JPH0289101A - Support system for process plant in case of emergency - Google Patents

Abstract

PURPOSE:To attain emergency support activity by means of appropriate timing by using an accident prediction/analysis code operation support device which sets the analysis conditions of an accident prediction/analysis code, identifies an accident event and derives the corresponding operation. CONSTITUTION:A derivation function for accident prediction analysis code analysis condition 10 in the accident prediction/analysis code operation support device 4 reads process data from a data collection device 3, sets the analysis condition of the accident prediction analysis code and outputs it as the input signal of the accident prediction/analysis code 5 provided in a prediction calculation device 6. Then, an accident event identification function 12 compares the analyzed result with past accident data stored in the data collection device 3, evaluates the identified result of the accident event and identifies the accident event. A corresponding operation derivation function 13 supports the derivation of the corresponding operation, and the prediction calculation device 6 performs the prediction calculation of plant behavior including the corresponding operation. Thus, the support activity with satisfactory timing can be performed.

Description

Detailed Description of the Invention [Objective of the Invention] (Industrial Application Field) The present invention is a process plant such as a nuclear power generation plan I, in which an emergency situation that affects outside the plant occurs. This is related to a system that supports staff who provide emergency support in evaluating long-term countermeasures and investigating causes, especially when it comes to
In addition to converting 7-speed information into A-line, the accident prediction analysis code is efficiently operated by the IT failure prediction analysis code operation support device, which enables quick response in emergencies. Regarding support systems.

BACKGROUND OF THE INVENTION The potential dangers of nuclear power plants, especially nuclear power plants containing large amounts of tIl radiation, are great. For this reason, we are implementing measures such as redundant functions to contain radioactivity through various safety facilities, periodic inspections, and inspections of operators' skills [
Countermeasures such as maintenance and improvement have been taken, and as a result of these countermeasures, the probability of an accident leading to the release of radioactivity into the surrounding environment is extremely low.

In the unlikely event that an accident occurs, an emergency technical assistance team will be organized to carry out long-term response operations at the plant based on various information. In addition to investigating the cause, it is conceivable to provide advice to residents in the area, including the creation of an evacuation plan.

In order to carry out such activities, it would be more effective if not only the current state of the plant but also the future behavior of the plant could be predicted, such as an evaluation of the effectiveness of possible countermeasures. Therefore, we have set 81 calculation conditions for the accident prediction analysis code using conventional knowledge and light experience to predict the behavior of the plant in an emergency. I am planning to do l\i.

(Problem to be Solved by the Invention) Conventionally, when executing an accident prediction analysis code, a high degree of specialized knowledge and knowledge are required, including setting up a huge amount of data and interpreting calculation results. experiments have been required. For this reason, even if the effectiveness of using Accident Prediction Analysis 21-D in an emergency situation is understood, accident prediction altIw? - It is difficult to quickly operate the analysis code, and it can be supported by tying it together at the appropriate timing. There was an operational soil issue regarding whether or not the fish could carry out activities.

The present invention has been made in consideration of the above circumstances, and is a professional system that allows emergency support staff to quickly and easily use accident prediction analysis codes and perform emergency support activities at appropriate times. The purpose is to provide a plant emergency support system.

[Structure of the invention]

(Measures to solve the problem) The prohysplant emergency support system according to the present invention includes:
A data collection device that captures and saves the 70 response data from Plan 1 to Supports the operation of the accident prediction analysis code that derives response operations! Plan 1-
a prediction correction system d that performs prediction 81n regarding the behavior of
A man-machine interface device for displaying data from the above-mentioned accident prediction analysis code operation support device and inputting requests; In addition to being equipped with a database that stores the knowledge and data necessary for setting, deriving the specifications of accident events and corresponding operations, the above-mentioned accident prediction analysis code operation support 1! ! @ place,
Data collection! Accident prediction, which sets the analysis conditions for the accident prediction analysis code based on the process data read from the i-system.
A 1m analysis code analysis condition derivation function, an accident event identification function that evaluates the analytical conclusion of the accident prediction analysis code by comparing it with the general process data stored in the data collection 'J equipment listed below, and the accident event identification function. This system is equipped with a corresponding operation derivation support function that supports the derivation of a corresponding operation to end the problem.

(Operation) First, the data convergence device takes in 71'' process data from the plant and stores it at /v. Accident prediction analysis code operation support The accident prediction analysis code analysis condition derivation function of the TL device reads plant process data from the data collection device.
Then, set the analysis conditions for the plant accident prediction analysis code, and set the analysis conditions in advance! It is output as an input signal for the skewer failure prediction analysis code provided in the control device.

Next, the accident event identification function compares the analysis result using the accident prediction analysis code with past plant accident data stored in the data collection device, evaluates the accident event identification results, and performs convergence calculations. Identify the event.

Then, the corresponding operation derivation function supports derivation of a corresponding operation for terminating the accident vehicle 9, and the prediction calculation device performs predictive calculation of the plant behavior including the corresponding operation.

The man-machine interface device stores data such as the setting of analysis conditions for the accident prediction analysis code, identification of accident events, derivation of response operations, analysis conditions for prediction ill iit tl, identification results of accident events, response operations, prediction 81 calculation results, etc. In addition to outputting data, requests for modification of these data can also be input.

In this way, process data from the plant is input online and the accident "Fall analysis code" is automatically operated, making it possible to efficiently operate the accident prediction analysis code and improve timing. It becomes possible to carry out good support activities.

(Example) An example of the present invention will be described with reference to the attached circular.

FIG. 1 is a block diagram showing an embodiment of the process plant emergency +15 support system according to the present invention.

The present invention is suitably used as a nuclear power plant emergency support system,
The data collection bag M3 that takes in the stator 2 and holds it at It is equipped with an accident prediction analysis code operation support device 4 that derives response operations, and an accident prediction analysis code 5 whose analysis conditions are set by the accident prediction analysis]-de operation support 4M device 4. a prediction calculation device 6 for performing predictive calculations, a man-machine interface device 7 for displaying data from the accident prediction analysis code operation support device 4 and for manually issuing requests, and an accident prediction analysis code operation support device 1 for displaying data from the accident prediction analysis code operation support device 4; ! Nuclear photoelectric plant 1 accident F a! Setting the analysis conditions for II analysis code,
A database 8 storing knowledge a3 and data necessary for identifying π events and deriving corresponding operations is provided.

Data collection bag δ3 is 1 Kyoko 517 tf Blanc 1-1
From mt secondary reactor water (H pressure, reactor containment m (PCV)
) Process data 2 such as pressure is saved manually.

The accident prediction analysis code operation support device 4 includes an accident prediction analysis code analysis condition derivation function 10 that sets analysis conditions for the accident prediction analysis code 5 based on the plan 1-process data 9 read from the data collection device 3; An accident event identification function 12 that compares and evaluates the analysis results by the accident prediction analysis code 5 with the actual machine process data 11 read from the data collection device 3, and derives a response operation to eliminate the accident event. A supporting corresponding operation derivation function 13 is provided.

The accident prediction analysis code analysis condition fl derivation function 10 sets analysis conditions as follows. First, the accident prediction analysis code analysis condition derivation function 10 detects loss of water supply, main steam isolation valve (MSIV) closure, safety relief valve (S/
RV) Stuck open, loss of power, loss of coolant accident (LOC8)
etc., estimate the plant status that will be the analysis condition. In particular, regarding the estimation of analysis conditions for cold 7Jl material loss accident,
Possible methods include comparing changes in major process values such as reactor containment power, which have been analyzed in advance, with the past history of the accident plant, and distinguishing between fracture area and gas phase/liquid phase. It will be done.

Next, the accident prediction analysis code analysis condition derivation function 10 performs an abnormality diagnosis of the year-start state of the main plant system for the main TL devices whose states can be directly input from the plant process data 9. Regarding other devices, the state that they should be in based on their design is provided to the support staff via the man-machine interface device 7 to assist in diagnosing the state.

Then, the accident prediction analysis code analysis condition derivation function 10
Consider the totality of input points for process values for which measured values are multiplexed, such as reactor water level, pressure, and reactor containment vessel temperature (Signal Validation)
) L/ and then aggregate it into one value that is suitable as an input for the accident prediction analysis code 5.

The accident prediction analysis code analysis condition 4 output function 10 outputs the analysis conditions 1q thus obtained to the prediction output 10 device 6 as the input signal 14 of the accident prediction analysis code 5.

The accident prediction analysis code 5 performs calculations for identifying accident events from the occurrence of the accident to the present time, and is designed to identify accident events. Therefore, the calculation is completed in a relatively short time.

The accident event identification function 12 first inputs the analysis result 15 based on the accident prediction analysis code 5, compares this analysis result 15 with the actual machine process data 11 read from the data collection device 3, and evaluates the accident event identification result. If the error between the analysis result 15 and the actual process data 11 is greater than or equal to the allowable iia, the analysis conditions are corrected to reduce the error. For example, for a coolant loss accident statuette, the initially estimated fracture location and fracture area are corrected to more accurate fracture locations and fracture areas. In this way, the accident event identification function 12
The calculated correction data is output as a correction signal 16 to the accident prediction analysis code analysis condition derivation function 10, and re-correction is performed until the error converges and the event is identified. The accident event identification function 12 is connected to the man-machine interface device 7 via an input/output signal 17. While providing support staff with convergence binding, correction binding, etc. exclusively for identification calculations, they also receive input on necessary corrections, termination of convergence calculations, etc.

Corresponding operation derivation expense l! Function 13 is equipped with a database of corresponding operations (1?u 1e-based) based on the operating procedure manual as rules, while the support staff considers via the man-machine interface device 7 based on Plan 1 to understanding the status. corresponding operation (nowledgc
-based) to consider accident termination procedures? It is becoming like this. The corresponding operation derivation support function 13 first manually inputs information 18 regarding the plant state from the accident prediction analysis code analysis condition derivation function 10 and provides it to the support staff via the man-machine interface device 17. Furthermore, corresponding operation derivation expenses l! Function 131, 1 run at the request of L support staff!・Revision data, such as plan I-4i4 construction of piping, valves, pumps, etc., or what kind of interlocks these devices are equipped with, etc.
Provide information. Corresponding operation derivation expense l! The function 13 inputs, via the man-machine interface device 7, changes to pre-ruled procedures or addition of new procedures performed by the support staff based on information provided via the man-machine interface device 7. Then, according to the request, the rule base is changed or added.

Predictor Sieve@Recommendation 6 is designed to predict the behavior of nuclear power generation plan 1-1 after the time when identification of accident events is completed. The prediction h10 device 6 outputs the preml calculation result 19 to the corresponding operation derivation support function 13,
This corresponding operation derivation support function 13 sequentially derives corresponding operations corresponding to the plant state based on predictive analysis. The prediction calculation device 6 inputs the corresponding operation as an operation request signal 20 from the corresponding operation derivation support function 13, and predicts the state of the nuclear power plant 1 including the corresponding operation. If the result of the predictor is not satisfactory, the corresponding operation derivation support function 13 re-evaluates the corresponding operation. In this case, if the Lister 1 function of the accident prediction analysis code 5 is used, the calculation can be performed efficiently.

The man-machine interface device 7 is connected to an accident event identification function 12, an accident prediction analysis code analysis condition derivation function 10, and a corresponding operation derivation support function 13 through a system not shown, and receives data and data via an input/output signal 17. It is now possible to input and output requests. In other words, the man-machine interface device 7
9, the 1 run 1-state from the accident prediction analysis code analysis condition derivation function 10 is provided to support staff via a display device, etc., and in the identification of accident events, the 1-run 1-state from the accident prediction analysis code analysis condition derivation function 10 is provided to the support staff. While providing support staff with the final I9 results and correction methods for the accident event identification function, necessary corrections, etc.
Input the termination of the convergence calculation etc. using a keyboard or mouse, etc., and use the corresponding operation suspension support I! to derive the corresponding operation. 8m
From Function 13 to Plan 1, we provide support staff with information on the status, and provide plant design data, etc. upon request from support staff, while also accepting changes and additions to the rule base from support staff, and making predictions. In calculations, predicted calculation results can be provided to support staff.

The database 8 is connected to an accident event identification function 12, an accident prediction analysis code analysis condition derivation device 110, and a corresponding operation derivation support function 13 through a route not shown. The database 8 includes knowledge for deriving analysis conditions for the accident prediction analysis code 5 from the plant process data 9, knowledge for identifying accident events, knowledge for deriving response operations, knowledge and plans regarding dialogue. Data related to design, etc. are stored. These knowledge and data are read as knowledge/data signals 21 in response to requests from the accident event identification function 12, the accident prediction analysis code analysis condition derivation function 10, and the corresponding operation derivation support function 13, and are read in as knowledge/data signals 21 by these functions 10.12. 13 is used for efficient operation.

Next, the operation of the above embodiment will be explained according to the frame 11 shown in FIG. The dashed line behind the processing shown in FIG. 2 indicates the processing by the accident prediction analysis code operation support device 4, and the solid line indicates the processing by the prediction calculation device 6. This indicates processing in which both devices 4 and 6 are integrated.

The process data 2 from the nuclear power plant 1 is collected via the data collection 1ff3, and the process data 9 is
The accident prediction analysis code analysis condition derivation function 10 first sets analysis conditions.

In setting the analysis conditions, estimate the plant state that will be the analysis condition (Step ■), perform an abnormality check on the operational status of the plant usage system (Step ■), and consolidate it into one value suitable as input for accident prediction analysis code 5. (Step ■). Next, the accident prediction analysis code 5 is used to identify the four accidents (step ■). Therefore, the accident prediction analysis code 5 is used to perform the analysis 15 using the accident 4-ba phenomenon identification function 12.
is compared with actual machine process data 11 to evaluate the identification consistency of accident events (step ψ)). Analysis conclusion! If the error between 1!15 and the solid line process data 11 is greater than the allowable value, the analysis conditions are corrected using the accident event identification function 12 so as to reduce the error (step 2).

1. When an accident event is identified, the response operation derivation function 11 performs a parsimonious response operation for mitigating Komaki and ending the accident (step ■). Then, the prediction calculation device 6 performs a prediction calculation on the behavior of the nuclear power plant 1, including the corresponding operations (step ①). If the result of the predictive calculation is not as desired, return to step (3) again and re-evaluate the corresponding operation. If the result of the predictive calculation is satisfactory, the flow ends there and the predictive calculation result is provided to the support staff via the man-machine interface device 7.

According to the above embodiment, by facilitating and speeding up the use of accident prediction analysis codes, emergency support activities can be carried out in a timely manner, and the safety and reliability of nuclear power plants can be further improved. can be achieved.

In addition, in the above embodiment, in a normal state where no emergency situation has occurred, the accident prediction analysis and code operation support device 4 can be used to easily set the accident state of the plant, such as pipe breakage, equipment failure, etc. It can also be used for learning plant behavior.

(Effect of the invention) The process plant W4 emergency support system according to the present invention includes a data collection device d that captures and stores process data from Plan I, and a data collection device d that reads process data from this data collection device. Accident prediction analysis code operation support lAt1l, which sets the analysis conditions of the accident prediction analysis code, identifies the first accident image, and derives the corresponding operation, and this accident prediction analysis code operation support! ! In addition to being equipped with an accident prediction analysis code with analysis conditions set by the device, a prediction calculation device that performs predictive measures regarding plant behavior, and the above-mentioned accident prediction analysis code operation support 1! ! A man-machine interface device for displaying data from the device and inputting requests, and setting analysis conditions for the accident prediction analysis code in the accident prediction analysis code operation support device d;
It is equipped with a database that stores the knowledge and data necessary for identifying accident events and deriving response operations, and the above-mentioned accident prediction analysis code operation support device is used to store process data read from the data collection device and perform accident prediction analysis. The above data includes the accident prediction analysis code analysis condition derivation function that sets the code analysis conditions, and the analysis results using the accident prediction analysis code! The actual machine block 0 [
1 (failure event identification function and response operation derivation support function that supports the derivation of response operations to quickly reduce the occurrence of the accident), generally require advanced knowledge and experience. It is possible to facilitate and speed up the continuous use of accident prediction and analysis codes that are required, and it is possible to carry out emergency support a! activities in a timely manner, and further improve the safety and reliability of process plants. can be achieved.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing an embodiment of the process plant emergency support system according to the present invention, and FIG. 2 shows the processing contents of the above embodiment. 1... Nuclear power, 51 electric plant, 3... Data collection equipment, 4... Accident prediction analysis code operation support i! device, 5
...Accident prediction analysis code, 6...Predictor sieve vl placement,
7...Man-machine interface device, 8...Database, 10...Accident prediction analysis code analysis condition derivation function, 12...Accident event identification function, 13...Response operation derivation support tlri.

Claims (1)

[Claims] A data collection device that captures and stores process data from the plant, and an accident prediction system that reads process data from this data collection device to set analysis conditions for accident prediction analysis codes, identify accident events, and derive countermeasures. An analysis code operation support device, an accident prediction analysis code whose analysis conditions are set by the accident prediction analysis code operation support device, and a prediction calculation device that performs predictive calculations regarding the behavior of the plant, and the above-mentioned accident prediction analysis A man-machine interface device for displaying data from the code operation support device and inputting requests, setting analysis conditions for the accident prediction analysis code in the accident prediction analysis code operation support device, identifying accident events, and responding to them. The accident prediction analysis code operation support device is equipped with a database that stores knowledge and data necessary for deriving operations, and sets analysis conditions for the accident prediction analysis code based on process data read from the data collection device in the accident prediction analysis code operation support device. A predictive analysis code analysis condition derivation function, an accident event identification function that compares and evaluates the analysis results from the accident prediction analysis code with the actual process data stored in the data collection device, and measures to end the accident event. A process plant emergency support system characterized by having a corresponding operation derivation support function that supports derivation of operations.