JP6251201B2 - Occupancy rate prediction device and availability rate prediction method - Google Patents

Description

  The present invention relates to an operation rate prediction apparatus and an operation rate prediction method for predicting an operation rate of a mechanical system including a plurality of components.

  In recent years, various plant systems such as thermal power plants, nuclear power plants, and chemical plants, as well as various mechanical systems such as aircraft, rockets, and ships have been increasingly required to guarantee operating rates, and have a competitive edge with competitors. In order to strengthen it, it is necessary to improve the operation rate. For this reason, it is important for machine system manufacturers to accurately determine whether or not the guaranteed operating rate can be achieved, and to take appropriate measures to improve the operating rate. It is necessary to establish a method for quantitatively and accurately evaluating the rate.

  As a conventional technique for quantitatively and accurately evaluating the operation rate of a mechanical system, there is a technique called RAM analysis. In the RAM analysis, the maintenance history data is analyzed from the three viewpoints of reliability, availability, and maintainability based on maintenance history data to determine whether the maintenance plan is performed in an optimal state with respect to productivity and cost. It is something to evaluate. Specifically, based on the system diagram of the entire mechanical system, a reliability block diagram showing the linkage of failures among the components of the mechanical system is created taking into account the redundancy of the mechanical system. Input rate data, recovery time, etc. and perform simulation to evaluate the operating rate of the entire mechanical system.

  In addition, there is a technique described in Patent Document 1 as a technique applicable to quantitatively evaluating the operation rate. In the technique described in Patent Document 1, a failure association expansion tree in which failure events assumed in the equipment constituting the plant are expanded in a tree is created, and the corresponding device is based on the operation history information using the failure association expansion tree. The unreliability (probability of failure by a certain time) is predicted by the probabilistic life evaluation means.

JP 2003-303243 A

  Since the above prior art uses failure record data such as failure rate, recovery time, and operation history information for predicting the operation rate and failure rate, devices with similar or identical configurations are already in operation and the failure record is If it is abundantly accumulated, the operation rate and failure rate can be accurately predicted by referring to the failure record. However, in the mechanical system including the new technology, there is a problem that it is difficult to accurately predict the operation rate with the above-described conventional technology because data that can be referred to as a failure record is not sufficiently accumulated.

  In addition, there are many types of failures in the initial stage of operation, initial failures in which the number of occurrences rapidly decreases with time, accidental failures that occur with a certain probability related to operating time, and operating time is advanced and near life. If so, there is a failure due to aging, where the number of occurrences increases rapidly. In the above-described RAM analysis, only accidental failures are considered, and the technique of Patent Document 1 does not focus on such a failure mode. Therefore, in any technique, the failure rate that changes according to the operation time is accurately determined. Unpredictable.

  The present invention was devised in view of the problems as described above, and an object thereof is to provide an operation rate prediction device and an operation rate prediction method capable of accurately predicting an operation rate based on a small amount of actual data. To do.

  [1] In order to achieve the above object, an operation rate prediction apparatus of the present invention is an operation rate prediction device that predicts an operation rate of a prediction target machine system including a plurality of components, and the plurality of configurations Each element has a map storage means in which a level is evaluated for an evaluation item related to a failure rate, and a map defining a correspondence relationship between the level of the evaluation item and a failure coefficient is stored, and the evaluation item of the component A failure coefficient setting means for setting the failure coefficient of the component based on the map, a failure rate estimation means for estimating a failure rate of the component using the failure coefficient, Operating rate estimating means for estimating the operating rate of the prediction target machine system based on the failure rate of each of a plurality of components, and the map includes the operating rate estimating means It is characterized in that it is fitted to fit more estimated actual operation rate similar mechanical system.

[2] the map storage means, the initial and early failure map for calculating a failure factor, and random failures map for calculating the random failure coefficients, for aging failure for calculating the aging failure factor And the failure coefficient setting means sets the initial failure coefficient for the component using the initial failure map, and sets the random failure coefficient using the random failure map. and, using said aging failure map sets the aging failure coefficients, the failure rate estimating means estimates the initial failure rate by using the initial failure coefficient, using said random failure factor total and estimates the contingent failure rate, calculates through annual deterioration failure rate by using the aging failure factor, the failure rate, the initial failure rate, the random failure rate and the aging failure rate To calculate as It is preferred.

  [3] The map storage means stores, as the initial failure map, a map that defines a correspondence relationship between the initial failure coefficient, the magnitude of the load of the component, and the novelty of the component, It is preferable that the failure coefficient setting means sets the initial failure coefficient of the component based on the initial failure map based on the load of the component and the novelty of the component.

  [4] The map storage means defines, as the random failure map, a correspondence relationship between the random failure coefficient, the occurrence frequency of the abnormal load of the component, and the magnitude of the influence of the abnormal load of the component The failure coefficient setting means stores the random failure of the component according to the frequency of occurrence of the abnormal load of the component and the magnitude of the influence of the abnormal load of the component based on the random failure map. It is preferable to set a coefficient.

  [5] The map storage means stores, as the aging deterioration fault map, a map that defines a correspondence relationship between the aging failure coefficient and design uncertainty of the component, and the failure coefficient setting means includes: It is preferable to set the aging failure coefficient of the component based on the design uncertainty of the component based on the aging failure map.

[6] the and the failure rate of the components, based on the similar mechanical system said component has failed in the recovery time it took to recover from actual and cumulative downtime results, each of said component estimating the influence on stop before SL prediction target machine system for, preferably comprises a influence estimating means.
[7] the operation information before Ki予 measuring target machine system acquires the failure information from the operation information storage means for periodically storing, providing the failure coefficient correction means for updating the fault factor based on said failure information Is preferred.

[8] Before Ki予 measuring target machine system is a plant, it is preferable that the similar mechanical system is similar plant.
[9] To achieve the above object, a method of operating a plant operating rate predicting apparatus of the present invention predicts the rate of operation of the machine system comprising a plurality of components, the operating method of operating Ritsu予 measuring device The failure factor and the evaluation related to the failure rate so that the performance rate of the similar machine system can be reproduced based on the failure rate of each of the plurality of components estimated using the failure factor. Creating a map that defines the correspondence with the level of the item, the first step, and setting the failure coefficient of the plurality of components based on the map from the level for the evaluation item of the component, respectively, estimating the failure rate of a plurality of components, a second step to calculate the rate of operation of the prior Symbol machinery system based on each of the failure rate of the plurality of components, a third step It is characterized by comprising.

According to the present invention, the level of the evaluation item related to the failure rate is evaluated for each of the plurality of constituent elements, and the map based on the actual operation rate of the similar mechanical system is calculated from the level for each of the plurality of constituent elements. Use it to set the failure factor and thus calculate the failure rate. And based on the failure rate of this some component, the operation rate of a prediction object machine system is estimated.
The operation rate of the target machine system is finally predicted using the level of the evaluation items evaluated for each of the plurality of components and the map based on the actual operation rate of similar machine systems. The future operating rate can be predicted accurately based on the rate performance.

It is a typical side view which shows the whole structure of the coal burning boiler which concerns on each embodiment of this invention. It is a functional block diagram which shows the structure of the operation rate prediction apparatus of the plant which concerns on 1st Embodiment of this invention. It is a schematic diagram which shows an example of the data set which concerns on the operation rate prediction apparatus of the plant which concerns on 1st Embodiment of this invention. It is a figure for demonstrating the calculation method of the initial failure rate h1 (t) based on 1st Embodiment of this invention, Comprising: (a) is a figure which shows the risk matrix for initial failure, (b) is (a). It is a figure which shows the initial stage failure rate curve created using the risk matrix for initial faults. It is a figure for demonstrating the calculation method of the random failure rate h2 (t) based on 1st Embodiment of this invention, Comprising: (a) is a figure which shows the risk matrix for random failures, (b) is (a). It is a figure which shows the accidental failure rate line created using the risk matrix for accidental failures. It is a figure for demonstrating the calculation method of the aged failure rate h3 (t) based on 1st Embodiment of this invention, Comprising: (a) is a figure which shows the risk matrix for aged failure, (b) is (a). It is a figure which shows the secular failure rate curve produced using the risk matrix for secular failure of. It is a schematic diagram which compares and shows the predicted value of the failure rate by the operation rate prediction apparatus of the plant as 1st Embodiment of this invention, and the predicted value of the failure rate by the conventional method. It is a schematic diagram which compares the predicted value of the plant operating rate by the plant operating rate prediction apparatus as 1st Embodiment of this invention, the predicted value of the plant operating rate by the conventional method, and the performance of a plant operating rate. . It is a functional block diagram which shows the structure of the operation rate prediction apparatus of the plant which concerns on 2nd Embodiment of this invention. It is a schematic diagram which shows the influence degree to the plant stop which concerns on 2nd Embodiment of this invention for every component apparatus. It is a functional block diagram which shows the structure of the operation rate prediction apparatus of the plant which concerns on 3rd Embodiment of this invention. It is a schematic diagram which shows the example of an input of the driving diary system which concerns on 3rd Embodiment of this invention.

Embodiments of the present invention will be described below with reference to the drawings. Note that each embodiment described below is merely an example, and there is no intention of excluding various modifications and technical applications that are not explicitly described in the following embodiment. Each configuration of the following embodiments can be implemented with various modifications without departing from the spirit thereof, and can be selected as necessary or can be appropriately combined.
In the following embodiments, an example in which the operation rate prediction device and the operation rate prediction method of the present invention are applied to a boiler plant will be described.

[1. First Embodiment]
[1-1. Overall configuration of boiler plant]
An overall configuration of a boiler plant (hereinafter also referred to as a prediction target plant) 1 as a prediction target machine system according to the present invention will be described with reference to FIG.
The boiler plant 1 has a boiler 2 that uses coal as fuel. The boiler 2 includes a furnace wall 4 in which a furnace (combustion chamber) 3 is formed, and a pulverized coal burner (hereinafter referred to as an upper stage, a middle stage, and a lower stage) installed in multiple stages (herein, the upper stage, the middle stage, and the lower stage). 5), an additional air nozzle 6 attached near the outlet of the furnace 3 and supplying air for two-stage combustion to the furnace, a flue 7 connected to the outlet of the furnace 3, and an upper part of the furnace 3 A superheater 8, a reheater 9, and an economizer 10 arranged in this order toward the flue 7, and a drum 11 provided at the upper portion of the furnace 3 are provided.

The furnace wall 4 is configured as a water-cooled wall in which a plurality of water pipes (not shown) are arranged with the pipe axis direction in the vertical direction. The upper and lower ends of each water pipe are connected to the drum 11 via a header, a riser pipe and a downcomer pipe not shown.
The boiler plant 1 further includes pulverized coal supply means 12 for supplying pulverized coal to the burner 5, air supply means 13 for supplying combustion air to the burner 5 and the additional air nozzle 6, and combustion exhaust gas discharged from the boiler 2. And an exhaust gas line 14 for discharging the gas to the outside.
Although the pulverized coal supply means 12 is omitted for the upper and middle burners 5 in FIG. 1, it is provided for each stage of the burners 5. The pulverized coal supply means 12 is interposed in the pulverized coal pipe 12c at the exit of the pulverized coal machine 12b, the pulverized coal pipe 12c connecting the bunker 12a, the pulverized coal machine 12b, the pulverized coal machine 12b, and the burner 5. The on-off valve 12d is provided.

  Connected to the air supply means 13 are a forced air blower 13 a that pressurizes and supplies air, a wind box 13 b provided outside the furnace wall 4, a forced air blower 13 a, the wind box 13 b, and the additional air nozzle 6. An air duct 13c and an air side of the regenerative air preheater 13d are provided.

The exhaust gas line 14 includes an exhaust gas duct 14a, a denitration device 14b interposed in this order from the upstream side in the flow direction of the combustion exhaust gas, the combustion exhaust gas side of the regenerative air preheater 13d, the electrostatic precipitator 14c, and attraction. A ventilator 14d, a desulfurization device 14e, and a chimney 14f connected to the downstream end of the exhaust gas duct 14a are provided.
That is, the boiler plant 1 includes the above-described components (components) 1 to 14 shown in FIG.

With the above configuration, in the boiler plant 1, in the boiler 2, the air taken in from the outside by the forced air blower 13a is pressurized and sent out, and this air circulates through the air duct 13c and is regenerative air preheater 13d. After being heated by exchanging heat with the combustion exhaust gas, the air is supplied to each burner 5 through the wind box 13b and also supplied to the additional air nozzle 6.
In addition, the coal of the coal bunker 12a is supplied to the pulverized coal machine 12b and pulverized to obtain pulverized coal having a size suitable for combustion (for example, several μm to several hundred μm). The pulverized coal is mixed with pressurized conveying air supplied from an air source (not shown) to become a pulverized coal mixture, and is conveyed to the burner 5 through the pulverized coal pipe 12c.

  The burner 5 injects combustion air and pulverized coal mixture into the furnace 3. The combustion air and pulverized coal mixture injected from the burner 5 form a flame at the injection port of the burner 5 and generate combustion gas. The combustion gas ascends the furnace 3, passes through the superheater 8 and the reheater 9 in this order, is attracted by the induction fan 14d, and is discharged from the furnace 3 to the flue 7.

  The combustion exhaust gas that has passed through the economizer 10 of the flue 7 flows through the exhaust gas duct 14a, nitrogen oxide (NOx) is removed by the denitration device 14b, and heat is exchanged with the combustion air by the regenerative air preheater 13d to generate a heat quantity. Is recovered (cooled), removed by the electrostatic precipitator 14c, sulfur oxide (SOx) is removed by the desulfurizer 14e, and then discharged from the chimney 14f to the atmosphere.

  Further, water supplied from a water supply pump (not shown) is preheated by exchanging heat with combustion gas in an economizer 10 disposed in the flue 7 and then sent to the furnace wall 4. The water sent to the furnace wall 4 is heated by the radiant heat of the flame of the burner 5 and the sensible heat of the combustion gas, and becomes a gas-liquid two-phase flow composed of water and saturated steam. This gas-liquid two-phase flow is separated into saturated steam and saturated water by the drum 11. The saturated water is sent back to the furnace wall 4 and again heated. The saturated steam is introduced into the superheater 8, heated by the radiant heat of the flame of the burner 5 and the sensible heat of the combustion gas, and becomes superheated steam. This superheated steam is supplied to the steam turbine for power generation via the main steam pipe 8a, and the steam taken out during the expansion process in the steam turbine is introduced into the reheater 9 via the low temperature reheat steam pipe. Then, it is superheated again by the combustion gas and supplied again to the steam turbine via the high temperature reheat steam pipe (steam turbine, low temperature reheat steam pipe and high temperature reheat steam pipe are not shown).

[1-2. Configuration of plant operating rate prediction device]
The configuration of the plant operation rate prediction apparatus 20 according to the first embodiment of the present invention will be described with reference to FIG.

First, an outline of the plant operation rate prediction apparatus 20 of the present embodiment will be described.
The operation rate prediction apparatus 20 predicts the operation rate of the prediction target plant 1 (hereinafter also referred to as a plant operation rate) based on the failure rate of each component device of the prediction target plant 1. The operation rate predicting device 20 determines the failure rate of each component device based on the failure rate of the initial failure (hereinafter referred to as the initial failure rate), the failure rate of the accidental failure (hereinafter referred to as the accidental failure rate), and aged deterioration. It is calculated as the total value of failure rates (hereinafter referred to as aging degradation failures) of failures (hereinafter referred to as aging degradation failures).

  Here, “initial failure” is a failure mode that occurs in the early stage of operation due to potential defects such as design failure, manufacturing failure, material selection failure, and installation failure that appear after implementation. The “accidental failure” is a failure mode that randomly occurs in time and is a failure mode that occurs at a substantially constant failure rate regardless of the operation time. “Aging failure” is a failure mode mainly occurring in the second half of the operation period due to aging deterioration such as wear and fatigue.

  The initial failure rate, the accidental failure rate, and the aging failure rate are the risk matrix (initial failure map) 24a for the initial failure, the risk matrix for accidental failure (map for the accidental failure) 24b, and the risk for the aging failure. It is calculated using a matrix (aged deterioration failure map) 24c. These risk matrices 24a, 24b, and 24c are fitted so as to meet the failure performance of similar plants. The risk matrices 24a, 24b, and 24c will be described in detail later.

Here, the similar plant (similar mechanical system) referred to in the present application refers to a plant in which a technology common to or similar to the new technology in the prediction target plant 1 is used. Typical similar plants are the demonstration plant and the test plant of the target plant 1.
The demonstration plant is intended to at least one of scale-down and simplification of the configuration with respect to the prediction target plant 1. Similarly, the test plant is one in which at least one of the scale-down and the simplification of the configuration is attempted with respect to the prediction target plant 1, but the degree of the scale-down and the simplification of the configuration is larger than that of the demonstration plant. Say. In this embodiment, a demonstration plant is used as a similar plant that refers to a failure record.
The new technology refers to a technology that is employed for the first time in the prediction target plant 1 or a similar plant, or a technology that has almost no adoption record other than the prediction target plant 1 and a similar plant.

Hereinafter, the plant operation rate prediction apparatus 20 of the present embodiment will be described in detail.
As shown in FIG. 2, the operating rate prediction apparatus 20 includes an input means 21 composed of a keyboard, a mouse, a recording medium reader, etc., an evaluation information storage means 22 composed of software, a risk matrix creation means 23, A risk matrix storage means (map storage means) 24, failure coefficient setting means 25, failure rate calculation means (failure rate estimation means) 26, plant operation rate calculation means (plant operation rate estimation means) 27, a monitor, a printer, and the like. The external output means 28 is provided.

  Note that the input unit 21, the evaluation information storage unit 22, and the risk matrix creation unit 23 are not essential as components of the operation rate prediction device 20, and may be devices independent of the operation rate prediction device 20. In this case, a risk matrix may be created by another device comprising the input means 21, the evaluation information storage means 22 and the risk matrix creation means 23, and stored in the risk matrix storage means 24 of the operation rate prediction device 20.

By obtaining the evaluation information of the demonstration plant that is already operating from the input means 21, a data set 22 </ b> A as shown in FIG. 3 is created and stored in the evaluation information storage means 22.
This data set 22A is an FMEA sheet composed of items necessary for FMEA (Failure Mode and Effect Analysis), and the side-by-side data constitutes one data set. Hereinafter, this data set 22A is referred to as FMEA sheet 22A.
The FMEA sheet 22A will be further described. The FMEA sheet 22A includes “component equipment”, “new technology”, “designed expected life (service life)”, “failure mode”, “degree of influence of failure on operating rate”, “failure matrix” for each failure record. "And" Effects of measures "are entered and stored.

In the item “component device”, the name of the component device in which the failure has occurred is entered. In the “new technology” item, “1” is input when the component device is evaluated to include the new technology, and “2” is input when the component device is evaluated not to include the new technology. Have been entered. In the “designed expected life (service life)” item, the expected life at the design stage of the corresponding component device is entered.
In the item of “failure type”, whether “initial failure”, “accidental failure”, or “aging deterioration” (aging deterioration failure) is evaluated and inputted.

  The item “degree of influence on operation rate due to failure” is composed of two items, “operational output P_a” and “recovery time t_r”. In the “operable output P_a”, an output that can be generated by the demonstration plant at the time of failure of the component device is input, and 0 (zero) is input when the plant trips. In “Recovery time t_r”, the time required to recover the constituent devices is entered.

  The item of “Fault Matrix” is divided into three items of “Initial Failure Evaluation”, “Random Failure Evaluation”, and “Aging Degradation Evaluation”. The evaluation in these items is evaluated based on the failure status and the opinions of a person who has knowledge about the demonstration plant or the predicted target plant 1 such as a developer or a contractor (hereinafter also referred to as a knowledge person). In particular, component devices with “1” entered in the “new technology” field are carefully interviewed by experts and carefully evaluated.

The “initial failure evaluation” is an item evaluated when the “failure mode” is “initial failure”. “Initial failure evaluation” has two evaluation items of “load magnitude X_1” and “newness X_2”.
“Load magnitude X_1” is an evaluation of the load level of the component device. The load referred to in this application is a load in a broad sense, and not only the output load of the component device but also the component device such as the load received by the component device, the environmental temperature of the component device, and the environmental corrosion of the component device. It includes stress. When the load is evaluated to be smaller than expected, “1” meaning low load is input to the evaluation item for “load size X_1” and the load is evaluated as expected Is inputted with “2” meaning medium load, and “3” meaning high load is inputted when it is evaluated that the load is larger than expected.

  "Novelty X_2" is an evaluation item based on two viewpoints, whether the component equipment is mass-produced or non-mass-produced, or whether it has been used in the same type of plant (here, boiler plant) . If the evaluation item for “load size X_1” is evaluated as a mass-produced component, “novel X_2” must be low (in other words, high reliability) If “1” meaning “No” is input and it is evaluated that the component equipment is a non-mass produced product and has a track record, “2” meaning “Novelty X_2” is in the middle level is input. When it is evaluated that the device is a non-mass produced product and has no track record, “3”, which means an evaluation that “novel X_2” is at a high level (in other words, low reliability), is input.

The “accidental failure evaluation” is an item evaluated when the “failure type” is “accidental failure”. The evaluation items of “accidental failure evaluation” include two evaluation items of “abnormal load occurrence frequency Y_1” and “abnormal load influence magnitude Y_2”.
The “abnormal load occurrence frequency Y_1” is an evaluation of the frequency with which the component device is subjected to an excessive load. In the evaluation item of “abnormal load occurrence frequency Y_1”, if this occurrence frequency is evaluated as low frequency (for example, occurrence frequency of once or less in 20 years), “1” is input, and medium frequency (for example, 20 “2” is input if it is evaluated as an occurrence frequency exceeding once a year and not more than once a month), and “3” if it is evaluated as a high frequency (for example, an occurrence frequency exceeding once a month). Is entered.

  “Magnitude of influence of abnormal load Y_2” is an evaluation item that affects the plant when the corresponding component equipment receives an excessive load. The standard is whether the plant will trip. If the plant is evaluated not to trip regardless of how many times the corresponding component equipment is subjected to the abnormal load, the item “The magnitude of the influence of the abnormal load Y_2” If the level is evaluated to be low and “1” is input and the corresponding component is repeatedly subjected to abnormal load multiple times, it is evaluated that the plant will trip. If it is evaluated that the level is medium level and “2” is input and the corresponding component equipment is subjected to an abnormal load only once, it will be evaluated that the plant will trip. Is evaluated as being at a high level, and “1” is input.

The item of “aging deterioration evaluation” is an item evaluated when the “failure mode” is “aging deterioration”. The item of “Aging degradation assessment” has one evaluation item of “Design uncertainty Z”.
Here, since it takes a considerable amount of time until a failure occurs due to aging deterioration, there is no track record of aging deterioration failure even in the demonstration plant, so the aging deterioration is caused by the variation in the load of the component equipment and the variation in the strength of the component equipment. Is evaluated. This is because the greater the variation in the load on the component equipment and the variation in the strength of the component equipment, the more likely the design conditions that were assumed to deviate from the actual conditions. This is because the aging failure rate increases. In the present invention, the variation in the load of the component device and the variation in the strength of the component device are expressed as “design uncertainty Z”. The variation in the load of the component equipment is evaluated based on the degree of progress of the deterioration of the component equipment in the demonstration plant, and the variation in the strength of the component equipment is evaluated based on the database obtained by the material test.

Furthermore, in the present embodiment, the “design uncertainty Z” takes into account whether or not the cause of the progress of aging deterioration earlier than expected is clear. This is because when the cause of aging deterioration is unknown, the progress of aging deterioration may be further accelerated for an unexpected reason.
Specifically, the evaluation item of “Design Uncertainty Z” is evaluated to be “1” because “Design Uncertainty Z” is evaluated as a low level if the progress rate of aging deterioration is as expected. If it is entered and the rate of progress of aging is higher than expected, but the cause is clear, “Design Uncertainty Z” is evaluated as a medium level and “2” is entered, and the progress of aging is progressing. If the speed exceeds the assumption and the cause is unknown, the “design uncertainty Z” is evaluated as a high level and “3” is input.

  In the item of “Effects of measures”, “A” is entered for failures that have been evaluated as being able to be prevented by reviewing the learning and work procedures even if a failure occurs. “B” is input for a failure that is evaluated as being preventable, and “C” is input for a failure that is not able to take countermeasures or that is evaluated to still have a risk even if countermeasures are taken. When the “effect of the countermeasure” is evaluated as “A” or “B” and the countermeasure is taken for the prediction target plant 1, this failure result is calculated based on the operation rate W (t ) Is not reflected in the forecast.

"Load magnitude X_1" and "Novelty X_2" in "Initial failure evaluation", "Abnormal load occurrence frequency Y_1" and "Abnormal load influence magnitude Y_2" in "Random failure evaluation", and "Aging deterioration" Each evaluation item of “Design Uncertainty Z” in “Evaluation” is “1”, “2”, or “3” in order of increasing influence (risk) on the initial failure rate, accidental failure rate, and aging failure rate. (In other words, the higher the risk of failure, the higher the value.)
That is, each of these evaluation items X_1, X_2, Y_1, Y_2, and Z is an evaluation item related to the initial failure rate, the accidental failure rate, and the aging failure rate in the present invention, and these evaluation items X_1, X_2, Y_1 , Y_2, and Z, the numerical values “1”, “2”, and “3” are levels for the evaluation items in the present invention.

  If no initial failure occurred in the component device, the “load magnitude X_1” and “novelty X_2” in the component device are evaluated as “1” with the lowest failure risk. If no accidental failure has occurred, the “abnormal load occurrence frequency Y_1” and “abnormal load impact magnitude Y_2” of the component equipment are evaluated as “1” with the lowest risk, respectively. If there is no sign of aging, the “design uncertainty Z” of the component is rated as “1” with the lowest risk.

  Therefore, each component device has “load magnitude X_1”, “newness X_2”, “abnormal load occurrence frequency Y_1”, “abnormal load impact magnitude Y_2”, and “design uncertainty Z”. One of “1”, “2”, and “3” is evaluated. And evaluation about the component apparatus of these demonstration plants is accumulate | stored in track record data storage information.

  Hereinafter, specific examples of evaluation will be described by taking “burner”, “on / off valve”, and “furnace wall” as constituent devices.

(A) Burner The burner corresponds to the burner 5 of FIG. Although the burner itself was a mass-produced product, its cooling pipe structure was a new structure and there was no knowledge, so the installation procedure for the cooling pipe was inappropriate. As a result, the heat load on the burner main body increased and burnout occurred. After the occurrence of burnout, the installation procedure of the cooling pipe was reviewed, and no similar failure occurred.
In this example, since the cooling pipe structure is a new structure, “1”, which means that the new technology is included, is entered in the “new technology” item, which is caused by a potential failure such as inappropriate installation procedure. Therefore, “failure mode” is evaluated as “initial failure”. Since it is an “initial failure”, two items of “load magnitude X_1” and “novel X_2” of “initial failure evaluation” are evaluated. In this case, “3” is entered for “load size X_1” because the burner body has been burned out, and “3” is entered in “novel X_2”. The burner itself is a mass-produced product. Since there is a track record, it is evaluated that the novelty is low and “1” is input.

  In the item “Effects of countermeasures”, since a similar failure has not occurred after the review of the cooling pipe installation procedure, “A” is input because it is evaluated that there is no recurrence due to proficiency. Since similar measures are taken in the prediction target plant 1, the burner failure record data is not reflected in the prediction of the operation rate of the prediction target plant 1. In other words, it is not used for risk matrix fitting data described later.

(B) On / off valve The on / off valve corresponds to the on / off valve 12d of FIG. The on / off valve accidentally bites pulverized coal once a year (pulverized coal enters the valve moving part and sliding part), and the valve moving part sticks (sticks). is there. If a stick occurs, an excessive load is applied to the on / off valve, and if the stick is repeated several times, the on / off valve may be damaged and malfunction, resulting in a trip of the plant. At present, no effective measures are taken against the pulverized coal biting in the demonstration plant and the prediction target plant 1.

In this example, the failure of the on / off valve is due to accidental biting into the pulverized coal, so the “failure mode” is evaluated as “accidental failure” and is “incidental failure”. “Evaluation” “Evaluation frequency Y_1 of abnormal load” and “Magnitude of influence of abnormal load Y_2” are evaluated. Here, “abnormal load occurrence frequency Y_1” is evaluated as a medium frequency and “2” is input, and “abnormal load influence magnitude Y_2” is a plant failure due to multiple failures. “2”, which is evaluated as inviting a trip and meaning a middle level, is input.
In addition, since no effective measures are taken for biting at present, it is evaluated that the risk of failure remains in the item “effect of measures” and “C” is input. Since effective measures are not taken in the prediction target plant 1, the failure record of the on / off valve is reflected in the prediction of the operation rate of the prediction target plant 1. That is, it is used for risk matrix fitting data described later.

(C) Furnace wall A furnace wall is equivalent to the furnace wall 4 of FIG. In the inspection after 5 years of operation at the demonstration plant, many cracks were found on the furnace wall, which has a service life of 10 years.
In this example, since the crack is a secular change due to the operation of the demonstration plant 1, the “failure mode” is evaluated as “aging degradation”. Since it was evaluated as “aging deterioration”, “design uncertainty Z” in “aging deterioration evaluation” is evaluated. The cause of the secular change was faster than expected, but the cause was evaluated to be clear, and “2” representing a medium level was entered in the item of “design uncertainty Z”.
In addition, since no effective countermeasures have been taken against the cracks in the furnace wall at present, “C” is entered in the item “Effects of countermeasures” because the risk of failure remains. Is done. Since no effective measures are taken in the prediction target plant 1, the failure wall data of the furnace wall is reflected in the prediction of the operation rate of the prediction target plant 1. That is, it is used for risk matrix fitting data described later.

The risk matrix creating means 23 is a risk for initial failure described later so that the plant operation rate W (t) obtained from the device failure rate h (t) of each component device of the demonstration plant can reproduce the operation performance of the demonstration plant. A matrix 24a, a risk matrix 24b for accidental failure, and a risk matrix 24c for aging failure are created.
More specifically, the equipment failure rate h (t) is obtained as the total value of the initial failure rate h1 (t), the accidental failure rate h2 (t), and the aging failure rate h3 (t) [h (t) = h1 (t) + h2 (t) + h3 (t)]. Then, the operation rate W (t) of the entire demonstration plant is obtained based on the device failure rate h (t) of each component device, so that this plant operation rate W (t) approximates the operation results of the demonstration plant. , Risk matrix 24a for initial failure (specifically, values of initial failure coefficients k11 to k33 defined by the risk matrix 24a), risk matrix 24b for accidental failures (specifically, contingent failure coefficients defined by the risk matrix 24b) C11 to C33) and a risk matrix 24c for aging failure (specifically, values of logarithmic standard deviations σ1 to σ3 as aging failure coefficients defined by the risk matrix 24c) are adjusted.

  Hereinafter, the risk matrix creating means 23 will be further described while explaining the risk matrices 24a, 24b, and 24c.

  The risk matrix 24a for initial failure is a two-dimensional matrix as shown in FIG. 4A, and the rows and columns are respectively “load magnitude X_1” and FMEA sheet 22A shown in FIG. This corresponds to “Novelty X_2”. As described above, the “load magnitude X_1” and “novelty X_2” are evaluated for the components of the demonstration plant, and this evaluation is stored in the evaluation information storage means 22. The risk matrix creating means 23 obtains an evaluation regarding “load magnitude X_1” and “novelty X_2” of each component device from the evaluation information storage means 22, and selects the component device in the column in the risk matrix 24a according to this evaluation. Sort to the corresponding column in X11 to X33. For example, in the example of the initial failure of the burner in the demonstration plant described above, since the countermeasure has been taken in the prediction target plant 1, it is not reflected in the risk matrix 24a. However, if it is reflected temporarily, the “load magnitude X_1” is “3”, Since “Novelty X_2” is “1”, it is distributed to the column X13.

ここで、リスクマトリクス２４ａの欄X11〜X33内に記入されているk11〜k33は、上式（１）の形状母数kとして使用される具体的な数値であり、時間とともに故障率が低くなる初期故障率を算出する場合には１未満の正数で設定される。上記バーナの例では、バーナの形状母数kには、欄X13に割り当てられたk13が使用される。 Then, the initial failure rate h1 (t) of each component device is obtained by calculating the operation time t, the shape parameter (initial failure coefficient) k, and the scale parameter λ according to the failure rate calculation formula (1) based on the Weibull distribution shown below. Calculated using. It is known that the failure rate calculation formula (1) based on the Weibull distribution can reproduce and predict the failure rate with a certain accuracy.

Here, k11 to k33 entered in the columns X11 to X33 of the risk matrix 24a are specific numerical values used as the shape parameter k in the above equation (1), and the failure rate decreases with time. When calculating the initial failure rate, a positive number less than 1 is set. In the example of the burner, k13 assigned to the column X13 is used as the shape parameter k of the burner.

  FIG. 4B is an example of a failure rate curve obtained by using the shape parameter k having three different values according to the above equation (1). As is apparent from FIG. 4B, the positive failure rate h1 tends to be higher as the value of the shape parameter k is smaller when the shape parameter k is less than 1. In the risk matrix 24a, the shape parameters k11 to k33 are set to smaller values as the “load magnitude X_1” is higher (as it is higher in FIG. 4A) (k11> k12> k13, k21> k22>). k23, k31> k32> k33), and the higher the level of “novelty X_2” (the more right it becomes in FIG. 4A), the smaller the value is set (k11> k21> k31, k12> k22> k32, k13> k23> k33). That is, the shape parameters k11 to k33 are set so that the initial failure rate h1 (t) becomes higher as the level of the “load magnitude X_1” is higher and the level of the “newness X_2” is higher. Yes.

  The risk matrix creating means 23 uses the shape parameters k11 to k33 so that the plant operation rate based on the failure rate of each component device obtained using these coefficients k11 to k33 can accurately reproduce the results of the demonstration plant. A specific numerical value is determined and a risk matrix 24a is created. The numerical values in parentheses in FIG. 4A are examples of specific numerical values of the coefficients k11 to k33 determined by the risk matrix creating means 23, and are not limited to these.

The failure rate calculation formula (1) based on the upper Weibull distribution is expressed by the following formula (2) when the shape parameter k is 1. That is, it is possible to calculate an accidental failure rate at which the device failure rate h (t) is constant with respect to the operation time t with constant accuracy. Here, the failure rate at the time when the operation time of the demonstration plant has passed for several years and the operation rate has become substantially constant is substituted for the device failure rate h (t) in the following equation (2), and back calculated. The scale parameter λ is obtained.

  The risk matrix 24b for accidental failure is a two-dimensional matrix as shown in FIG. 5A, and the rows and the columns are “abnormal load occurrence frequency Y_1” in the FMEA sheet 22A shown in FIG. And “Magnitude of influence of abnormal load Y_2”. As described above, the evaluation equipment is evaluated for “abnormal load occurrence frequency Y — 1” and “abnormal load influence magnitude Y — 2”, and this evaluation is stored in the evaluation information storage means 22. The risk matrix creating means 23 obtains the evaluation regarding the “abnormal load occurrence frequency Y_1” and the “magnitude of the influence of the abnormal load Y_2” of each component device from the evaluation information storage unit 22, and selects the component device according to this evaluation. The risk matrix 24b is assigned to the corresponding column in the columns Y11 to Y33. For example, in the above-described example of the on-off valve accident, the “abnormal load occurrence frequency Y_1” is “2”, and the “abnormal load influence magnitude Y_2” is “2”.

ここで、リスクマトリクス２４ｂの欄Y11〜Y33内に記入されているC11〜C33は、上式（３）の故障係数Cとして使用される具体的な数値である。上記オンオフバルブの例では、オンオフバルブの故障係数Cには、欄Y22に割り当てられたC22が使用される。 Then, the random failure rate h2 (t) of each component device is calculated using the random failure coefficient C and the conventional random failure rate h2_0 of the conventional boiler plant for thermal power generation according to the following equation (3). The conventional boiler plant for thermal power generation (hereinafter referred to as a conventional plant) used for the conventional accidental failure rate h2_0 is preferably similar in configuration to the boiler plant 1 to be predicted.

Here, C11 to C33 entered in the columns Y11 to Y33 of the risk matrix 24b are specific numerical values used as the failure coefficient C in the above equation (3). In the example of the on / off valve, C22 assigned to the column Y22 is used as the failure coefficient C of the on / off valve.

FIG. 5B is an example of a failure rate line obtained by using the random failure coefficient C having three different values according to the above equation (3). As is clear from the above equation (3) and FIG. 5 (b), the larger the random failure coefficient C, the higher the random failure rate h2 (t). In the risk matrix 24b, the accidental failure coefficients C11 to C33 are set to have larger values as the “abnormal load occurrence frequency Y_1” becomes higher (as shown in FIG. 5A, the higher) (C11 <C12 <C13, C21 <C22). <C23, C31 <C32 <C33), a larger value is set as “the magnitude of the influence Y_2 of the abnormal load” becomes higher (as shown to the right in FIG. 5A) (C11 <C21 <C31, C12 < C22 <C32, C13 <C23 <C33). That is, the contingency failure coefficients C11 to C33 are set such that the contingency failure rate h2 (t) increases as the “abnormal load occurrence frequency Y_1” increases and the “abnormal load influence magnitude Y_2” increases. ing.
Here, C11 to C33 entered in each column Y11 to Y33 of the risk matrix 24b are specific numerical values used as the random failure coefficient C in the above equation (3). The risk matrix creating means 23 uses the failure coefficients C11 to C33 so that the plant operation rate based on the failure rate of each component device obtained using these coefficients C11 to C33 can accurately reproduce the performance of the demonstration plant. A specific numerical value is determined and the risk matrix 24b is created. The numerical values in parentheses in FIG. 5A are examples of specific numerical values of the coefficients C11 to C33 determined by the risk matrix creating means 23, and are not limited thereto.

  As shown in FIG. 6A, the risk matrix 24c for predicting the aging failure rate is a one-dimensional matrix, and the row thereof is the “design uncertainty Z” of the FMEA sheet 22A shown in FIG. Corresponding. As described above, the “design uncertainty Z” is evaluated for the components of the demonstration plant, and this evaluation is stored in the evaluation information storage means 22. The risk matrix creation means 23 obtains an evaluation regarding “design uncertainty Z” of each component device from the evaluation information storage means 22, and in accordance with this evaluation, the component device is selected from the columns Z1 to Z3 in the risk matrix 24c. Sort to the corresponding column. For example, in the example of the accidental failure of the furnace wall described above, since “design uncertainty Z” is “2”, it is assigned to the column Z2. As mentioned above, “Design Uncertainty Z” means the variation in the load and strength of the component equipment, and is used to predict the aging failure rate of the component equipment of the target plant entered in columns Z1 to Z3. [sigma] 1 to [sigma] 3 are logarithmic standard deviations indicating the magnitudes of variations in loads and strengths of component devices.

図６（ｂ）は、上式（４）により３つの異なる値の対数標準偏差σを使用して求めた故障率ラインの例であり、図６（ｂ）からも明らかなように対数標準偏差σが大きいほど経年劣化故障率h3(t)は高い傾向を示す。リスクマトリクス２４ｃでは、対数標準偏差σ1〜σ3は、「設計の不確かさZ」が高くなるほど（図６（ａ）で上になるほど）大きい値が設定されている（σ1<σ2<σ3）。すなわち、「設計の不確かさZ」が高いほど経年変化故障率h3(t)が高くなるように対数標準偏差σ1〜σ３が設定されている。
リスクマトリクス作成手段２３は、これらの対数標準偏差σ1〜σ33を用いて得られた各構成機器の故障率に基づくプラント稼動率が、実証プラントの実績を精度よく再現できるように対数標準偏差σ1〜σ3の具体的な数値を決定して、リスクマトリクス２４ｃを作成する。図６（ａ）における括弧内の数値は、リスクマトリクス作成手段２３により決定された対数標準偏差σ1〜σ3の具体的な数値の一例であり、これに限定されるものではない。 Here, the logarithmic standard deviations σ1 to σ3 entered in the columns Z1 to Z3 of the risk matrix 24c are specific numerical values used for the logarithmic standard deviation σ as a failure coefficient in the following equation (4). In the example of the furnace wall, σ2 assigned to the column Z2 is used as the logarithmic standard deviation σ of the furnace wall.
The aging failure rate h3 (t) is modeled by a lognormal distribution, and the logarithmic standard deviation σ, logarithmic mean μ, and error function erf are used as the aging failure factor by the following equation (4). One of the specific numerical values σ1 to σ3 is substituted for the logarithmic standard deviation σ. The logarithmic average μ is set so that the design life expectancy of the component equipment matches the 95% lower limit value of the lognormal distribution. That is, since two conditions of the logarithmic standard deviation σ and the 95% lower limit value are given, the logarithmic average μ can be obtained.

FIG. 6B is an example of a failure rate line obtained by using the logarithmic standard deviation σ of three different values according to the above equation (4), and as can be seen from FIG. 6B, the logarithmic standard deviation. The larger σ is, the higher the aging failure rate h3 (t) tends to be. In the risk matrix 24c, the logarithmic standard deviations σ1 to σ3 are set to have larger values (σ1 <σ2 <σ3) as the “design uncertainty Z” becomes higher (as shown in FIG. 6A). That is, the logarithmic standard deviations σ1 to σ3 are set so that the aging failure rate h3 (t) increases as the “design uncertainty Z” increases.
The risk matrix creating means 23 uses logarithmic standard deviations σ1 to σ1 so that the plant operation rate based on the failure rate of each component device obtained using these logarithmic standard deviations σ1 to σ33 can accurately reproduce the performance of the demonstration plant. A specific numerical value of σ3 is determined, and a risk matrix 24c is created. The numerical values in parentheses in FIG. 6A are examples of specific numerical values of the logarithmic standard deviations σ1 to σ3 determined by the risk matrix creating means 23, and are not limited thereto.

  The risk matrix storage unit 24 is a unit that acquires information from the risk matrix creation unit 23 and stores the risk matrices 24a, 24b, and 24c created by the risk matrix creation unit 23.

The failure coefficient setting unit 25 uses the risk matrices 24a, 24b, and 24c stored in the risk matrix storage unit 24 to calculate the device failure rate h (t) of each component device of the prediction target plant 1 shown in FIG. To do.
The failure coefficient setting means 25 receives from the input means 21 an initial failure evaluation (that is, “load magnitude X_1” and “newness X_2”) and an accidental failure evaluation (that is, “abnormal” for each component of the prediction target plant 1. Load generation frequency Y_1 "and" abnormal load influence magnitude Y_2 ") and aged deterioration evaluation (that is," design uncertainty Z ") are input.

For each of the evaluation items X_1, X_2, Y_1, Y_2, and Z, the same value as that of the corresponding verification plant is input for the same configuration device as that of the verification plant unless there are special circumstances. The special circumstances are, for example, situations where the constituent equipment of the target plant 1 is different from the constituent equipment of the demonstration plant, and it is natural that the knowledgeable person can predict that the level of the evaluation item will be different due to the difference. It is.
Furthermore, for the component equipment that does not have the same thing in the demonstration plant, the respective levels of the evaluation items X_1, X_2, Y_1, Y_2, and Z are set and inputted based on the opinions of the inspectors.
The failure coefficient setting means 25, for each component device, from each evaluation item X_1, X_2, Y_1 and Z, based on each risk matrix 24a, 24b, 24c stored in the risk matrix creation means 23, each failure coefficient k, Set C and σ.

  The failure rate calculation means 26 acquires each failure coefficient k, C, σ in each component device from the failure coefficient setting means 25, and calculates each failure in the measurement target plant 1 by the above equations (1), (3), (4). Based on the initial failure rate h1 (t), accidental failure rate h2 (t), and aging degradation of component equipment, the failure rate h3 (t) and the total device failure rate h (t) are calculated (h (t) = h1 (t) + h2 (t) + h3 (t)).

The plant operation rate calculation unit 27 acquires the device failure rate h (t) in each component device from the failure rate calculation unit 26, and predicts it using a method set in advance based on the connection of failures between component devices. The operation rate W (t) of the target plant 1 is predicted (calculated). Similar to the RAM analysis described in the background art, the failure connection between the component devices is obtained in advance by a reliability block diagram created based on the system diagram of the entire plant.
The external output means 28 is a monitor or printer connected to the computer, and displays the operating rate W (t) calculated by the plant operating rate calculating unit 27 as an operating rate curve with respect to the operating time t.

For example, in the case where there are a plurality of push-in ventilators 13a in parallel because the plant 1 to be predicted has a spare push-in ventilator 13a, even if one of the push-in ventilators 13a fails, the other push-in ventilator 13a Since combustion air can be supplied to the burner 5, the plant does not stop completely. Therefore, when calculating the plant operating rate W (t), the degree of influence of the equipment failure rate h (t) of the forced air blower 13a is set to be low.
On the other hand, when the predicted target plant 1 has only one forced draft fan 13a, if the forced draft fan 13a fails, the combustion target cannot be supplied to the burner 5 and the predicted target plant 1 trips (stops). Therefore, when calculating the plant operating rate W (t), the degree of influence of the equipment failure rate h (t) of the forced air blower 13a is set higher.

  As described above, the prediction of the operation rate W (t) of the prediction target plant in the actual operation rate predicting apparatus 20 is made to the device failure rate h (t) in each component device in consideration of the connection of failures between the component devices. This is common to the conventional RAM analysis in that the operation rate W (t) of the plant 1 to be predicted is predicted. However, the prediction in the actual operation rate predicting device 20 reflects the failure performance of the demonstration plant in the prediction of the operation rate W (t) of the prediction target plant. Which of the failure items is classified into any of the failure types of deterioration failure, and in which evaluation items X_1, X_2, Y_1, Y_2, and Z each component device affects the failure rate h1 (t), h2 (t), h3 (t) It differs greatly from the conventional RAM analysis in that it is analyzed whether it corresponds to a certain level (hereinafter, the prediction of the operation rate W (t) in the actual operation rate prediction device 20 is also referred to as knowledge-based RAM analysis).

[1-3. Plant utilization rate prediction method]
A plant operation rate prediction method by the plant operation rate prediction apparatus 20 of the present embodiment will be described with reference to FIGS.
The failure record in the component equipment of the demonstration plant is input from the input unit 21 and is accumulated in the evaluation information storage unit 22. The risk matrix creating means 23 refers to the failure record of the demonstration plant stored in the evaluation information storage means 22, and uses the risk matrix 24a for initial failure, the risk matrix 24b for accidental failure, and the risk matrix 24c for aging failure. It is created by fitting the actual operation rate of this demonstration plant.

That is, the risk matrix creating means 23 is for the shape parameters k11 to k33 defined by the risk matrix 24a for initial failures, the random failure coefficients C11 to C33 defined by the risk matrix 24b for random failures, and for aged deterioration failures. The logarithmic standard deviations σ1 to σ3 defined by the risk matrix 24c are set as specific numerical values so that the actual performance of the demonstration plant can be accurately reproduced.
The risk matrix storage unit 24 stores the risk matrices 24a, 24b, and 24c created by the risk matrix creation unit 23, and the failure coefficient setting unit 25 stores the risk matrices 24a, 24b, and 24c stored in the risk matrix storage unit 24. Referring to FIG. 4, the shape parameter k, the random failure coefficient C, and the logarithmic standard deviation σ are set for each component device of the prediction target plant 1.

Next, the failure rate calculation means 26 determines the initial failure rate for each component device of the prediction target plant 1 based on the shape parameter k, the random failure coefficient C, and the logarithmic standard deviation σ set by the failure coefficient setting means 25. h1 (t), accidental failure rate h2 (t), aged failure rate h3 (t), and hence device failure rate h (t) are calculated.
FIG. 7 shows the failure rate h (t) of the component device calculated by the failure rate calculation means 26 (estimated based on the knowledge-based RAM analysis), and the failure rate of the component device predicted based on the conventional RAM analysis. It is a figure shown with h_RAM (t). As is clear from FIG. 7, the failure rate h_RAM (t) of the component device based on the conventional RAM analysis is constant regardless of the operation time t because only the accidental failure is considered. In the equipment failure rate h (t) of the constituent equipment based on the knowledge-based RAM analysis by the device 20, the behavior of the failure rate can be accurately reproduced in the initial operation period, the intermediate operation period, and the late operation period (time near the life).

  In other words, in the device failure rate h (t) of the component device by the prediction device 20, in the initial operation (hereinafter also referred to as the initial failure period), there is an initial failure that frequently occurs at the beginning of operation and the number of occurrences rapidly decreases with time. In the middle period of operation (hereinafter also referred to as “accidental failure period”), an accidental failure with the failure rate having a substantially constant small value becomes the main cause of failure, and the latter period of operation (hereinafter also referred to as “aging degradation failure period”). ) Accurately reproduces the behavior of the failure rate h associated with the operation time t, such as aged deterioration failures with a sudden increase in the number of occurrences being the main cause of failure.

  Then, the plant operation rate calculating means 27 calculates the predicted target plant from the equipment failure rate h (t) of the constituent devices of the predicted target plant 1 (predicted based on the knowledge base RAM analysis) calculated by the failure rate calculating means 26. A plant operating rate W (t) of 1 is predicted (calculated), and is output to the external output means 28 as an operating rate curve for the operating time t as shown in FIG. In FIG. 8, for comparison, together with the plant operation rate W (t) calculated by the plant operation rate calculation means 27 (prediction device 20), the plant operation rate W_RAM (t) predicted based on the conventional RAM analysis. And the results W1, W2, W3, and W4 of the operation rate of the demonstration plant used for fitting the risk matrices 24A, 24b, and 24c are shown.

As is clear from FIG. 8, the plant operation rate W_RAM (t) based on the conventional RAM analysis has a large error with the actual operation rates W1, W2, W3, and W4 of the demonstration plant. Furthermore, the plant operation rate W_RAM (t) based on the conventional RAM analysis is constant regardless of the operation time t because only accidental failures are taken into account, and the operation rate decreases when the operation time is short. The behavior of W1, W2, W3, W4 has not been reproduced. On the other hand, in W (t) based on the knowledge-based RAM analysis by the prediction device 20, there is little error with the actual operation rate results W1, W2, W3, and W4 of the demonstration plant, and the actual operation rate results W1 with respect to the operation time t. Therefore, the behavior of W2, W3, and W4 can be reproduced accurately.
Note that FIG. 8 shows the prediction of the operation rate from the initial operation to the middle operation, and does not show the operation rate in the operation late stage in which the aging degradation failures rapidly increase.

[1-4. effect]
According to the plant operation rate prediction apparatus (plant operation rate prediction method) of the first embodiment, there is an advantage that the future operation rate W (t) can be accurately predicted based on a small amount of actual data.
That is, the levels of the evaluation items X_1, X_2, Y_1, Y_2, Z related to each failure rate h1 (t), h2 (t), h3 (t) of the component equipment (details are the magnitudes that affect the failure rate) ) And the failure coefficients k, C, and σ that correlate with the magnitudes of the failure rates h1 (t), h2 (t), and h3 (t) of the component devices are represented by risk matrices 24a, 24b, and 24c. It is prescribed. The evaluation items X_1, X_2, Y_1, Y_2, and Z related to failure are subdivided according to the level that affects the failure rate of the component equipment, and risk matrixes 24a, 24b, and 24c are configured to indicate the operation rate of the demonstration plant. By dropping the actual results W1, W2, W3, W4 (the failure coefficients k, C, σ are defined by the risk matrices 24a, 24b, 24c systematically subdivided according to the degree of influence on the failure rate of the component equipment. This failure coefficient k, C, σ is set so that the actual W1 performance W1, W2, W3, W4 of the demonstration plant can be accurately reproduced). A certain level of reliability can be ensured. Therefore, the future operation rate W (t) can be accurately predicted based on a small track record.

  In addition, the initial failure rate h1 (t), the accidental failure rate h2 (t), and the aging failure rate h3 (t) are calculated separately for different failures, and the initial failure rate h1 (t), the accidental failure rate h2 (t) and the aged failure rate h3 (t) are used as the equipment failure rate h (t), so the equipment failure rate h (t) whose trend changes according to the operating time t and the plant operation of the forecast target plant 1 There is an advantage that the rate W (t) can be predicted with high accuracy.

In recent years, a contract regarding a plant operating rate may be concluded between a plant maker and a plant delivery company. In such a contract, for example, if the actual operating rate of the plant falls below the guaranteed operating rate, the plant manufacturer pays a penalty to the delivery company, and conversely, the actual operating rate of the plant is the guaranteed operating rate or the guaranteed operating rate. When the operating rate specified separately from the rate is exceeded, the operating rate excess achievement bonus is paid from the delivery company to the plant maker.
The plant operation rate prediction device (plant operation rate prediction method) can predict the operation rate W (t) of the target plant 1 with higher accuracy than before, so the probability of achieving the guaranteed operation rate and the guaranteed operation rate There is an advantage that an unachieved risk that cannot be achieved can be accurately evaluated. Thereby, both the plant maker and the delivery destination company can quantitatively and accurately evaluate the business performance of the contract contents.
Further, economic indicators such as NPV (internal return) and IPR (net present value) relating to the business investment value for the customer (delivery destination company) can be calculated stochastically according to the operation rate.

[2. Second Embodiment]
The plant operation rate prediction apparatus and prediction method of the present invention will be described with reference to FIGS. In addition, the same code | symbol is attached | subjected about the same element as 1st Embodiment, and the description is abbreviate | omitted.
In addition, the prediction object plant 1 which concerns on this embodiment is comprised as shown in FIG. 1 similarly to 1st Embodiment, The description is abbreviate | omitted.

[2-1. Configuration of plant operating rate prediction device]
The plant operation rate prediction apparatus 20A of the present embodiment is configured as shown in FIG. 9, and an influence degree calculating means 29 is added to the operation rate prediction apparatus 20 of the first embodiment shown in FIG. It has a configuration.
The influence degree calculation means 29 acquires the equipment failure rate h (t) of each component device from the failure rate calculation means 26, and from the FMEA sheet 21A (see FIG. 3) stored in the evaluation information storage means 22, the demonstration plant The operationable output P_a and the recovery time t_r when the corresponding component device fails are acquired. Further, the influence degree calculating means 29 adds the recovery time t_r when the operable output P_a is 0 (zero), that is, the time when the demonstration plant is tripped, thereby obtaining the integration time t_trip of the demonstration plant stop time. calculate.
Then, the influence degree calculating means 29 calculates the influence degree E_trip on the plant stop of each component device from the recovery time t_r and the accumulated stop time t_trip by the following equation (5).
E_trip = h (t) × t_r / t_trip (5)

Note that the recovery time t_r of the component equipment is the verification plant if there are special circumstances such as the fact that the recovery time of the component equipment of the target plant 1 is clearly different from the recovery time of the corresponding component equipment of the verification plant. A time obtained by correcting the recovery time of the corresponding component device may be used.
The recovery time t_r is the time required for recovery when a failure actually occurs in the demonstration plant. When no failure occurs in the component equipment in the demonstration plant, the recovery time t_r and consequently the influence E_trip of the corresponding component equipment of the prediction target plant 1 is 0 (zero).

Then, the influence degree calculating means 29 outputs the influence degree E_trip of each constituent device to the external output means 29, and the external output means 29 converts the influence degree E_trip of each constituent equipment into the influence degree E_trip as shown in FIG. Display or print these items in order of size. In addition, in FIG. 10, description of a name is abbreviate | omitted about structural apparatuses other than a burner, denitration equipment, and a forced air fan.
As a result, high-risk equipment with a high impact E_trip and a high risk of lowering the plant operating rate W (t) is divided into a low-risk device with a low impact E_trip and a low risk of reducing the plant operating rate W (t). be able to.
Since other configurations are the same as those of the first embodiment, description thereof is omitted.

[2-2. effect]
According to the plant operation rate prediction apparatus and prediction method of the second embodiment, in addition to the same effects as the plant operation rate prediction apparatus and prediction method of the first embodiment, the following effects are obtained. It is done.
That is, by calculating the influence level E_trip of each component device, it is possible to divide the component device into a component device with a large impact level E_trip and a component device with a small impact level E_trip. In other words, there is a high risk of lowering the plant operation rate W (t), a high-risk device that requires countermeasures, and a low risk of reducing the plant operation rate W (t), and a low-risk device that does not require countermeasures. Can be divided. Then, by taking necessary measures for high-risk equipment, the plant operation rate W (t) can be improved efficiently.

  In the example shown in FIG. 10, the burner 5 is the component device having the largest influence degree E_trip, and the influence degree E_trip is the largest in the order of the denitration equipment 14b and the forced air blower 13a. Here, for the burner, measures are taken such as shortening the inspection interval and the regular replacement interval of the consumables, and for the denitration device 14b, a measure is taken to ensure spare parts that take time to deliver. For the push-in ventilator 13a, a countermeasure is taken such that a spare machine is provided to make two lines.

  For the components that have taken such countermeasures, the levels of the evaluation items X_1, X_2, Y_1, Y_2, and Z in the risk matrices 24a, 24b, and 24c, the reliability block diagram, and the device failure rate h (t) are used. The method for calculating the plant operation rate W (t) is appropriately corrected. Reflecting this correction, the plant operation rate W (t) is calculated again, and if the plant operation rate W (t) still fails to achieve the target operation rate, the range of component devices to which countermeasures are taken can be expanded. Increase the level of countermeasures (for example, shorten the interval between regular replacements of consumables). On the other hand, when the target operating rate is excessively exceeded, the range of component devices for which countermeasures are taken is reduced and the level of countermeasures is reduced (for example, the interval for periodic replacement of consumables is extended). That is, the measures taken for the component devices can be quantitatively evaluated, and the range of component devices for which measures are taken and the level of measures can be rationally selected.

[3. Third Embodiment]
The plant operation rate prediction apparatus and prediction method of the present invention will be described with reference to FIGS. In addition, the same code | symbol is attached | subjected about the same element as 1st Embodiment, and the description is abbreviate | omitted.
In addition, the prediction object plant 1 which concerns on this embodiment is comprised as shown in FIG. 1 similarly to 1st Embodiment, The description is abbreviate | omitted.

[3-1. Configuration of plant operating rate prediction device]
The plant operation rate prediction device 20B of the present embodiment is configured as shown in FIG. 11, and a failure coefficient correction means 25a is added to the operation rate prediction device 20 of the first embodiment shown in FIG. It has a configuration. The failure coefficient correction unit 25a acquires the operation information input from the operation diary system 30 via the operation information storage unit 31, and corrects the failure coefficients k, c, and σ based on the operation information.
Here, the operation diary system 30 is used for sequentially inputting operation information such as events of the day, and is disposed, for example, in the central control room of the prediction target plant 1. Here, as shown in FIG. 12, the operation diary system 30 has input items for event start time, event end time, event target device, event category, event content, plant load, and remarks. In the example shown in FIG. 12, it is input that the pulverized coal machine stopped at 18:02 due to a failure due to a defective bearing bearing of the coal feeder and was restored at 20:02.
The driving information input by the driving diary system 30 is sequentially stored in the driving information storage means 31.

The failure coefficient correction unit 25a acquires the operation information from the operation information storage unit 31 sequentially or every few days, and if the acquired operation information includes information on the failure, reflects this in the failure coefficient. .
In the example shown in FIG. 12, the pulverized coal machine 12 fails due to a defective coal bearing bearing, and this failure belongs to an accidental failure. Therefore, the failure coefficient correction means 25 a is the pulverized coal machine set by the correction coefficient setting means. The random failure coefficient C is corrected by multiplying or adding the 12 random failure coefficient C by a correction coefficient. This correction coefficient can be determined based on, for example, a ratio between a time during which normal operation is performed in the pulverized coal machine 12 and a recovery time.

  The failure coefficient correction unit 25a changes k11 to k13, k21 to k23, and k31 to k33 defined based on each evaluation item in each risk matrix 24a, 24b, and 24c stored in the risk matrix storage unit 24. You may make it do. That is, the risk matrices 24a, 24b, and 24c themselves may be corrected.

[3-2. effect]
According to the plant operation rate prediction apparatus and prediction method of the third embodiment, in addition to the same effects as the plant operation rate prediction apparatus and prediction method of the first embodiment, the following effects are obtained. It is done.
That is, since the failure coefficient guarantee unit 25a corrects (updates) the failure coefficient based on the information about the failure acquired from the operation information storage unit 31, the failure coefficient is sequentially updated in accordance with the operation performance of the prediction target plant 1 itself. As a result, it is possible to predict the equipment failure rate h (t) and consequently the plant operation rate W (t) with higher accuracy.

[4. Others]
The plant operation rate prediction apparatus and the prediction method of the present invention are not limited to the above-described embodiments, and may take the following aspects, for example.

  (1) In each of the above embodiments, one demonstration plant is used as a similar plant used for fitting the risk matrices 24a, 24b, and 24c. However, a plurality of similar plants may be used. For example, two plants, a demonstration plant and a test plant, may be used as similar plants, and both results may be used for fitting the risk matrices 24a, 24b, and 24c. In addition, when the target plant has multiple new technologies, the results related to the new technology of similar plants including a part of the new technology and the results related to the new technology of another similar plant including other parts of the new technology You may use for fitting of risk matrix 24a, 24b, 24c.

  (2) In each of the above embodiments, a burner, an on / off valve, and the like are exemplified as the constituent elements of the present invention. However, the constituent elements do not have to be in units of such devices, and the constituent elements are further subdivided into devices. It is good also as a part unit. For example, the on / off valve may be further subdivided, and the valve body, the drive portion of the valve body, the valve seat, and the like may be used as components.

(3) The types and number of evaluation items (in other words, evaluation axes of the risk matrices 24a, 24b, and 24c) for setting the respective failure coefficients k, C, and σ are not limited to the above embodiment. For example, the evaluation items for setting the random failure coefficient C include "whether the design standards and maintenance standards for component equipment are in place" and "completion degree of component equipment (new product or technically mature product). ”) May be used instead of or in addition to“ the abnormal load occurrence frequency Y_1 ”and“ the abnormal load magnitude Y_2 ”which are the evaluation items in the embodiment.
Further, the number of rows and columns of each risk matrix 24a, 24b, 24c (that is, the degree of subdivision of the risk matrices 24a, 24b, 24c) is not limited to the above embodiment. For example, the risk matrix 24a for initial failure may be a 4 × 4 matrix.

  (4) The failure rate h (t) may not be calculated using the risk matrices 24a, 24b, and 24c for all the constituent devices of the prediction target plant 1. For example, a known failure rate curve may be used for a component device that has already been proven and has a known failure rate curve with high reliability.

  (5) In each of the above embodiments, the operation rate prediction device and the operation rate prediction method of the present invention are applied to the prediction of the operation rate of the pulverized coal fired boiler plant. It can also be applied to predicting the operating rates of various plants such as power plants, nuclear power plants and chemical plants, and various mechanical systems such as aircraft, rockets and ships. In short, what is to be predicted for the operation rate is not limited at all.

1 Boiler plant (forecast target machine system, forecast target plant)
20, 20A, 20B Occupancy rate prediction device 24a Risk matrix for initial failure (map for initial failure)
24b Risk matrix for accidental failure (Map for accidental failure)
24c Risk matrix for aging failure (Map for aging failure)
24 Risk matrix storage means (map storage means)
25 Failure coefficient setting means 25a Failure coefficient correction means 26 Failure rate calculation means (failure rate estimation means)
27 Plant operation rate calculation means (plant operation rate estimation means)
31 Operation information storage means 32 Failure coefficient correction means
C, C11 to C33 Random failure coefficient
E_trip Influence of component machine system on plant shutdown
h (t) Estimated equipment failure rate based on knowledge-based RAM analysis
h_RAM (t) Failure rate predicted based on conventional RAM analysis
h1 (t) Initial failure rate
h2 (t) Random failure rate
h3 (t) Aging failure rate
k, k11 to k33 Shape parameter (initial failure coefficient)
W (t) Predicted plant utilization based on knowledge-based RAM analysis
X_1 Load size
X_2 Novelty
Y_1 Frequency of abnormal load
Y_2 Influence of abnormal load
Z Design uncertainty σ, σ1 to σ3 Logarithmic standard deviation (aging deterioration factor)

Claims (9)

An operation rate prediction device that predicts the operation rate of a prediction target machine system including a plurality of components,
Each of the plurality of components is evaluated for the evaluation item related to the failure rate,
Map storage means storing a map defining a correspondence relationship between the level of the evaluation item and the failure coefficient;
From the level related to the evaluation item of the component, failure coefficient setting means for setting the failure coefficient of the component based on the map;
A failure rate estimation means for estimating a failure rate of the component using the failure coefficient;
An operation rate estimating means for estimating the operation rate of the prediction target machine system based on the failure rate of each of the plurality of components;
The map is fitted so as to match the actual operation rate of the similar mechanical system estimated by the operation rate estimation unit.
An operation rate predicting device characterized by that. It said map storage means, the initial failure map for calculating the initial failure factor, and random failures map for calculating the random failure coefficients and aging failure map for calculating the aging failure factor Remember,
The failure coefficient setting means sets the initial failure coefficient using the initial failure map for the component, sets the random failure coefficient using the accidental failure map, and the aging failure Set the aging failure coefficient using the map for
The failure rate estimating means estimates the initial failure rate by using the initial failure factor, using the random failure coefficient estimates the contingent failure rate, after a year using the aging failure factor degradation failure rate is calculated, the failure rate, the initial failure rate, and calculating a sum of the random failure rate and the aging failure rate, operation rate prediction apparatus according to claim 1. The map storage means stores, as the initial failure map, a map that defines a correspondence relationship between the initial failure coefficient, the magnitude of the load of the component, and the novelty of the component,
The failure coefficient setting means sets the initial failure coefficient of the component based on the initial failure map, based on the load of the component and the novelty of the component, The operation rate prediction apparatus according to claim 2. The map storage means stores, as the random failure map, a map that defines a correspondence relationship between the random failure coefficient, the occurrence frequency of the abnormal load of the component, and the magnitude of the influence of the abnormal load of the component. ,
The failure coefficient setting means sets the random failure coefficient of the component based on the occurrence frequency of the abnormal load of the component and the magnitude of the influence of the abnormal load of the component based on the random failure map. The operating rate predicting apparatus according to claim 2 or 3, wherein The map storage means stores, as the aging failure map, a map that defines a correspondence relationship between the aging failure coefficient and the design uncertainty of the component,
5. The failure coefficient setting means sets the aged deterioration failure coefficient of the component based on the design uncertainty of the component based on the aged deterioration failure map. The operation rate prediction apparatus according to any one of the above. And the failure rate of the component, the similar mechanical system the components in the based on the failure to the recovery time of the actual and cumulative stopping time record it took to recover from, for each pre Symbol of the component The operation rate prediction apparatus according to claim 2, further comprising an influence degree estimation unit configured to estimate an influence degree to stop of the prediction target machine system. The operation information before Ki予 measuring target machine system acquires the failure information from the operation information storage means for periodically storing, further comprising a failure factor correcting means for updating the fault factor based on said failure information The operating rate prediction apparatus according to any one of claims 1 to 6. Before Ki予 a measurement target machine system plant, wherein the similar mechanical system is similar plant operating rate prediction apparatus according to any one of claims 1-7. An operation method of an operation rate prediction apparatus for predicting an operation rate of a mechanical system composed of a plurality of components,
The failure factor and the level of the evaluation item related to the failure rate so that the performance rate of similar mechanical systems can be reproduced based on the failure rate of each of the plurality of components estimated using the failure factor. A first step of creating a map that defines a correspondence relationship with
A second step of setting the failure coefficient of each of the plurality of components based on the map from the level of the evaluation item of the component, and estimating a failure rate of the plurality of components;
Calculates the rate of operation of the prior Symbol machinery system based on each of the failure rate of the plurality of components, and a third step, the operation method of the operation rate predictor.

Images