JPS6057522B2 - How to detect abnormality in gas supply source - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an abnormality detection method for early detection of a failure in a gas supply source by monitoring the flow rate of gas flowing through a gas pipeline.

In recent years, liquefied natural gas (LNG) has come to be widely used as a fuel for power generation due to emphasis on the living environment and resource conservation.

Moreover, this proportion is likely to increase in the future. Therefore, the proportion of the power generated by these power plants in the entire power generation system is increasing, and once an accident occurs in the LNG gas supply source, the impact will be large. The present invention detects a failure of the vaporizer at an early stage by constantly monitoring the rate of change in gas flow rate and the duration of the abnormal change rate in the gas pipeline that connects the vaporizer that vaporizes LNG and the power plant. The purpose is to communicate this information and encourage appropriate response measures to minimize the impact of the accident.

Such a device is not found in the prior art and is completely new. The amount of gas flowing through the gas vibe line is
It constantly fluctuates due to the properties of the conduit itself, noise in the detection system, and changes in the operating status of the power plant.

In the present invention, in order to prevent false alarms caused by fluctuations in the gas flow rate, a method is employed to smooth the measured flow rate and monitor how long a state of change exceeding a certain rate continues based on this smoothing. There is. The main points of the present invention are shown below. The present invention analyzes the behavior of the gas vibrine due to an accident in the gas supply source (gas vaporizer) through simulation, confirms that the method of monitoring the rate of change is the earliest method to detect an accident, and furthermore, This invention was created after investigating the gas flow rate of the vibrine and thoroughly analyzing the fluctuations in the flow rate. First, the detection device of the present invention takes in the gas flow rate and performs a moving average to prevent false alarms due to periodic fluctuations. This moving averaged flow rate value is used to calculate the rate of change in flow rate. However, if an accident is determined only based on the magnitude of the instantaneous value of the flow rate change rate, a false alarm may be issued due to irregular or sudden fluctuations in the flow rate. Therefore, when determining an accident, a certain allowable rate of change (negative value) is required.
The feature is that an accident is determined when a value greater than the absolute value continues for a predetermined period of time.
Figure 1 shows the carburetor, power plant, and accident detection equipment. vinegar.

Here, the LNG stored in the LNG tank 2 is gasified by the LNG vaporizer 3 and sent to each thermal power plant 91 to 9n via a gas vibe line 10. Accident detection device 1
takes in the flow rate signal 5, temperature signal 6, and pressure signal I 7 of the gas flowing through the guy pipeline 9, and uses the above method to detect an accident at an early stage and output it to the alarm display device 8.
Also, depending on the severity of the accident, the optimal value of load distribution for each power plant is calculated, and the target load of each power plant after the accident occurs and the load reduction rate from the load before the accident is sent to each power plant. In addition to automatically preventing the power plant from tripping,
Extend the time until the trip.

The method for determining the target load and load reduction rate is as follows. A gas flow rate accident level L is determined according to the reduction in gas flow rate. Further, let the load of each power plant at t immediately before the occurrence of the accident be G (where j indicates each power plant), and let τ be the elapsed time from the occurrence of the accident to the present moment. At this time, the target load GO and load reduction rate RO of each power plant are as follows:
,bWOi Ko! FIG. 2 is a block diagram of an accident detection device showing an embodiment according to the present invention.

The input input unit 21 inputs analog signal inputs of current or voltage such as the gas main pipe flow rate F1 temperature T and pressure P due to the gas differential pressure of the vibration line at a 1-second cycle, and converts them into the engineering conversion unit 22.
Converts the input into a value in engineering units.The flow rate calculation unit 23 corrects the flow rate based on the gas differential pressure converted into the engineering unit value using temperature and pressure, and calculates a corrected gas flow rate close to the true value. The general correction formula is as follows (
3) It is shown in the formula. F: Gas flow rate (Nm3) K: Proportional constant (value considering reference temperature and pressure) P: Actual pressure (K9/CFll) T: Actual temperature (°C) h: Actual differential pressure (Tfr!Ft) Average smooth part 24 performs average smoothing processing on the correction gas flow rate, etc. to remove periodic fluctuations, thereby preventing false alarms due to fluctuations.

The average smoothing process is performed using the moving average method shown by the following formula. x: Input data Y: Average smoothed value k: Number of averages Further, the rate of change calculation unit 25 calculates the rate of change of the average value obtained by moving average processing, and the duration calculation unit 26 calculates the time period exceeding a predetermined level. It calculates how long it lasted.

Accident detection 27 is then performed after this has continued for a certain period of time. Then, depending on the situation, an operation signal is output to the operation instruction section 28. FIG. 3 shows an example of a change in the gas flow rate signal before the moving average smoothing process.

FIG. 4 is an example of the gas flow rate after the average smoothing process. However, the value when k=8 is shown. It can be seen that the periodic fluctuations in the gas flow rate have been eliminated and smoothed out. Rate of change calculation part 2
In step 5, the rate of change is determined using the value from which periodic fluctuations have been removed.

The duration calculation unit 26 compares the rate of change with several predetermined threshold values and calculates the time during which each threshold value is continuously exceeded. In this calculation, complicated time calculations can be simplified by dividing the rate of change into the following regions and calculating the duration of each region. D1
:Divided area, R: Calculation of the actual measurement rate of change duration is based on the above divided area (from D1 to DN
) in which region the measured change rate R is found is calculated by adding the current time to the previous duration corresponding to that region.

When the respective durations in the rate of change region DN~1 up to the previous time are t1~N, the actual measured value R this time is 1 α1 × R
If ≦-α1-1 (RCDl) and the duration is CT, the duration up to this time is as follows.

- -ΔT indicates the difference in measurement time between the previous time and this time (1 second in this embodiment), and RCD1 indicates that R falls into the region D1.

That is, as shown in the diagram, the method of adding the duration time is as follows. In Fig. 5, the rate of change is in the area DN (where N=1, 2, .
・N), if it continues from ζ to T5, the rate of change is considered to have continued for regions D1 to DN-1 that are smaller than the absolute value of DN and its rate of change -α, -1, and the rate of change is assumed to be continuous, and the duration is changed to the previous time. Add the time ΔTN from time ζ to T that continued in DN.

That is, the rate of change in region D1 is T. -T, considered continuous. Next, the magnitude of the rate of change is from T5 to T in region D2.
When continuing up to 6, the change rate areas Dl and D2 are obtained by adding the time ΔT2 from T5 to T6 to the previous duration time as in the previous time, and conversely, the change rate areas D3 to DN with large change rate absolute values are the same as the previous time. Reset the duration of the rate of change to ゜゜0゛. That is, since the change rate regions D3 to DN are not continuous, they are reset. FIG. 6 shows that the rate of change region in D2 continues from T4 to T5 from the state in FIG. This is a diagram of the situation.

By using the above-described duration calculation method, irregular and sudden fluctuations in the flow rate are removed and only signals under normal conditions are captured.

It should be noted that the threshold values -α1 to -αN are determined so that an accident can be most easily identified based on the average smoothed signal actual measurement value and the simulation result of the accident.

This method of calculating the duration adopts a moving average approach to time. That is, the feature is that the duration in the distant past is reset, and the duration in an area close to the present is monitored.

In the accident detection section 27 in FIG.
The duration determined in step 6 is compared with a predetermined accident area to detect an accident.

When determining the accident area pattern, a sufficient margin should be taken for the actually measured pattern of the rate of change duration to prevent misjudgments.
In addition to trying to prevent false alarms, it is necessary to accurately capture anticipated accidents through simulations in advance. FIG. 7 shows an example of a flowchart outlining the accident detection section and the operation instruction section, and FIG. 8 shows an example of the accident area.

That is, when -α1, a failure is determined when the duration reaches T1, and when -α2, T2L, . . . , -α, a failure is determined when TNL is reached. TNL<T(
There is a relationship of N-0L<...2L<TlN, and when the absolute value of the rate of change is large, it is determined that there is a failure even if the duration is short. Further, as shown in FIG. 6, if the process does not continue, it is reset and the continuation time is monitored. In the flowchart of FIG. 7, in block 71, based on the result of the duration calculation in block 26 of FIG. 2, it is determined, for example, where in the area shown in FIG. It is possible to determine whether a drop in main pipe pressure has occurred, and if there is a pressure drop, notify each power plant of the extent of the accident and perform corresponding controls, such as load distribution control. In addition, if there is an increase in main pipe pressure even if the area is not an accident area, or if the main pipe pressure increases rather than decreases while it is in an accident area, it is determined that a valve failure has occurred, and processing is performed such as issuing an alarm to each power station. will be carried out. In block 28 of FIG. 2, the gas flow rate before the accident occurrence and the current gas flow rate are compared, the degree of the accident is determined, and the occurrence of the accident and its degree are reported to each power station.

For example, depending on the severity of the accident, load target values and
The load reduction rate is determined and output as follows.

Now, if L is the flow rate reduction rate, the load reduction amount ΔG of each power plant
O, is determined by.

That is, the target load GO, is GO,=G,−ΔGO,. G
, is the amount of power generated at each power plant immediately before the accident occurred, and n is the number of power plants. In addition, the load reduction rate RO is determined by:

Here a is the allowable reduction rate. γ is a coefficient that determines the magnitude of the load reduction rate, and its optimum value is determined by the simulation lane. According to an embodiment of the present invention, by detecting a carburetor accident early and outputting the target load and load reduction rate to each power plant, tripping of the power plant can be automatically prevented and the time until tripping can be reduced. Since it is possible to extend the period and necessary measures can be taken during that period, there is also the additional effect of improving reliability.

As described above, according to the accident detection device according to the present invention, the gas flow rate change rate and By monitoring the duration, accidents in the vaporizer can be detected quickly and reliably.

[Brief explanation of the drawing]

Figure 1 is a general power generation system diagram using LNG, Figure 2 is a computer processing block diagram, Figure 3 is the value before average smoothing of the gas flow rate of the Vibration line, and Figure 4 is the value of the gas flow rate of the Vibration line. Values after average smoothing processing, Figures 5 and 6 are diagrams showing the change rate region and duration time, Figure 7 is a schematic flow chart of the accident detection section and operation instruction section, and Figure 8 is the flow rate change rate and its duration time. FIG. 1... Accident detection device, 2... LNG tank, 3...
・LNG vaporization equipment, 4... Gas main pipe main valve, 10...
gas vibe line.

Claims (1)

[Claims] 1. In a method for detecting an abnormality in a gas supply source to a power generation plant, the gas flow rate of a main pipe supplying gas is corrected and detected according to the gas condition, and the detected flow rate is moved. The average value is calculated for each predetermined period, and the rate of change in that period is calculated, and a limit value for the duration is set in advance according to the magnitude of the rate of change, and the calculated rate of change value is the change rate. A method for detecting an abnormality in a gas supply source, characterized in that an abnormality is determined when a duration limit value set corresponding to a rate is reached. 2. An abnormality in a gas supply source, characterized in that the duration limit value set forth in claim 1 is compared when the rate of change value is small, and a smaller limit value is set when the value of the rate of change is large. Detection method. 3. Abnormality detection of a gas supply source, characterized in that in detecting the gas flow rate as set forth in claim 1, the detected gas flow rate is corrected by gas pressure or gas temperature and used as a detected value of the gas flow rate. Method.