, in this embodiment, the sampling rate is set so that the inspection effectiveness index is at least 20% or higher. (Table 1) TIFF2026067692000002.tif95152

[0065] According to the correspondence table between the calculated score and the sampling rate of tube 102 shown in the far right column of Figure 6, the sampling rate of tube 102 is changed to 1%, 3%, or 5% depending on the calculated score. On the other hand, even if the sampling rate is set based on this correspondence table, when the test effectiveness index is calculated, the value may exceed 50%, resulting in a rating of C or higher. In this case, the sampling rate at which the test effectiveness index becomes 50% or less can be calculated in reverse, and this reverse-calculated value may be set as the new sampling rate.

[0066] According to the method described above, the sampling rate determined based on the calculated score shown in Figure 6 is compared with the sampling rate determined based on the inverse value of the inspection effectiveness index, and the smaller sampling rate is adopted. When setting the sampling rate based on the inverse value of the inspection effectiveness index, for example, the sampling rate is changed in 1% increments, with a lower limit of 1%. The number of tubes to be inspected can be determined by multiplying the sampling rate obtained in this way by the total number of tubes 102 provided in the heat exchanger 10 to be inspected (processing (1) of the inspection tube determination unit 24).

[0067] Next, a specific example of the process (2) of selecting inspection tubes based on the results of the statistical analysis described above for each area where tubes 102 are located will be explained. The areas are formed by dividing the region where many tubes 102 are located in each type of heat exchanger 10 (horizontal heat exchanger 10A, or vertical heat exchanger 10B, AFC 10C), and each area contains multiple tubes 102.

[0068] Figure 7A shows an example of area settings for a horizontal heat exchanger 10A or a vertical heat exchanger 10B, and Figure 7B shows an example of area settings for an AFC 10C. As shown in these figures, the "area in which the tubes 102 are arranged (hereinafter also simply referred to as the "arrangement area")" is set so that the arrangement positions of the tubes 102 can be individually identified, for example, within the cross-section of the shell 101 in the case of a horizontal heat exchanger 10A or a vertical heat exchanger 10B, and within the cross-section of the tube bundle 114 in the case of an AFC 10C.

[0069] In the examples shown in Figures 7A and 7B, the arrangement area for tubes 102 in each heat exchanger 10 is divided into 16 areas labeled A to P. There are no particular limitations on the method of dividing the arrangement area for tubes 102 or the number of areas formed, but each area is set to contain several to about one hundred tubes 102. If the arrangement area includes a tube 102 consisting of a U-shaped tube, the tube 102 is considered to be divided into two at the curved part of the U-shape. The divided tubes 102 may belong to different areas. Furthermore, when multiple paths 106 are formed in relation to the heat exchanger 10, each area may be set independently of the range of tubes 102 included in each path 106, or each area may be set to coincide with the range of tubes 102 included in each path 106.

[0070] Figure 8 shows the distribution of data where the wall thinning rate measured from the inner surface of the tube 102 was an outlier in the horizontal heat exchanger 10A of type 1 shown on the left side of Figure 5. As previously described, the inspection data includes not only information related to the type classified by the type classification unit 23, but also information indicating the arrangement position of the tube 102 within the heat exchanger 10. Therefore, the distribution map in Figure 8 can be created by combining this data with the results of the statistical analysis used to create the box plot shown in Figure 5.

[0071] The numbers shown in Figure 8 represent the percentage [%] of the number of identical data points in each area A to P relative to the total number of outliers used to create Figure 8. According to Figure 8, the larger the numbers shown in these areas, the greater the proportion of tubes 102 with a thinning rate greater than the upper limit, and the higher the risk of the thinning rate being an outlier for the tubes 102 located in that area.

[0072] Therefore, in this embodiment, the area containing, for example, 8% or more of the total number of outliers evaluated in Figure 8 is designated as the "recommended area." The recommended area is the area containing the tube 102 that should be selected as the inspection tube. In the example shown in Figure 8, areas B, F, N, and O are selected as the recommended areas. The inspection results from the outer surface of the shell 101 are also evaluated in the same way as in Figure 8, and the recommended area is selected (processing (2) of the inspection tube determination unit 24).

[0073] Figure 9 shows the distribution of data where the maximum value of the outlier in the wall thinning rate exceeded 0.05 mm / year, based on the inspection results from the inner surface of tubes 102 of type 1 (60 units in total). The numbers shown in the figure represent the percentage [%] of the number of such data in each area A to P relative to the total number of maximum outlier values ​​used to create Figure 9. Figure 9 allows us to understand which areas tend to have a higher wall thinning rate in the horizontal heat exchanger 10A where the maximum outlier value exceeds a predetermined threshold (0.05 mm / year). In Figure 9, the larger the numbers shown in these areas, the greater the proportion of tubes 102 where the maximum outlier value is greater than the threshold, and it can be evaluated that there is a higher risk of the outlier in the wall thinning rate becoming even larger for the tubes 102 located in those areas.

[0074] Therefore, in this embodiment, areas containing, for example, 8% or more of the total number of outliers evaluated in Figure 9 are also designated as "recommended areas." In the example shown in Figure 9, in addition to areas B, F, N, and O already selected in Figure 8, area C is selected as a recommended area. The inspection results from the outer surface of the shell 101 are also evaluated in the same way as in Figure 9, and recommended areas are selected (processing of the inspection tube determination unit 24 (2)).

[0075] It is not a mandatory requirement to select the recommended area using both methods described in Figures 8 and 9. For example, the recommended area may be selected using either the method described in Figure 8 or Figure 9.

[0076] The inspection tube determination unit 24 selects a recommended area using the process (2) described above, and then selects an inspection tube from within each recommended area. Examples of methods for selecting an inspection tube include selecting the tube 102 with the smallest wall thickness within the entire heat exchanger 10 (all areas), selecting the tube 102 with the smallest wall thickness within a recommended area, and randomly selecting a tube 102 within a recommended area. The method for selecting an inspection tube from a recommended area while satisfying the number of inspection tubes determined in process (1) will be explained later in the operation description of the inspection planning support device 2, with reference to Figures 10A and 10B.

[0077] Furthermore, the inspection tube selection unit 24 may also select inspection tubes for one or more paths 106, which are fluid passages through the inner surface of the tube 102, in addition to the areas described using Figures 7A and 7B. For example, the inspection tube selection unit 24 selects the tube 102 that showed the greatest reduction in wall thickness when its wall thickness was measured last time from each path 106 as the inspection tube.

[0078] In both the case where the area described above is used as the basis and the case where path 106 is used as the basis, the method for selecting the inspection tubes while satisfying the condition for the number of inspection tubes determined in process (1) will be described later in Figures 10A and 10B. However, it is not a mandatory requirement to perform both the selection of inspection tubes based on the area and the selection of inspection tubes based on path 106. The selection of inspection tubes may be performed based on either the area or path 106. For example, if the selection of inspection tubes is performed based only on path 106, the inspection tube determination unit 24 will also perform a risk assessment on path 106 using a method similar to at least one of the methods described using, for example, Figures 8 and 9, and will determine the tube 102 located in a high-risk position as the inspection tube.

[0079] The number of tubes to be inspected used when determining which tubes to inspect may be calculated to be greater than the number of passes 106 in the heat exchanger 10 to be inspected. In this case, the inspection tube determination unit 24 can, for example, determine the number of inspection tubes for the heat exchanger 10 to be inspected as follows. First, for each pass 106, the inspection tube determination unit 24 determines the tube 102 with the greatest wall thinning or the smallest wall thickness from among the tubes 102 belonging to that pass 106 as the inspection tube. Next, if the number of tubes 102 determined as inspection tubes is less than the number to be inspected, the inspection tube determination unit 24 further determines a tube 106 located in a high-risk position as an inspection tube. If the number of tubes 102 determined as inspection tubes is equal to or greater than the number to be inspected, the inspection tube determination unit 24 terminates the inspection tube determination process (does not determine a new tube 106 as an inspection tube).

[0080] Furthermore, the risk assessment may be performed for each heat exchanger 10 under inspection, for each pass 106, or for each area. Also, when the inspection tube determination unit 24 determines the tube 102 with the greatest wall thinning for each pass 106, if there are multiple tubes 102 with the same amount of wall thinning, it may count these multiple tubes 102 together as one inspection tube, or it may count these multiple tubes 102 individually as separate tubes. Also, when the inspection tube determination unit 24 determines the tube 102 with the smallest wall thickness for each pass 106, if there are multiple tubes 102 with the same wall thickness, it may count these multiple tubes 102 together as one inspection tube, or it may count these multiple tubes 102 individually as separate tubes.

[0081] The processes performed by the type classification unit 23 and the inspection tube determination unit 24 described above are stored as a computer program (inspection planning support program) that executes these inspection planning support functions in the storage unit 22, which is a storage medium within the inspection planning support device 2, or in an external storage medium. The computer program is then read from this storage medium and installed in the inspection planning support device 2.

[0082] The operation of the inspection planning support device 2 will be explained with reference to Figures 10A and 10B. Figure 10A is a first flowchart illustrating an example of the process for determining the tube to be inspected. Figure 10B is a second flowchart illustrating an example of the process for determining the tube to be inspected. The order and content of each step described below can be changed as appropriate.

[0083] First, information regarding the wall thickness of the heat exchanger 10 installed in plant 1 is acquired as raw data via the information acquisition unit 21 and stored in the storage unit 22 (start in Figure 10A). Next, the type classification unit 23 performs type classification of the heat exchanger 10 using this raw data (step S101).

[0084] Subsequently, the inspection tube determination unit 24 performs statistical analysis and risk assessment on the inspection data, which has been modified with information on the type of heat exchanger 10, as explained using Figures 5 to 9, to determine the sampling rate and recommended area of ​​the tubes 102 (step S102). The information on the recommended area is stored, for example, in the storage unit 22. Next, the information acquisition unit 21 acquires data on the total number of tubes 102 for each heat exchanger 10 that is the subject of inspection, for example, from an external source. The inspection tube determination unit 24 then multiplies the total number by the sampling rate to calculate the number of tubes to be inspected (step S103).

[0085] Furthermore, the information acquisition unit 21 acquires data from an external source regarding the number of passes and the arrangement of passes 106 for each heat exchanger 10 that is the subject of inspection. The inspection tube determination unit 24 then determines whether the number of tubes to be inspected calculated earlier is equal to or greater than the number of passes (step S104). If the number of tubes to be inspected is less than the number of passes (step S104: NO), the inspection tube determination unit 24 makes adjustments to add tubes 102 to the number of tubes to be inspected (step S105) and performs a comparison with the number of passes again (step S104).

[0086] If the number of tubes to be inspected is greater than or equal to the number of passes (step S104: YES), the inspection tube determination unit 24 determines whether the number of tubes to be inspected is greater than or equal to the number of recommended areas (step S106). If the number of tubes to be inspected is less than the number of recommended areas (step S106: NO), the inspection tube determination unit 24 adjusts by adding tubes to be inspected for tube 102 (step S107) and performs a comparison with the number of passes again (step S106). If the number of tubes to be inspected is greater than or equal to the number of recommended areas (step S106: YES), the inspection tube determination unit 24 determines the number of tubes to be inspected for tube 102 with this number (step S108).

[0087] Next, the information acquisition unit 21 acquires data from an external source regarding the inspection history of the individual heat exchangers 10 that are to be inspected. Then, the inspection tube determination unit 24 selects the tubes 102 that had the smallest wall thickness measured from either the inner or outer surface during the previous inspection for the heat exchangers 10 whose number of tubes to be inspected has been determined, as the inspection tubes (step S109).

[0088] Subsequently, the information acquisition unit 21 acquires information on recommended areas that have been pre-selected by the inspection tube determination unit 24 and stored in the memory unit 22, etc., according to the type of individual heat exchanger 10 to be inspected. The inspection tube determination unit 24 then selects one inspection tube from each of the recommended areas for the heat exchanger 10 to be inspected (step S110, the process of determining the tube 102 located in a high-risk position as the inspection tube). At this time, since the tube 102 with the minimum wall thickness in the entire heat exchanger 10 has already been selected, the inspection tube determination unit 24 can exemplify the next selection method by selecting the tube 102 with the minimum wall thickness within each recommended area. Alternatively, if the tube 102 with the minimum wall thickness within a recommended area has already been selected, a tube 102 may be selected randomly within the recommended area.

[0089] Next, the test tube determination unit 24 sets the counter count n to "n=1" (step S111) and determines whether or not n (1 at this stage) or more tubes 102 have been selected for all paths 106 (step S112). If at least one tube 102 has not been selected from all paths 106 (step S112: NO), the test tube determination unit 24 selects one test tube from each of the remaining paths 106 (step S113).

[0090] Subsequently, the inspection tube determination unit 24 re-determines whether one or more tubes 102 have been selected for all passes 106. If the condition is met (step S112: YES), it determines whether there are any remaining tubes to be inspected (step S114). If there are remaining tubes (step S114: YES), the inspection tube determination unit 24 increments the counter count by one (step S115), and repeats the determination and processing from steps S112 to S114 according to the count (n≧2).

[0091] Here, an example can be given where the inspection tube selection unit 24 selects the tube 102 with the minimum wall thickness within each pass 106. Alternatively, if the inspection tube selection unit 24 has already selected the tube 102 with the minimum wall thickness within each pass 106, it may randomly select a tube 102 as the inspection tube within each pass 106. After selecting the inspection tubes from each pass 106, the inspection tube selection unit 24 completes the inspection tube selection operation (end) once it confirms that there are no remaining inspection tubes (step S114: NO).

[0092] Figure 11 shows the arrangement positions of the inspection tubes determined by the operations described above. The inspection tube determination unit 24 can display the arrangement diagrams of the selected inspection tubes for the horizontal heat exchanger 10A, vertical heat exchanger 10B, and AFC 10C on the input / output terminal 3, in response to a user request. As shown in the legend in Figure 11, marks or hatches may be added to the tubes 102 in the arrangement diagram to identify the reason why each tube 102 was selected as an inspection tube.

[0093] The inspection planning support device 2 of this embodiment has the following advantages. Based on the results of statistical analysis of information regarding the wall thickness of tubes 102 in a heat exchanger 10 of the same type as the multi-tube heat exchanger 10 to be inspected, it identifies tubes 102 located in positions with a high risk of wall thinning, so that the tubes 102 to be inspected can be determined individually.

[0094] Here, the configuration of the inspection planning support device 2 is not limited to the case where it is configured using a computer used by the user, as shown in the example in Figure 1. For example, as illustrated in Figure 12, the inspection planning support system that performs the aforementioned processes (1) and (2) and the selection of inspection tubes may be configured on a cloud computing system 20.

[0095] In this case, the cloud computing system 20 provides the functions of the type classification unit 23 and the inspection tube determination unit 24. Figure 12 also shows an example in which the storage unit 22 for storing raw data, etc., is located outside the cloud computing system 20. However, this configuration can be changed as appropriate, and the cloud computing system 20 may also provide the functions of the storage unit 22. [Explanation of symbols]

[0096] 1 Plant 10 Heat exchanger 10A Horizontal Heat Exchanger 10B Vertical Heat Exchanger 10C Air Fin Cooler (AFC) 101 Shell 102 Tube 103 Fixed tube sheet 104 Baffle 105 Partition Plate 106 Pass 110 fans 111 Moving blade 112 Rotation axis 113 Duct 114 Tube Bundle 115 Drive unit 2. Inspection planning support device 20 Cloud Computing Systems 21 Information Acquisition Department 22 Memory section 23. Type Classification Section 24. Inspection tube determination section 3 Input / Output Terminals

Claims

1. An inspection planning support device that assists in planning inspections for multi-tube heat exchangers equipped with numerous tubes, An inspection planning support device is characterized by comprising an inspection tube determination unit that, based on the results of statistical analysis of inspection data that includes the type common to the heat exchanger to be inspected, among inspection data that associates information on the wall thickness of each of the many tubes contained in the heat exchanger with the type of the heat exchanger and the position of the tubes arranged in the heat exchanger, determines from among the many tubes arranged in the heat exchanger to be inspected that are located in positions where there is a high risk of wall thickness reduction exceeding a preset threshold, as inspection tubes to be inspected.

2. The inspection planning support device according to claim 1, characterized in that the type is classified based on the type of heat exchanger and the phase state of the heat supply fluid and heat receiving fluid flowing on the inner or outer side of the tube.

3. The inspection planning support device according to claim 2, characterized in that the aforementioned type is selected from horizontal, vertical, or air fin cooler (AFC).

4. The inspection data includes information regarding the wall thickness measured from both the inner and outer surfaces of the tube, The inspection planning support device according to claim 1, characterized in that the statistical analysis is based on the inspection data for the inner and outer surfaces of the tube, respectively.

5. In the aforementioned statistical analysis, the region in the heat exchanger where the numerous tubes are arranged is divided into multiple areas, each containing multiple tubes. The inspection planning support device according to claim 1, characterized in that, for a plurality of divided areas, the higher the proportion of tubes in which the wall thickness is greater than the threshold, the higher the risk is considered to be for the tubes placed in that area.

6. The inspection planning support device according to claim 5, characterized in that, in the statistical analysis, an area in which the proportion is greater than the threshold is selected from among the divided areas as a recommended area including the inspection tube.

7. In the aforementioned statistical analysis, the variability in the amount of tube wall thinning in the inspection data was evaluated for each type. The inspection planning support device according to claim 1, characterized in that the inspection tube determination unit adjusts the number of inspection tubes to be inspected based on the evaluation of the variability.

8. The inspection tube determination unit is characterized in that the number of tubes to be inspected increases as the proportion of tubes that deviate in the direction in which the wall thinning becomes greater than a preset range in the evaluation of the variation increases.

9. The inspection planning support device according to claim 8, characterized in that the evaluation of the variability is performed using a box plot created from the inspection data with an upper limit set, and that the thinning of the wall material deviates in a direction that is greater than a preset range, by the fact that the thinning of the wall material is an outlier that is greater than the upper limit.

10. The heat exchanger to be inspected includes a path through which the heat supply fluid or heat receiving fluid passes on the inner surface side of the tube, When determining the number of inspection tubes to be inspected, which is predetermined for the heat exchanger to be inspected, The aforementioned inspection tube determination unit is, Based on the inspection data, for each pass, the tube with the greatest wall thinning or the smallest wall thickness among the tubes belonging to that pass is determined to be the inspection tube. When the number of tubes determined is less than the number of tubes to be inspected, the tubes located in the high-risk positions are further determined to be the inspection tubes. The number of inspections is greater than the number of passes. The inspection planning support device according to feature 1.

11. The inspection planning support device according to claim 1, further comprising a type classification unit that identifies which of the above types the heat exchanger from which the inspection data was measured belongs, based on statistical quantities obtained from the information regarding the wall thickness.

12. The inspection planning support device according to claim 11, characterized in that the type classification unit performs type classification of the heat exchanger using a learning model created by pre-machine learning the correspondence between the statistical quantity and the type.

13. The inspection planning support device according to claim 12, characterized in that the type classification unit creates a learning model by machine learning the correspondence between the type and a statistical quantity selected from a group of statistical quantities consisting of maximum value, minimum value, data range, mean, median, interquartile range, first quartile, third quartile, and standard deviation, with respect to the wall thickness information obtained for the inner or outer surface of the tube, using the random forest method.

14. An inspection planning support method that assists in planning the inspection of a multi-tube heat exchanger equipped with numerous tubes, An inspection planning support method characterized by including a step of determining, based on the results of statistical analysis of inspection data that includes the same type as the heat exchanger to be inspected, from among the many tubes arranged in the heat exchanger to be inspected, the tubes that are located in positions where there is a high risk of wall thickness reduction exceeding a preset threshold, from among the many tubes arranged in the heat exchanger to be inspected, as inspection tubes to be inspected.

15. A storage medium that stores a computer program that enables a computer to perform inspection planning support to assist in the planning of inspections of a multi-tube heat exchanger equipped with many tubes, The storage medium is characterized in that the computer program has a set of steps arranged to execute the inspection planning support method described in claim 14.

16. An inspection planning support program that assists in planning inspections of multi-tube heat exchangers equipped with numerous tubes, Computers, An inspection tube determination unit determines, based on the results of a statistical analysis of inspection data that includes the same type as the heat exchanger being inspected, from among the many tubes arranged in the heat exchanger being inspected, the tubes that are located in positions where there is a high risk of wall thickness reduction exceeding a preset threshold, from among the many tubes arranged in the heat exchanger being inspected, as inspection tubes to be inspected. A testing plan support program that functions as such.

Images