JP6844969B2 - Maintenance support equipment and maintenance support programs - Google Patents

Description

The present invention relates to a maintenance support device and a maintenance support program.

Conventionally, as a technique for determining a maintenance method, inspection specifications, etc. of a structure of a plant, for example, an estimated value of corrosion rate is corrected according to the type of the structure, and the corrected estimated value is used to correct the structure. There is a technique for estimating the life of a structure from the period of use and the plate thickness (see, for example, Non-Patent Document 1). Here, as the maintenance method, for example, there are preventive maintenance and post-maintenance. Further, the inspection specifications include, for example, an inspection method for preventive maintenance, an inspection range, an inspection frequency, and an inspection time.

"Analysis of actual machine inspection results of external surface corrosion under heat insulating materials in chemical plants and examination of probability estimation method", Materials and Environment, 59, 291-297 (2010)

However, in the technique described in Non-Patent Document 1, although the life of the structure can be estimated, there is no viewpoint of calculating the probability of reaching the life, that is, the probability of the structure becoming unusable. Therefore, it has been difficult to appropriately determine the maintenance method and inspection specifications of the structure.
Focusing on the above points, it is an object of the present invention to provide a maintenance support device and a maintenance support program capable of presenting the probability that a plant structure becomes unusable.

In order to solve the above problems, measurement one aspect of the maintenance support system of the present invention, a resulting unit preparative you get a number in association with measurements of degree of damage of the structure of the plant and their attributes, obtained by the obtaining unit grouping value attribute for each the group, to measure each belonging to the group, as the measured value according to a predetermined probability distribution to calculate the scaling variable corresponding to the cumulative probability of the probability distribution calculating parts and, for each attribute and the thickness of the combination of a predetermined structure, Hazuki group standardized variables calculated by the calculation unit, the life end thereof structures by damage of the structure is a probability of unusable A calculation unit for calculating the probability is provided, and the attribute includes the usage period of the structure .

According to one aspect of the present invention, it is possible to provide a maintenance support device capable of presenting the probability that the structure of the plant becomes unusable (probability of reaching the end of life).

It is a block diagram of the maintenance support device which concerns on embodiment. It is a scatter diagram which shows the relationship between the attribute (use period) and the measured value x. It is a figure which shows the Gumbel distribution which shows the relationship between the measured value x and the probability density function f. It is a table showing the relationship between the cumulative probability F and the reference variable y. It is a scatter diagram which shows the relationship between the measured value x and the reference variable y. It is a table which shows the relationship between a predetermined cumulative probability Fo and a predetermined standardization variable yo. It is a table showing the relationship between a predetermined cumulative probability Fo and the first wall thinning depth x1. It is a scatter diagram which shows the relationship between the mean value xa of the use period, and the 1st wall thinning depth x1. It is explanatory drawing for demonstrating the weighting of the 1st wall thinning depth × 1. It is a table showing the relationship between a predetermined cumulative probability Fo and the second wall thinning depth x2. It is a table which shows the relationship between the 2nd wall thinning depth x2 and a predetermined standardization variable yo. It is a table (first table) showing the relationship between the second standardization variable y2, the predetermined thinning depth xx, and the predetermined usage period uo. It is a table (second table) showing the relationship between the cumulative probability Fs after standardization, the predetermined wall thinning depth xx, and the predetermined usage period uo. It is a table (third table) showing the relationship between the life arrival probability P, the plate thickness x3, and the predetermined usage period uo. It is a table (third table) showing the relationship between the life arrival probability P, the margin wall thickness, and the predetermined usage period uo. It is a figure which shows the image data which made the coordinates with respect to the combination of the predetermined use period uo and the plate thickness x3 into the color which represents the life arrival probability P. It is a figure which shows the image data which made the coordinates with respect to the combination of the temperature of a structure and the plate thickness x3, which represents the lifetime arrival probability P, which concerns on the modification. It is a figure which shows the image data which made the coordinates with respect to the combination of the type of a structure and the plate thickness x3 which corresponded to the modification example into the color which represents the lifetime arrival probability P. It is a figure which represents the image data which made the coordinates with respect to the combination of the predetermined use period uo and the plate thickness x3 as the color which represents the life arrival probability P for each type of structure. It is a block diagram of the maintenance support device which concerns on a modification. It is a table showing the calculation results such as the estimated cost, the inspection priority, and the inspection cost guideline. It is a table showing the calculation result of the forecast cost, the inspection priority, the inspection cost guideline, etc. related to the modification.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
The present embodiment applies the present invention to a maintenance support device that supports determination of maintenance methods, inspection specifications, and the like of a plant structure. This maintenance support device calculates the probability that the structure will reach the end of its life, that is, the probability that the structure will become unusable due to damage to the structure (hereinafter, also referred to as "life end probability") P. Structures include, for example, pipes, heat exchangers, and towers. In this embodiment, as a structure, a pipe or the like covered with a heat insulating material is targeted. Further, as the damage, the corrosion under the heat insulating material that occurs on the outer surface of the structure (pipe, etc.) is targeted.
It should be noted that the embodiments shown below exemplify devices and methods for embodying the technical idea of the present invention, and the technical idea of the present invention describes the shape, structure, arrangement, etc. of components. It is not specific to the following. The technical idea of the present invention may be modified in various ways within the technical scope specified by the claims.

(Constitution)
As shown in FIG. 1, the maintenance support device 1 of the present embodiment includes an input device 2 such as a keyboard and a mouse, a display device 3 for displaying image data, and a device main body (hereinafter, "computer") for performing various arithmetic processes. (Also called) 4 is provided. The computer 4 is composed of a CPU (not shown), an HDD (Hard Disk Drive), and the like. The HDD stores one or more programs that realize various arithmetic processes. Further, the CPU executes various arithmetic processes according to one or more programs stored in the HDD. When the input device 2 starts the calculation of the life arrival probability P, the CPU executes the life arrival probability calculation process according to the program stored in the HDD. In the life arrival probability calculation process, the program (in a broad sense, the "maintenance support program") causes the computer 4 to function as the acquisition unit 410, the calculation unit 420, and the calculation unit 430.

The acquisition unit 410 acquires a large number of measured values x of the wall thinning depth (in a broad sense, “damage degree”) of the structure of the plant in association with the attributes. As the measured value x, for example, the maximum value (local depth) of the depth of the corroded portion is used. The attributes include, for example, numerical attributes such as the period of use and temperature of the structure, the type of structure, the environment (distance from the sea, distance from the cooling tower, annual rainfall, etc.), the presence or absence of coating, and the type of coating. There is at least one of the qualitative attributes (either alone or in combination thereof) such as the composition of the internal solution, the companion and flow velocity, the presence or absence of contact with the gantry, the presence or absence of coating with a heat insulating material, and the type of heat insulating material. In this embodiment, the usage period is adopted as an attribute. Further, as the measured value x, for example, data acquired at the time of inspection of various plants can be adopted. The acquisition unit 410 stores the acquired measured value x and the like in the HDD of the computer 4.
It is not realistic to perform inspections at various plants in order to acquire a large number of measured values x after the start of the life arrival probability calculation process, and in fact, they are acquired at the time of past inspections at various plants. The measured value x is collected (acquired) in advance.

As shown in FIG. 2, the calculation unit 420 has a plurality of measurements stored in the HDD of the computer 4 in a two-dimensional coordinate system in which the horizontal axis corresponds to the attribute (use period) and the vertical axis corresponds to the measured value x. Create scatter plot image data that plots coordinates consisting of a set of values x and attributes (use period). Then, the created image data is displayed on the display device 3. As a result, the operator is made to determine the interval of grouping according to the attribute (use period) based on the displayed image data.
Subsequently, when the operator operates the input device 2 to input the grouping interval, the calculation unit 420 inputs a plurality of measured values x stored in the HDD of the computer 4 based on the input interval, that is, , The plurality of measured values x acquired by the acquisition unit 410 are grouped by the attribute (use period). In this embodiment, it is divided into 6 groups of 10 to 15 years, 16 to 20 years, 21 to 25 years, 26 to 30 years, 31 to 35 years, and 36 to 40 years. Subsequently, for each of those groups, the standardized variable y corresponding to the cumulative probability F of the probability distribution is set so that the measured value x (random variable) follows a predetermined probability distribution for each of the measured values x belonging to the group. calculate.

Here, as the probability distribution, various distributions such as a Gunbell distribution (maximum value distribution, double exponential distribution), a wibble distribution, a normal distribution, an exponential distribution, and a beta distribution can be adopted. In this embodiment, as shown in FIG. 3, the Gumbel distribution is adopted. The cumulative probability F is a value obtained by accumulating the probability density function f of the probability distribution from the smaller measured value x (random variable). As a method for calculating the cumulative probability F, for example, at least one of an average rank method, a median rank method, and a mode rank method can be adopted. In this embodiment, the average rank method is adopted.
In the average rank method, first, as shown in FIG. 4, the measured values x are arranged in ascending order for each group, and a number indicating the rank (hereinafter, also referred to as “data number”) is associated with each of the arranged measured values x. Subsequently, for each of the measured values x, the cumulative probability F is calculated according to the following equation (1) based on the associated data number and the total number of data in the group (maximum value of the data number).
F = (data number) / (total number of data + 1) ……… (1)
Subsequently, the calculation unit 420 converts the calculated cumulative probability F into the standard variable y of the Gumbel distribution according to the following (2). In the example of FIG. 4, the reference variable y is about -1 to +6.
y = -ln (-ln (F)) ……… (2)

As shown in FIG. 1, the calculation unit 430 includes a first regression equation calculation unit 431, a first damage degree calculation unit 432, a second regression equation calculation unit 433, and a second damage degree calculation unit 434. A third regression equation calculation unit 435, a second variable calculation unit 436, a life arrival probability calculation unit 437, and a presentation unit 438 are provided.
As shown in FIG. 5, the first regression equation calculation unit 431 has a measured value x stored in the HDD of the computer 4 for each group set by the calculation unit 420, that is, a measured value acquired by the acquisition unit 410. A regression equation (hereinafter, also referred to as “first regression equation”) representing the relationship between x and the reference variable y corresponding to the measured value x is calculated. In this embodiment, a linear approximation formula is adopted as the first regression equation. Further, as the calculation method of the first regression equation, the least squares method is adopted. Subsequently, the first regression equation calculation unit 431 calculates the coefficient of determination R ² for the first regression equation.
Subsequently, the first regression equation calculation unit 431 associates the calculated first regression equation with a representative value of the usage period that defines the group. Typical values for the period of use include, for example, an average value, a median value, and a mode value. In this embodiment, an average value is adopted as a representative value of the usage period.

Subsequently, as shown in FIG. 5, the coordinates consisting of the set of the measured value x and the standardized variable y are set in the two-dimensional coordinate system in which the horizontal axis corresponds to the measured value x and the vertical axis corresponds to the standardized variable y. Plot and create scatter plot image data on which the approximate straight line of the first regression equation and the coefficient of determination R ^{2 are drawn.} Then, the created image data is displayed on the display device 3. Thus, the original image data (coefficient of determination R ^2, etc.) that is displayed, the operator, first regression equation was calculated to determine whether it is appropriate. For example, when the coefficient of determination R ² is 0.5 or more, preferably 0.7 or more, it is determined that the first regression equation is appropriate. It should be noted that such a determination may be made by the calculation unit 430.

Subsequently, when the operator operates the input device 2 to input the determination that the first regression equation is appropriate, the first damage degree calculation unit 432 is as shown in FIG. 6 according to the following equation (3). The standardized variable (hereinafter, also referred to as “predetermined standardized variable”) yo corresponding to the predetermined cumulative probability (hereinafter, also referred to as “predetermined cumulative probability”) Fo is calculated. In the example of FIG. 6, 0.5, 0.9, 0.95, and 0.99 are adopted as the predetermined cumulative probabilities Fo, and 0.366513, 2.250367, 2.970195, as the predetermined reference variable yo. 4.600149 is calculated.
y = -ln (-lnF)) ……… (3)

Subsequently, as shown in FIGS. 5 to 7, a plurality of first damage degree calculation units 432 are used for each of the first regression equations calculated by the first regression equation calculation unit 431 according to the first regression equation. A plurality of wall thinning depths (hereinafter, also referred to as "first wall thinning depths") x1 corresponding to a predetermined standardized variable yo are calculated. As a result, a plurality of first wall thinning depths x 1 corresponding to a plurality of predetermined cumulative probabilities Fo = 0.5, 0.9, 0.95, 0.99 (in a broad sense, "first degree of damage". ") Is calculated.
On the other hand, when the operator operates the input device 2 and inputs the determination that the first regression equation is not appropriate, the first damage degree calculation unit 432 repeats the above procedure from the grouping in the scatter diagram of FIG. Do.

As shown in FIG. 8, the second regression equation calculation unit 433 has a predetermined cumulative probability Fo (that is, a predetermined reference) for each of the plurality of predetermined cumulative probability Fos (that is, a predetermined reference variable yo). A regression equation representing the relationship between a plurality of first wall thinning depths x1 corresponding to the conversion variable yo) and an average value xa of a plurality of usage periods corresponding to the plurality of first wall thinning depths x1 (hereinafter). , Also called "second regression equation"). In this embodiment, a power curve approximation formula is adopted as the second regression equation. Further, as the calculation method of the second regression equation, the least squares method is adopted. Subsequently, the second regression equation calculation unit 433 calculates the coefficient of determination R ² for the second regression equation.

The second regression equation calculation unit 433 weights the first wall thinning depth x1 to calculate the second regression equation. In this embodiment, as shown in FIG. 9, the first wall thinning depth x1 used in the calculation of the second regression equation is weighted. As a weighting method, a method of setting the weight according to the number of measured values x is adopted. Specifically, the average value xa of the usage period is selected from the average value xa of a plurality of usage periods, and the larger the measured value x of the usage period represented by the average value xa of the selected usage period, the more the selected usage period. The weight of the first wall thinning depth x1 corresponding to the average value xa is increased.

Subsequently, as shown in FIG. 8, the horizontal axis corresponds to the usage period and the vertical axis corresponds to the first thinning depth x1 in the two-dimensional coordinate system, and the first thinning depth x1 and the attribute (use period). ) Is plotted, and the image data of the scatter plot in which the approximate curve of the second regression equation and the coefficient of determination R ^{2 are drawn is created.} Then, the created image data is displayed on the display device 3. As a result, the operator is made to judge whether the calculated second regression equation is appropriate based on the displayed image data. For example, when the coefficient of determination R ² is 0.5 or more, preferably 0.7 or more, it is determined that the second regression equation is appropriate. It should be noted that such a determination may be made by the calculation unit 430.

Subsequently, when the operator operates the input device 2 to input the determination that the second regression equation is appropriate, the second damage degree calculation unit 434 receives the second damage degree calculation unit 434 as shown in FIGS. 8 and 10. For each of the second regression equations calculated by the regression equation calculation unit 433, a plurality of reductions corresponding to a plurality of predetermined usage periods (hereinafter, also referred to as “predetermined usage periods”) according to the second regression equation. The wall depth (hereinafter, also referred to as “second wall thinning depth”) × 2 is calculated. In the example of FIG. 10, 15 years, 20 years, 25 years, 30 years, 35 years, and 40 years are adopted as predetermined usage periods. Subsequently, the second damage degree calculation unit 434 applies the calculated second wall thinning depth x2 (in a broad sense, "second damage degree") to the second regression equation used for the calculation. The corresponding predetermined cumulative probability Fo is associated.
On the other hand, when the operator operates the input device 2 and inputs the determination that the second regression equation is not appropriate, the second damage degree calculation unit 434 repeats the above procedure from the grouping in the scatter diagram of FIG. Do.

As shown in FIG. 11, the third regression equation calculation unit 435 has, for each of the plurality of predetermined usage periods uo, a plurality of second wall thinning depths x2 corresponding to the predetermined usage period uo, and a plurality of them. Regression equation (hereinafter, also referred to as “third regression equation”) expressing the relationship with a plurality of predetermined reference variables yo (that is, predetermined cumulative probability Fo) corresponding to the second wall thinning depth × 2 of calculate. In this embodiment, a linear approximation formula is adopted as the third regression equation. Further, as the calculation method of the third regression equation, the least squares method is adopted.

Subsequently, as shown in FIG. 11, a second thinning depth x is defined in the two-dimensional coordinate system in which the horizontal axis corresponds to the second thinning depth x2 and the vertical axis corresponds to the reference variable y. The coordinates consisting of the set with the reference variable yo of are plotted, and the image data of the scatter plot in which the approximate straight line of the third regression equation is drawn is created. Then, the created image data is displayed on the display device 3. As a result, the operator is made to judge whether the calculated third regression equation is appropriate based on the displayed image data. Here, it can be said that FIG. 11 is a scatter plot in which the scatter plot of FIG. 5 is standardized (arranged by using regression or the like).

Subsequently, when the operator operates the input device 2 to input the determination that the third regression equation is appropriate, the second variable calculation unit 436 makes a third regression as shown in FIGS. 11 and 12. For each of the third regression equations calculated by the formula calculation unit 435, a plurality of wall thinning depths (hereinafter, also referred to as “predetermined wall thinning depths”) xo corresponding to a predetermined thickness according to the third regression equation. The standardized variable (hereinafter, also referred to as “second standardized variable”) y2 is calculated. In the example of FIG. 12, as the predetermined wall thinning depth xo (in a broad sense, “predetermined degree of damage”), for example, 2 [mm], 3 [mm], 4 [mm], 5 [mm], 6 Nine pieces of [mm], 7 [mm], 8 [mm], 9 [mm], and 10 [mm] are adopted. Based on these calculation results, the table shown in FIG. 12 (hereinafter, also referred to as “first table”) is formed.
On the other hand, when the operator operates the input device 2 and inputs the determination that the third regression equation is not appropriate, the second variable calculation unit 436 repeats the above procedure from the grouping in the scatter diagram of FIG. ..

As shown in FIG. 13, the life arrival probability calculation unit 437 sets the second standardized variable y2 (the second standardized variable y2 in the first table) calculated by the second variable calculation unit 436 as follows ( 4) Convert to Fs, which is the cumulative probability of the Gunbell distribution (hereinafter, also referred to as "cumulative probability after standardization") according to Eq. As a result, the table shown in FIG. 13 (hereinafter, also referred to as “second table”) is formed.
Fs = exp (−exp (−y2)) ……… (4)

Subsequently, as shown in FIG. 14, the life arrival probability calculation unit 437 converts the converted cumulative probability Fs after standardization (cumulative probability Fs after standardization in the second table) into the life arrival probability P according to the following equation (5). Convert. Subsequently, the predetermined usage period uo is defined as the usage period u1 of the target structure. Subsequently, assuming a situation in which the wall thinning depth becomes equal to the plate thickness and a through hole is formed and becomes unusable, the plate of the target structure is set to the predetermined wall thinning depth xo of the second table. The thickness is x3. As a result, a table (hereinafter, "third table") representing the life arrival probability P is formed for each combination of the predetermined attribute of the structure (use period of the structure) and the plate thickness x3.
P = 1-F ……… (5)

In the present embodiment, in the third table (FIG. 14), an example in which the predetermined wall thinning depth xx of the second table (FIG. 13) is used as it is as the plate thickness x3 is shown, but other configurations are shown. Can also be adopted. For example, in consideration of the lower limit of the wall thickness (hereinafter, also referred to as “required wall thickness”) in which a through hole is not formed on the outer surface of the structure due to the pressure of a liquid or the like in the structure, a second table (FIG. As shown in FIG. 15, the margin wall thickness calculated by subtracting the required wall thickness from the predetermined wall thickness reduction depth xo of 13) is used instead of the plate thickness x3 of the third table (FIG. 14). It may be.

As shown in FIG. 16, the presentation unit 438 has a horizontal axis corresponding to an attribute (structure usage period u1) and a vertical axis based on the third table (FIG. 14) generated by the life arrival probability calculation unit 437. Create image data in which each coordinate of the two-dimensional coordinate system corresponding to the plate thickness x3 is colored to represent the attribute (structure usage period u1) represented by the coordinate and the life arrival probability P corresponding to the plate thickness x3. .. In this image data, the horizontal axis corresponds to the attribute (use period u1 of the structure), the vertical axis corresponds to the plate thickness x3, and the height direction axis corresponds to the life arrival probability P. It can be said that it is image data when viewed. As for the color of each coordinate, for example, when the life arrival probability P is ^{larger than 10 -1} [1 / (year / location)], it is red, and 10 ^-1 to ^10-2 [1 / (year / location)]. ] Is orange, 10 ^-2 to 10 ^-3 [1 / (year / location)] is yellow, and 10 ^-3 to 10 ^-4 [1 / (year / location)] is yellow. Adopt green. In the example of FIG. 16, each color of red, orange, and green is represented by hatching. Here, as the color, one in which the luminous intensity of any one color is changed, such as gray scale, may be used. Further, a pattern such as hatching may be used together with the color or instead of the color. Further, in the example of FIG. 16, although a margin is provided between the color regions of each coordinate, the size of each color region may be increased so that no margin is provided between the color regions.

Then, the presentation unit 438 causes the display device 3 to display the created image data. As a result, based on the displayed image data, the operator is notified of the life arrival probability P of the structure (piping, etc.) in colors corresponding to the usage period u1 of the target structure (piping, etc.) and the plate thickness x3. Let me grasp.
Here, when planning an inspection in an actual plant, first of all, there is a budget that can be spent on the inspection, and it is important to decide which structure should be inspected within the budget. This is because when inspecting in a chemical plant, etc., it is necessary to install a large-scale scaffolding for the inspection, so the inspection cost tends to be high, and every year, all the structures of the plant are inspected. Because it is not realistic to do. Therefore, by using the table as shown in FIG. 16 obtained by the present embodiment, it is possible to make a decision as to which structure should be preferentially inspected by using the life arrival probability P as a determination index. That is, it is possible to support the determination of the maintenance method, inspection specifications, and the like. This can greatly contribute to the decision maker making rational decisions, as compared with the method that relies on the intuition of a skilled worker.

(Effect of this embodiment)
(1) As described above, in the present embodiment, the measured value x of the wall thinning depth of the acquired structure (pipe, etc.) is divided into groups according to the period of use, and the measured value x is measured for each group. Assuming that the value x follows the Gumbel distribution, the reference variable y corresponding to the cumulative probability F of the Gumbel distribution is calculated. Subsequently, the probability that the structure (piping, etc.) becomes unusable (life arrival probability P) was calculated for each combination of the predetermined usage period uo and the plate thickness x3. Thereby, it is possible to present the probability (life arrival probability P) that the structure (piping, etc.) becomes unusable due to damage to the structure. In addition, the maintenance method and inspection specifications of structures (piping, etc.) can be determined by statistical methods regardless of the inspector's intuition and experience. Therefore, it is possible to more appropriately determine the maintenance method, inspection specifications, etc. of the structure (piping, etc.). Further, since the life arrival probability P can be obtained, it is easy for the budget decision-maker of the plant to understand the necessity of inspection.

(2) In addition, the attributes are the period of use of the structure, temperature, type, environment, presence / absence of coating and type of coating, composition of internal solution, companion and flow velocity, presence / absence of contact with pedestal, and presence / absence of coating with heat insulating material. Since at least one of these (single or a combination thereof) is used, it is possible to group by factors having a large influence on the corrosion rate, and the life arrival probability P can be calculated with higher accuracy.
(3) Further, the second regression equation calculation unit 433 continues to weight the first wall thinning depth x1 to calculate the second regression equation. Therefore, the weight of the highly accurate data can be increased to calculate the second regression equation, so that the second regression equation can be calculated with higher accuracy.

(4) Further, since the second regression equation calculation unit 433 uses a power curve approximation equation as the second regression equation, the degree of matching can be improved and the second regression equation can be calculated with higher accuracy.
(5) Further, in the calculation unit 420, at least one of the Gumbel distribution and the Weibull distribution is used as the probability distribution, and the average rank method is used as the calculation method of the cumulative probability F. Therefore, the measured value x of the wall thinning depth of the structure (pipe, etc.) due to corrosion can be expressed more appropriately.

(6) Further, in the presentation unit 438, the attribute (the coordinate represents each coordinate of the two-dimensional coordinate system in which one axis corresponds to the attribute (predetermined usage period uo) and the other axis corresponds to the plate thickness x3. Image data is presented as either a color or a small number of patterns representing the life arrival probability P corresponding to the predetermined usage period uo) and the plate thickness x3. Therefore, for example, even if the life arrival probability P differs greatly for each coordinate, the operator can appropriately grasp the life arrival probability P.

(Modification example)
(1) In the present embodiment, an example in which the usage period of the structure (pipe, etc.) is used as an attribute is shown, but other configurations can also be adopted. For example, the temperature of a structure (pipe, etc.) can be used as an attribute. In this case, the calculation unit 420 groups the measured values x by temperature. Further, the calculation unit 430 calculates the life arrival probability P for each combination of temperature and plate thickness x3. As a result, as shown in FIG. 17, it is possible to generate image data in which the life arrival probability P is represented by a color or a pattern for each combination of temperature and plate thickness x3.

Further, for example, the qualitative attribute of the structure (pipe, etc.) may be used as the attribute. In this case, the calculation unit 420 groups the measured values x by qualitative attributes. Further, the calculation unit 430 calculates the life arrival probability P for each combination of the qualitative attribute of the structure (pipe, etc.) and the plate thickness x3. As a result, as shown in FIG. 18, image data representing the life arrival probability P in color or pattern can be generated for each combination of the qualitative attribute of the structure (pipe, etc.) and the plate thickness x3. In the example of FIG. 18, the type (piping, heat exchanger, tower, etc.) is used as the qualitative attribute.

Further, for example, as an attribute, a combination of the period of use of the structure and the type of the structure (piping, heat exchanger, tower, etc.) may be used. In this case, the calculation unit 420 groups the measured values x according to the combination of the usage period and the type. Further, the calculation unit 430 calculates the life arrival probability P for each combination of the usage period and the plate thickness x3 for each type of structure. As a result, as shown in FIG. 19, image data representing the life arrival probability P in color or pattern can be generated for each combination of the attribute and the plate thickness of the structure for each type of structure. In the example of FIG. 19, as attributes, (a) uses a pipe, (b) uses a heat exchanger, and (c) uses a tower.

(3) Further, in the present embodiment, an example used for determining the maintenance method, inspection specifications, etc. of Corrosion Under Insulation (CUI) of the structure coated with the heat insulating material is shown, but other maintenance is shown. It can also be used to determine the method and inspection specifications. For example, maintenance methods and inspection specifications for gantry contact parts (corrosion at the gantry contact part between pipes and gantry, etc.), pitting corrosion due to heat medium cooling water, crevice corrosion origin stress corrosion cracking, erosion corrosion, etc. Etc. may be used to determine. In this case, the degree of damage is the depth of wall thinning due to corrosion of the gantry contact part, the depth of wall thinning due to pitting corrosion, the depth of cracking due to stress corrosion cracking, the depth of cracking due to stress corrosion cracking at the origin of crevice corrosion, and the depth of wall thinning due to erosion. At least one of the depth and the wall thinning depth due to erosion corrosion will be used.

(4) Further, in the present embodiment, an example is shown in which image data representing the life arrival probability P in color or pattern is presented for each combination of the attribute of the structure (piping, etc.) of the plant and the plate thickness x3. Other configurations may be adopted. For example, it may be configured to present the amount of damage that occurs.
Specifically, in this modification, as shown in FIG. 20, the program of the life arrival probability calculation processing uses the computer 4 as the prediction cost estimation unit 439, the priority setting unit 440, the guideline calculation unit 441, and the expected value presentation. It is also designed to function as a unit 442.
The estimated cost estimation unit 439 estimates the expected value of the damage amount (hereinafter, also referred to as “estimated cost”) based on the life arrival probability P calculated by the life arrival probability calculation unit 437. As a method of calculating the estimated cost, for example, the life arrival probability P is multiplied by the number of corroded parts estimated in the structure (piping, etc.) (hereinafter, also referred to as "estimation of the number of CUI parts per system"), and the calculation method thereof is performed. A method is used in which the amount of damage incurred when a through hole is formed in a structure (pipe, etc.) (hereinafter, also referred to as “leakage cost”) is multiplied by the multiplication result.

The priority setting unit 440 sets the inspection priority of each structure (piping, etc.) and each member constituting the structure (piping, etc.) based on the predicted cost calculated by the forecast cost estimation unit 439. For example, the higher the forecast cost, the higher the inspection priority.
The guideline calculation unit 441 sets the estimated cost calculated by the forecast cost estimation unit 439 as a guideline for the upper limit of the cost for inspecting a structure (piping, etc.) (hereinafter, also referred to as “inspection cost guideline”).
As shown in FIG. 21, the expected value presenting unit 442 sets the estimated cost estimated by the estimated cost estimation unit 439, the inspection priority set by the priority setting unit 440, and the guideline calculation unit for each structure (piping, etc.). The image data of the fourth table representing the inspection cost guideline set in 441 is displayed on the display device 3. In the example of FIG. 21, in addition to the predicted cost, the inspection priority, and the inspection cost guideline, the piping number, the plate thickness, the period of use, the life arrival probability P, the estimation of the number of CUI points per system, the true penetration probability, and the leakage cost are also included. I am presenting it.
As a result, the operator can more easily determine the priority, timing, and budget distribution of the inspection of the structure (piping, etc.) based on the displayed image data, and the maintenance method of the structure (piping, etc.). And inspection specifications can be determined more appropriately. In addition, since the forecast cost, inspection priority, and inspection cost guideline can be obtained, it is easy for the plant budget decision-maker to understand the necessity of inspection.

(5) Further, the acquisition unit 410 may be configured to acquire the measured value of the degree of damage per unit area or unit length. In this case, based on the acquired measured values, the calculation unit 420 and the calculation unit 430 perform the above calculation to calculate the damage (thickness) probability per unit area or unit length, and the size of the structure (thickness reduction). The life arrival probability P can be calculated according to the area (area, length). Therefore, for example, when calculating the predicted cost, as shown in FIG. 22, the life arrival probability P is multiplied by the size of the structure, and the multiplication result is multiplied by the leakage cost. As a result, the predicted cost, the inspection priority, the inspection cost guideline, and the like can be calculated relatively easily from the size of the structure. In the example of FIG. 22, the surface area outside the pipe is adopted as the size of the structure.

Here, the estimated value of the degree of damage per unit area or unit length (measured value acquired by the acquisition unit 410) is, for example, the degree of damage of each structure (pipe, etc.) (for example, the wall thinning depth by CUI). ) May be derived based on the distribution. As the distribution of the degree of damage of each structure (piping, etc.), for example, one structure (for example, piping system) has a plurality of (for example, two or more points) of damage (for example, the measured value of the wall thinning depth by CUI). In some cases, a distribution of degree of damage (eg, Gunbell distribution) obtained assuming a unit area or unit length of measurement can be adopted.

1 ... maintenance support device, 2 ... input device, 3 ... display device, 4 ... computer, 410 ... acquisition unit, 420 ... calculation unit, 430 ... calculation unit, 431 ... first regression equation calculation unit, 432 ... first Damage degree calculation unit 433 ... Second regression equation calculation unit 434 ... Second damage degree calculation unit 435 ... Third regression equation calculation unit 436 ... Second variable calculation unit 437 ... Lifespan arrival probability calculation Department, 438 ... Presentation unit, 439 ... Forecast cost estimation unit, 440 ... Priority setting unit, 441 ... Estimated calculation unit, 442 ... Expected value presentation unit

Claims (11)

An acquisition unit that acquires a large number of measured values of the degree of damage to a plant structure in association with its attributes,
The measured values acquired by the acquisition unit are grouped according to the attributes, and for each of the measured values belonging to the group, the measured values follow a predetermined probability distribution, and the probability distribution is accumulated. A calculation unit that calculates the standardized variables corresponding to the probabilities,
For each of the predetermined combinations of structure attributes and plate thickness, the life arrival probability, which is the probability that the structure becomes unusable due to damage to the structure, is calculated based on the standardized variable calculated by the calculation unit. comprising a calculation unit for, a,
The attribute is a maintenance support device including the period of use of the structure. The degree of damage is the depth of wall reduction due to corrosion under the heat insulating material, the depth of wall reduction due to corrosion at the contact part of the gantry, the depth of wall reduction due to erosion corrosion, the crack depth due to stress corrosion cracking, and the cracking due to crevice corrosion origin stress corrosion cracking. The maintenance support device according to claim 1, wherein the depth is at least one of a depth, a wall thinning depth due to erosion, and a wall thinning depth due to erosion corrosion. The attributes are the period of use of the structure, the temperature, type, environment, presence / absence of coating and coating type, composition of internal solution, companion and flow velocity, presence / absence of contact with pedestal, presence / absence of coating with heat insulating material, and so on. The maintenance support device according to claim 1 or 2, further comprising at least one of the types of heat insulating materials. The attribute consists of a combination of the period of use of the structure and the type of structure.
The calculation unit groups the measured values according to the combination of the period of use and the type.
The maintenance support device according to claim 3, wherein the calculation unit calculates the life arrival probability for each combination of a usage period and a plate thickness for each type of structure. Before Symbol computing unit,
For each group set by the calculation unit, a first regression equation representing the relationship between the measured value and the standardized variable corresponding to the measured value is calculated, and the first regression equation is used for the first regression equation. The first regression equation calculation unit that associates the representative values of the usage period that defines the group, and
For each of the first regression equations, a first damage degree calculation unit that calculates a plurality of first damage degrees corresponding to a plurality of predetermined cumulative probabilities according to the first regression equation.
For each of the plurality of predetermined cumulative probabilities, the plurality of the first damage degrees corresponding to the predetermined cumulative probabilities and the representative values of the plurality of usage periods corresponding to the plurality of first damage degrees. A second regression equation calculation unit that calculates a second regression equation that represents a relationship,
For each of the second regression equations, a plurality of second damage degrees corresponding to a plurality of predetermined usage periods are calculated according to the second regression equation, and for each of the plurality of second damage degrees, A second damage degree calculation unit that associates the predetermined cumulative probability corresponding to the second regression equation, and a second damage degree calculation unit.
For each of the plurality of predetermined usage periods, a plurality of the second damage degrees corresponding to the predetermined usage period and a plurality of the predetermined cumulative probabilities corresponding to the plurality of the second damage degrees. A third regression equation calculation unit that calculates a third regression equation that represents a relationship,
For each of the third regression equations, a second variable calculation unit that calculates a plurality of second reference variables corresponding to a plurality of predetermined damage degrees according to the third regression equation.
3. The third aspect of the present invention is characterized in that, for each combination of the predetermined usage period and the plate thickness, a life arrival probability calculation unit for calculating the life arrival probability based on the second reference variable is provided. The maintenance support device according to 4. The fifth aspect of claim 5, wherein the second regression equation calculation unit weights the first degree of damage according to the number of measured values to calculate the second regression equation. Maintenance support device. The maintenance support device according to claim 5 or 6, wherein the second regression equation calculation unit uses a power curve approximation equation as the second regression equation. The probability distribution is at least one of the Gumbel distribution and the Weibull distribution.
The maintenance support device according to any one of claims 1 to 7, wherein at least one of an average rank method, a median rank method, and a mode rank method is used as the method for calculating the cumulative probability. The maintenance support device according to any one of claims 1 to 8, wherein the acquisition unit acquires a measured value of the degree of damage per unit area or unit length. In the calculation unit, each coordinate of the two-dimensional coordinate system in which one axis corresponds to an attribute and the other axis corresponds to a plate thickness is a color representing the lifetime arrival probability corresponding to the attribute represented by the coordinate and the plate thickness. The maintenance support device according to any one of claims 1 to 9, further comprising a presentation unit for presenting image data in at least one of the symbols. Computer,
Acquisition unit, which acquires a large number of measured values of a plant structure in association with its attributes.
The acquired measured values are grouped according to the attributes, and each group corresponds to the cumulative probability of the probability distribution, assuming that the measured values follow a predetermined probability distribution for each of the measured values belonging to the group. Calculation unit that calculates the standardized variable,
The probability of reaching the end of life, which is the probability that the structure becomes unusable due to damage to the structure, based on the standardized variables calculated by the calculation unit for each of the predetermined combinations of the attributes and the plate thickness of the structure. together to function as a calculator for calculating a
The attribute is a maintenance support program that includes the period of use of the structure.

Images