JPH0545490A - Monitoring device for reactor power - Google Patents

Abstract

(57) [Summary] [Purpose] Detects local vibrations or vibrations that cause a phase difference that are difficult to detect with an average output range monitor (APRM), and contributes to safe driving by suppressing vibrations. [Constitution] An adder 4 and an A / D converter 5 for inputting an LPRM signal 3 from a neutron detector 2 arranged in a nuclear reactor core 1 are installed. A statistical calculator 6 for inputting the digital signal converted by the A / D converter 5 is connected. First and second memories 7 and 10 for inputting the reference signal calculated by the statistical calculator 6 are installed. The signal of the first memory 7 is input to the weight multiplier 8 and the first averaging device 9 to which the output signal of the weight multiplier 8 is input is connected to transmit the first new APRM signal. The signal of the second memory 10 is sequentially updated within the time set by the time delay, and the second averaging device 12 is provided to
The new APRM signal of is transmitted. Current A from adder 4
An output fluctuation mode determiner 16 that compares the PRM signal with the first and second new APRM signals is installed to estimate the output fluctuation mode.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a reactor power monitoring device used for power monitoring of a boiling water reactor.

[0002]

2. Description of the Related Art Boiling water reactor (hereinafter referred to as BWR)
About 16 neutron flux detectors are arranged for 16 fuel assemblies in the core, and four neutron detectors are vertically arranged for each one neutron flux detector. These neutron detectors are called local power range monitors (hereinafter referred to as LPRMs), and for example, 1100 MW class BWR core has 43 x
4 = 172 detectors. Further, each of these detector signals is averaged by about 20 signals, which is called an average output area monitor (hereinafter referred to as APRM). 1100MW class B
In WR there are 8 channels of APRM signals, all of which are analog signals. Normally, these APRM signals are monitored, and trip signals such as a scrum signal are issued when the signals exceed a certain set point to prevent the reactor from operating in a dangerous state.

[0003]

The APRM signal is an LPR signal.
Since the M signal is averaged evenly, it is possible to detect when the output of the entire core fluctuates, but when it fluctuates locally in the core or when it fluctuates with a spatial phase difference, the fluctuation amount is averaged. It may be difficult to detect because it is fooled. As an example of fluctuating locally in the core, a phenomenon called so-called channel oscillation, in which a fuel assembly that is severe in terms of thermo-hydraulic force generates an oscillation phenomenon called density wave oscillation, is considered to be diffused by neutron flux oscillation. May vary only within a relatively narrow range. In addition, as a variation with a spatial phase difference, there is an oscillation phenomenon called a regional vibration that vibrates with a phase difference of 180 degrees at target positions in the core, and it is actually observed in some overseas plants. .. For example, CAORS from Italy
In the regional vibration observed in the O plant, the maximum amplitude of the APRM was at most about 10%, while the maximum amplitude of the LPRM was up to 60%. This is because the half and half of the core are oscillating with a phase difference of 180 degrees, so the maximum and minimum values of LPRM are averaged at the same time, and canceling occurs between them.

As described above, there is an output fluctuation phenomenon that is difficult to detect if only the APRM is monitored.
In addition, monitoring all LPRMs and linking them with the trip signal poses a reliability problem. That is, APRM is L
Since the PRMs are uniformly averaged over the entire core, there is a problem that it becomes difficult to detect a local output vibration or an output vibration that causes a phase difference.

The present invention has been made to solve the above problems, and it is possible to detect an output fluctuation that may be overlooked by simply monitoring a conventional signal, using a conventional signal, It is intended to provide a reactor output monitoring device capable of improving safety and operating rate.

[0006]

SUMMARY OF THE INVENTION The present invention is directed to a number of local power range monitors located within a nuclear reactor core, an A / D converter for digitally converting the analog signals from these local power range monitors, and A statistical arithmetic unit for inputting the digital signal converted by the A / D converter, a first memory and a second memory for storing the reference signal calculated by the statistical arithmetic unit, and a connection to the first memory A weight multiplier for inputting the reference signal thus converted, and a first new average output area monitor signal connected to the weight multiplier for averaging
And an analog from the local output area monitor, and a second averager that inputs the lag time from the second memory and outputs an averaged second new average output area monitor signal. It is characterized by comprising an adder for inputting a signal and an output fluctuation mode judging device for inputting a current average output area monitor signal from the adder.

[0007]

According to the present invention, when the APRM is calculated from the LPRM, the current analog signal is AD-converted by the A / D converter, and the averaged APRM signal is calculated by the statistical calculator, the weight multiplier and the averaging device. In addition, digitized LP
The APRM is calculated from the RM signal by two different calculation methods. First, a low pass filter is applied to remove high frequency noise from the digitized LPRM signal, and the variance is obtained for each signal. When calculating the variance,
The history is updated every certain set short time, and the past history is stored in the first and second memories for a certain period. The variance value is used to judge the importance of the signal and to judge the detector failure. A signal whose variance is large on average is the reference signal,
Compute the cross-correlation function of that signal with other signals. Like the variance, this cross-correlation function is updated every set short time.

Considering the sine wave oscillation, since the amplitude is proportional to the square root of the variance, that is, the standard deviation, it is considered that the standard deviation of each signal represents the importance of the signal to the APRM. The ratio of the standard deviation of each signal to the standard deviation is used as a weight when calculating the APRM. Further, the delay time (lag time) at which the cross-correlation function with the reference signal becomes maximum is considered to correspond to the phase difference between both signals.

The first APRM signal assigns this weight to each L
The PRM signal is multiplied and averaged, and the second APRM signal is averaged using a signal of a time corresponding to the delay time from the reference signal (or this may be weighted). Since the two types of APRMs thus obtained and the current APRM signal have different types of output fluctuations that are most easily detected, it is possible to determine the type of output fluctuations by comparing the three types of signals. ..

Further, it is possible to detect the failure of the neutron detector from the history of changes in the dispersion value stored in the memory. For example, if the dispersion value is almost 0, it means that the signal has not come in and the neutron detector has failed or has been broken. Therefore, by issuing the alarm and removing the signal from the APRM calculation signal, it is possible to obtain more reliability. Highly adaptive APRM
The signal is obtained. In addition, if the dispersion fluctuates abnormally, that is, even if it suddenly increases in a short time, it is also possible that the neutron detector is malfunctioning, in addition to the fact that the output locally fluctuates greatly. Therefore, if the standard of the fluctuation rate is set in advance, and it exceeds the standard, and is abnormally larger than the dispersion value of the other 3 neutron detectors on the same axis, or the nearby neutron detectors, the neutron detector is malfunctioning. Given that there is, take the same measures as before. By such a measure, a reliable APRM signal can always be obtained.

Further, the two types of averaging give almost the same response as the current APRM signal to the phenomenon that the output of the entire core fluctuates uniformly. There is no need to change the convenience.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a reactor power monitoring device according to the present invention will be described with reference to the drawings. In FIG. 1, reference numeral 1 indicates a reactor core arranged in a reactor pressure vessel of a boiling water reactor. In this reactor core 1, about 16 neutron fluxes are detected in 16 fuel assemblies. The device is located. Four neutron detectors 2 are arranged vertically in each one neutron flux detection device. LP from neutron detector 2
An adder 4 for inputting the RM signal 3 via a signal line and an A / D converter 5 are installed. A statistical calculator 6 for inputting the converted digital signal is connected to the output side of the A / D converter 5. A first memory 7 and a second memory for inputting the calculated reference signal to the output side of the statistical calculator 6
10 are installed. A weight multiplier 8 and a first averager 9 are connected in series on the output side of the first memory 7.
The first averager 9 outputs the first new APRM signal 14. On the other hand, the output side of the second memory 10 has a lag time of 11
Is connected to the second averaging device 12, and the second averaging device 12 outputs a second new APRM signal 15.

From the output side of the adder 4, the current AP
The RM signal 13 is output, and this APRM signal 13 is input to the output fluctuation mode determination unit 16, and this determination unit 16 is connected to the signal lines for comparison and examination with the first and second new APRM signals 14 and 15 input. Has been done. The output side of the statistical calculator 6 is connected to the weight multiplier 8 and the input side of the lag time 11.

Next, a method of monitoring the output of the nuclear reactor by the nuclear reactor output monitoring device having the above configuration will be described. In FIG.
A large number of neutron detectors 2 are present in the core 1, and each LPRM signal 3 is extracted from these neutron detectors 2 as an analog signal. The LPRM signals 3 are collected by the adder 4 and are evenly averaged every 20 signals to become the APRM signal 13. The APRM signal 13 is constantly monitored, and when a certain set level is exceeded, a trip signal such as scrum is issued. Up to here, the method according to the conventional example,
The current APRM signal 13 is referred to, and the APRM signal 13 is input to the output fluctuation mode determiner 16.

The respective analog signals are digitized from the respective LPRM signals 3 using the A / D converter 5. In the statistical calculator 6, the digitized LPRM signal is first subjected to appropriate preprocessing, that is, noise removal by a low-pass filter, a trend component test, etc., and then the variance is calculated within a certain set time (a few minutes). To be done. Here, the variance σ ² (k) of the k-th signal is defined by the following equation (1).

[0016]

[Equation 1]

Here, the calculated variance value is stored in the first memory 7 for a certain period which is sufficiently longer than the interval for which the variance is calculated. From the variance value of each signal, the standard deviation σ (k), which is the square root of the signal, is calculated, and the ratio of the signal required for each APRM calculation to the average standard deviation value is calculated and used as the weighting coefficient. Each signal is multiplied by the weighting coefficient by the weighting multiplier 8 and averaged by the first averaging device 9 to obtain the first new AP.
It becomes the RM signal 14.

Further, the statistical calculator 6 selects a reference signal from the importance of the signal based on the calculated variance value, and calculates the cross-correlation function between this reference signal and other signals. Here, the cross-correlation function C _k (m) of the k-th signal is generally defined by the following equation (2).

[0019]

[Equation 2]

Here, y (t) is a reference signal, M is the number of signals sampled in the time used for the cross-correlation function calculation, at the same level as the variance calculation, and m is the delay time (lag time).
11) is the number of sampling times, and L is the maximum value thereof, which is the number of sampling times within the basic cycle of vibration and about 2 seconds. The delay time at which the value of this cross-correlation function becomes maximum corresponds to the phase difference from the reference signal. That is, if the delay time is 0, there is no phase difference between the two, but if the delay time is half of the vibration period, the phase difference is 180 degrees, that is, a completely opposite phase. The lag time 11 obtained here is stored in the second memory 10. Then, the time-series data referenced from the lag time 11 is extracted from the second memory 10, and the signals are averaged by the second averaging device 12 to obtain the second new APRM signal 15.

Thus, three types of APRM signals 13, 14 and 15 have been calculated. The current APRM signal 13 is used as a trip signal as before, but the first and second new APRM signals 14, 15 can be used as well. At that time, the output fluctuation mode determiner 16 compares the three types of signals to make a more reliable determination.

The effects of the present invention will be described below with respect to three types of output vibration phenomena. First, the first new APRM signal 14
Is particularly excellent in detecting local vibration phenomena. Eg 20
If we average the signals, one of them has an amplitude of 50%,
The remaining 19 signals are 5%. If this is simply averaged as before, the amplitude of the APRM signal is (19 × 5 + 1 × 50) /20=7.25%, whereas if the weighting of the present invention is used, (19 × (5 / 7.25) X5 + 1 x (50 / 7.25) x 50) / 20 = 20.52%, which shows that the sensitivity is nearly three times that of the conventional example. Using a three-dimensional dynamic characteristic analysis code that can actually simulate the core in detail and calculate all LPRM signals, one specific fuel assembly serves as a vibration source and the output locally vibrates. FIG. 2 shows a comparison between the APRM of the conventional example and the APRM signal obtained by using the weight of the present invention by simulating the existing state. While the amplitude of the APRM of the conventional example is about 9%, the amplitude of the APRM of the present invention is about 39%, which is four times or more of the sensitivity, and the present invention can cope with such a local output fluctuation. It turns out that it is very effective.

Next, the second new APRM signal 15 is excellent in detecting an event that spatially fluctuates with a phase difference in the core. An example of this is the phenomenon called regional vibration, which is actually observed in some overseas plants. Here again, a three-dimensional dynamic characteristic analysis code is used to simulate the regional vibration phenomenon and compare it with the conventional example. First, FIG. 3 shows a correlation function obtained from the simulated result. As the reference signal, the signal having the largest amplitude and closest to the vibration source was selected. From this figure, the time delay at which the cross-correlation function becomes maximum is about 1.05 seconds, which corresponds to 1/2 of the oscillation period, indicating that both signals oscillate in completely opposite phases. Therefore, if this is simply averaged, it can be seen that the peaks and troughs of the vibrations overlap and cancel, and the amplitude is damped. Further, in this case, even though the phases are opposite to each other, the variances are almost the same, and therefore the above-mentioned weighting does not improve. Therefore, it can be seen that if the averaging is performed in consideration of this time delay, the phases are aligned, and the peaks and peaks of the vibrations and the valleys and valleys of the vibrations match and canceling is not performed. FIG. 4 shows a comparison between the conventional example and the APRM signal obtained by the present invention. The amplitude of the conventional example is about 4%, whereas the amplitude of the present invention is about 19%, an improvement of about 5 times. You can see that it is done.

The noise level of a neutron detector is usually 10%
The above two examples are the APRM of the conventional example.
However, according to the present invention, it is possible to detect any of them.

On the contrary, the conventional method and the method according to the present invention will be compared with respect to an event that can be detected sufficiently even in the conventional example and uniformly fluctuates in the whole core. FIG. 5 shows a phenomenon in which the whole core is vibrating uniformly by comparing three signals of the conventional APRM 13, the first new APRM 14 and the second new APRM 15 according to the present invention. The amplitudes are almost the same. From this, it is understood that the level of the trip signal determined using the APRM signal of the conventional example does not need to be changed. Therefore, the mode of output fluctuation can be determined by comparing these three types of signals. This determination is performed by the output fluctuation mode determiner 16 in FIG. That is, from the three types of examples described above, the APRM of the conventional example, the first new APRM, and the second new APR
M is almost the same response → Core uniform fluctuation 1st new APRM changes significantly larger than others → Core local fluctuation 2nd new APRM changes significantly larger than others → Large fluctuations with spatially phase difference It can determine 3 different modes. If the output fluctuation mode can be specified, the optimum trip level and countermeasure can be set for each fluctuation. For example, the local core fluctuation is locally stricter than the uniform core fluctuation in terms of safety margin, and the trip level needs to be set lower than that of the uniform core fluctuation, but it is set low. In order to prevent the core uniform variation from outputting a trip signal at the level, it can be avoided by first determining the mode by the power variation mode determiner 16.

Further, although the output fluctuation mode described above can be monitored by monitoring the individual LPRM signals as they are, the number of detector signals becomes very large, and therefore the present invention is reliable in terms of determination. There is a problem compared to.

That is, in the reactor output monitoring apparatus of this embodiment, the AP for averaging the analog signal as it is at present
Compared to the RM signal, it is possible to perform detection and determination more accurately, including output fluctuations that are difficult to detect with conventional signals, and it is possible to improve safety and improve the operating rate.

[0028]

As described above, according to the present invention, it becomes possible to monitor the reactor output fluctuation more accurately than in the conventional case, and it is possible to improve the safety and the operation rate.

[Brief description of drawings]

FIG. 1 is a system diagram showing an embodiment of a reactor output monitoring device according to the present invention.

FIG. 2 is a weighted APRM signal and a conventional APRM signal.
The characteristic diagram which shows a signal by comparing about a local output fluctuation event.

FIG. 3 is a characteristic diagram showing a relationship between a cross correlation function and a lag time.

FIG. 4 is an APRM signal using lag time and A of a conventional example.
The figure which compared the PRM signal about the output fluctuation event with a spatial phase difference.

FIG. 5: AP of a conventional example regarding a uniform core power fluctuation event
FIG. 6 is a characteristic diagram showing a comparison between an RM signal and two types of signals according to the present invention.

[Explanation of symbols]

1 ... Reactor core, 2 ... Neutron detector, 3 ... LPRM signal, 4 ...
Adder, 5 ... A / D converter, 6 ... Statistical calculator, 7 ... First
Memory, 8 ... Weight multiplier, 9 ... First averaging device, 10 ...
Second memory, 11 ... lag time, 12 ... second averaging device,
13 ... Current APRM signal, 14 ... First new APRM signal,
15 ... Second new APRM signal, 16 ... Output fluctuation mode determiner.

Claims (1)

[Claims] 1. A large number of local power range monitors arranged in a nuclear reactor core, an A / D converter for digitally converting analog signals from these local power range monitors, and the A / D converter.
Connected to the first memory, a statistical calculator for inputting the digital signal converted by the / D converter, a first memory and a second memory for storing the reference signal calculated by the statistical calculator A weight multiplier for inputting the reference signal, a first averaging device connected to the weight multiplier for outputting an averaged first new average output area monitor signal, and the second averaging device.
A second averager for inputting the lag time from the memory to output an averaged second new average output area monitor signal, and an adder for inputting the analog signal from the local output area monitor, A reactor output monitoring device, comprising: an output fluctuation mode determination device for inputting a current average output region monitor signal from an adder.