JPH042919B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to a fuel failure detection device that detects a core fuel failure accident in a nuclear power plant with high sensitivity.

[Technical background of the invention and its problems] Early detection of nuclear fuel failure accidents is important in order to ensure the safety of nuclear plants and reduce radiation exposure during repair. By the way, when nuclear fuel is damaged, fission products are released into the coolant, so delayed neutrons released from ¹⁸⁷ I, ⁸⁷ Br, etc. contained in the released fission products are transferred to the vicinity of the reactor core exit. A known method is to detect fuel damage by detecting it with a delayed neutron detector installed outside the coolant pipe, but this method
It is affected by background components such as neutrons that reach the delayed neutron detector directly from the reactor core, and beta and gamma rays emitted from the coolant or impurities in the coolant.

Therefore, by shielding the area around the delayed neutron detector, efforts are being made to reduce the effects of the background components mentioned above, but one constraint is that the detection delay time must be small, so this detection Since the installation location of the device is limited, it is difficult to completely eliminate the influence of background components as described above.

In particular, in sodium-cooled fast breeder reactors,
The largest background component is neutrons arriving directly from the core, and the second largest background component is γ-rays from ²⁴ Na, which is the activation of sodium in the coolant.
Since the value of the count rate signal due to these background components changes depending on the reactor output and its history, the threshold value for determining the presence or absence of fuel damage must be changed depending on the operating conditions. There was a problem in that there was a high possibility of overlooking changes in the counting rate when they occurred.

[Object of the invention] The present invention has been made in view of the above circumstances, and
The objective is to monitor the correlation coefficient between the count rate signal of delayed neutrons emitted from fissile material released into the coolant and the coolant temperature signal, even in the presence of high levels of background components. It is an object of the present invention to provide a nuclear fuel damage detection device capable of detecting fuel damage with high sensitivity without being influenced by the above.

[Summary of the Invention] In order to achieve the above object, the present invention includes a delayed neutron detector installed in a coolant pipe near the reactor core exit, and a linear counter into which a detection signal from the delayed neutron detector is input. a count rate meter, a thermometer for measuring the reactor outlet coolant temperature, a digital converter for digitally converting the outputs of both the linear count rate meter and the thermometer, and a count rate signal X _K input from the digital converter. and the correlation coefficient ρ _K of the monitoring index from the temperature signal T _K
A fuel damage detection device comprising: a monitoring index calculation unit that calculates the _value of It is something. In addition, the correlation coefficient ρ _K is calculated online using the count rate signal X _K and the temperature signal T _K , and if the correlation between the reactor output and the coolant temperature signal is strong _, The predicted count rate signal X^ _K is created in advance using a regression model, and instead of X _K , (X _K
−X^ _K ).

[Embodiments of the invention] An embodiment of the present invention will be described with reference to the drawings.

FIG. 1 shows a block diagram of an embodiment of the present invention. In the figure, a delayed neutron detector 1 is placed outside the coolant pipe near the reactor core exit, and the detection signal detected by the delayed neutron detector 1 is amplified by a preamplifier 2 and a main amplifier 3 and then linearly counted. After being input to rate meter 4, its output is A/D
The signal is input to a converter 7. On the other hand, a temperature signal measured by a thermometer 5 for measuring the temperature of the reactor outlet coolant is amplified by an amplifier 6 and input to an A/D converter 7. Both input signals input to the A/D converter 7 are
It is configured so that it is compared with a threshold value α via a monitoring index calculation unit 8, which will be described later, in a comparison and determination unit 9, and when it is larger than this threshold value α, it is determined that fuel damage has occurred, and an alarm is displayed on an alarm display unit 10. There is.

The block diagram in Figure 1 shows the measurement system for one delayed neutron detector.
When delayed neutron detectors are placed in multiple coolant pipes, the coolant temperature signals in the corresponding pipes are used. Further, the detection device according to the present invention is realized by a microcomputer or the like, and can handle a plurality of measurement systems simultaneously.

First, the function of the monitoring index calculating section 8 in the block diagram of FIG. 1 will be explained.

Now, from the count rate signal X _K quantized by the A/D converter 7 and the temperature signal T _K at each time interval Δt, the correlation coefficient ρ _K as a monitoring index is calculated by the following equations (1) to (4). Calculated by However, the saphix K is at time t _K , K−
1 _{represents time} t _{K-1 = t K − △ t .} ) ...(1) Y′ _K = Y _K-1 + (X _K −X _K-1 ) ² U′ _K = U _K-1 + (T _K −T _K-1 ) ² U′ _K = U _{K -1} + (T _K −T _K-1 ) ² V′ _K = V _K-1 + (X _K −X _K-1 )・(T _K −T _K-1 )……(2)
Y _K =Y′ _K +ω{(X _K −X _K ) ² −Y′ _K } U _K =U′ _K +ω{(T _K −T _K ) ² −U′ _K } U _K =U′ _K +ω{ (T _K −T _K ) ² −U′ _K } V _K =V′ _K +ω{(X _K −X _K )(T _K −T _K )−V′ _K }……(3)
ρ _K = V _K / (Y _K・U _K ) ^1/2 ...(4) In other words, the above equations (1) to (4) become sequential calculation formulas to obtain the correlation coefficient ρ _K for every △t. ing.

Note that ω in the above equations (1) and (3) is a weighting coefficient.

Next, the function of the comparison/judgment section 9 will be explained.
This comparison/determination section 9 compares the latest value of the correlation coefficient ρ _K with a threshold value α determined by taking into account the range of variation in the value of the correlation coefficient ρ _K obtained under normal operating conditions without damage. If ρ _K >α, it is determined that fuel damage has occurred, and the information is input to the alarm display unit 10.

The configuration described above is the main component of the nuclear fuel failure detection device of the present invention, but when there is a strong correlation between the reactor output and the coolant temperature signal T _K , the background component of the count rate signal X _K and the cooling Material temperature signal
The correlation with T _K becomes stronger and the threshold value α becomes larger, which may lead to a decrease in detection sensitivity. In such a case, a regression model such as the following equation (5) that calculates the change in the count rate signal predicted by the temperature signal T _K ,
That is, X^ _K = _M 〓 ^m=1 (a _n・X _Kn + b _n・T _Kn )
...(5) is created in advance, and (X _K −X^ _K ) is expressed in (1) ~
Sensitivity can be improved by using it instead of X _K in equation (3).

Next, the nuclear fuel damage detection procedure according to the present invention is carried out in a second manner.
The flowcharts in Figures A and B will be explained. In this flowchart, "Start"
From or until "end", it is necessary to use the comparison judgment threshold α and the above equation (5) (F=1) or not (F
= 0), and if necessary, the parameter values of the above equation (5) and the procedure for determining them.

Now, when a nuclear fuel damage detection start command is issued, the parameters K _s , K _n , ω and α _n are set as the first step 101, where K _n is the comparison judgment threshold α
and the number of data points used to determine the parameters in equation (5) above, and ω is the weighting coefficient used in equations (1) and (3) above, taking into account the frequency characteristics of the signal and the time interval of A/D conversion. shall be determined. K _s is calculated according to (1) to (3) above when evaluating the fluctuation of ρ _K under normal operating conditions.
The initial values of _K , _K , U _K , Y _K , and V _K in the equation _{are 0} , ₀ ,
This gives the maximum time during which the influence of U ₀ , Y ₀ , and V ₀ remains on ρ _K , and the calculated value within this time (t _K < t _Ks ) must not be taken into consideration. (e.g. ₀ = ₀
=O, Y ₀ =U ₀ =V ₀ =0, then ρ ₁ =1.0 regardless of the values of X ₁ and T ₁ . ) Therefore, K _s is given by considering the value of ω and each of the above initial values. α _n is K=K _s
If the maximum value of ρ _K obtained between ~K _n is larger than α _n given here, it is determined that it is necessary to use the regression model of equation (5) above.

Next, first set F=0 and start the second step.
102 gives initial values ₀ , ₀ , Y ₀ , U ₀ , V ₀ . Then, with K=1, in the third step 103, X _K and T _K at time t _K are input, and in the fourth step 104, F
If ≠0, then in the fifth step 105, this X _K , T _K
Save. F≠0 and K at the 6th step 106
If ≧M, in the seventh step 107, X^ _K is calculated and _XK = _XK - X^ _K is calculated. Also, the 6th step
If K≧M is not satisfied in 106, the process returns to the third step 103 with K=K+1.

In the eighth step 108, the average values _K , _K are calculated using the data from the fifth step 105 and the seventh step 107.
(Equation (1)) and Y′ _K , U′ _K , V′ _K (Equation (2)) and variance value Y _K ,
Calculate U _K , V _K (Equation (3)) and correlation coefficient ρ _K (Equation (4)). In the next ninth step 109, K≧K _s ,
If ρ _K ≦α in the 10th step 110, the procedure of the 11th step 111 is directly executed, and if ρ _K ≦α, α=ρ _K and the next procedure of the 11th step 111 is executed. Further, if K≧K _s is not found in the ninth step 109, and if K=K _n is not found in the eleventh step 111, the process returns to the third step 103 with K=K+1.

If K=K _n in the 11th step 111, the next 12th step 112 is executed. i.e. the 12th
In step 112, if α > α _n, proceed to the 13th step
Step 113 is executed, and if α>α _n does not hold, then step 115 is executed.

However, if F=0 in the 13th step 113, then the
14 Data saved in step 114
Using X _K and T _K (K=1 to K _n ), a regression model X^ _K = _M 〓 ^m=1 (a _n ·X _Kn +b _n ·T _Kn ) is determined. Then, the second step is performed with F=1.
Return to 102. Also, if F=0, in the 15th step 115, α=Cα is set as the final threshold.
Here, C is given in advance, and the online monitoring threshold is set to be C times the maximum value of ρ _K obtained from the data of K=K _s to _{K n} .

In the next 16th step 116, the following initial values for online monitoring are set. ₀ = _Kn ,
T ₀ = _Kn , Y ₀ = Y _Kn , U ₀ = U _Kn , V ₀ = V _Kn. Then, in the next seventeenth step 117, it is determined whether to proceed to online monitoring. If the process does not proceed to online monitoring, the nuclear fuel damage detection procedure ends. By the way, the processing up to the 17th step 117 is based on F, α
This is a preparation for determining M, a _n , b _n (m=1 to M) in the case of F=1.

Moving on to online monitoring, X _K and T _K are input in the 18th step 118 with K=1. In the next 19th step 119, if F≠0, the procedure of the 20th step 120 is executed, and if F≠0, the procedure of the 22nd step 122 is executed, skipping the 20th step 120 and the 21st step 121. executed. 20th step 120
If K≧M does not hold, the process returns to the 18th step 118 and a similar procedure is executed for K=K+1. Also, if K≧M holds true, the 21st step 121
Calculate X^ _K by calculating equation (5) above,
Let X _K =X _K −X^ _K. Furthermore, the next 22nd step
122 executes the calculations of equations (1) to (4) above and calculates the monitoring index ρ _K
Calculate. In the next 23rd step 123, ρ _K >
α is determined, and if this determination is established, the next
In the 24th step 124, an alarm is generated, and in the 25th step 125, the result is displayed.

And again, the 18th step for K=K+1
The steps 118 to 25th step 125 are repeated.

[Effects of the Invention] According to the nuclear fuel damage detection device of the present invention, it is possible to remove the influence of high-level background radiation that exists in a state where no damage has occurred, so changes in counting rate due to fuel damage can be detected with high sensitivity. can. Therefore, it is possible to detect fuel damage accidents at an early stage, help reduce radiation exposure during repairs, and further improve the safety of nuclear reactor plants.

[Brief explanation of drawings]

FIG. 1 is a block diagram of an embodiment of the present invention.
FIGS. 2A and 2B are flowcharts for explaining the nuclear fuel damage detection procedure according to the present invention. 1...Late neutron detector, 4...Linear count rate meter, 5
...Reactor outlet coolant thermometer, 7...A/D converter,
8...Monitoring index calculation section, 9...Comparison/judgment section, 10...Alarm display section.

Claims (1)

[Claims] 1. A delayed neutron detector disposed in a coolant pipe near the reactor core exit, a linear count rate meter that inputs a detection signal from the delayed neutron detector, and a reactor exit coolant temperature detector. A thermometer that measures the temperature, a digital converter that converts both the outputs of the linear count rate meter and the thermometer into digital signals, and a digital converter that calculates the phase of the monitoring index from the count rate signal X _K and temperature signal T _K that are input from the digital converter. Consisting of a monitoring index calculation section that calculates the correlation coefficient ρ _K , and a comparison determination section that compares the correlation coefficient ρ _K input from the monitoring index calculation section with a threshold value α to determine the presence or absence of fuel damage. A fuel damage detection device featuring: 2 Correlation coefficient ρ _K from count rate signal X _K and temperature signal T _{K
are expressed} by the following equations (1) _to ( ₄ ), _that is, _{X K =} …(1) Y′ _K = Y _K-1 + (X _K −X _K-1 ) ² U′ _K = U _K-1 + (T _K −T _K-1 ) ² U′ _K = U _K-1 +(T _K −T _K-1 ) ² V′ _K =V _K-1 +(X _K −X _K-1 )・(T _K −T _K-1 )……(2)
_{Y K = Y} ^′ _{K + ω {} ⁽ _{X _ _ _} (T _K −T _K ) ² −U′ _K } V _K =V′ _K +ω{(X _K −X _K )(T _K −T _K )−V′ _K }……(3)
ρ _K = V _K / (Y _K・U _K ) ^1/2 ...(4) Therefore, saphix K is at time t _K , K-1 is at time t _K-1 = t _K -△t, and ω is 2. The fuel failure detection device according to claim 1, wherein the fuel failure detection device calculates on-line using a weighting factor. 3 If there is a strong correlation between the reactor output and the coolant temperature signal, a regression model expressed by the following formula that predicts the count rate from the temperature signal T _K , that is, X^ _K = _M 〓 ^m=1 (a _n・X _Kn + b _n・T _Kn ) is created in advance, and instead of X _K (X _K −X^ _K )
The fuel damage detection device according to claim 2, which uses the following.