JPH07209473A - Atomic reactor core performance calculating device - Google Patents

Abstract

PURPOSE:To perform an estimation calculation in a short time by providing a means for storing the midway convergence values of a characteristic value calculable every void repeat, and providing a means for estimating and calculating the final convergence result of the characteristic value as the function of the stored midway convergence values. CONSTITUTION:As the measurement data of the core 11 of a reactor 10, the read value phim of the present control rod position CR and flow rate F are read to an input processing means 2. In a nuclear heat hydraulic power calculating part 3, neutron bundle distribution phi and output distribution P, and void distribution V are calculated in a nuclear calculating part 3a and a heat hydraulic power calculating part 3b, respectively, while the mutual output results are taken as the other input. When an estimation calculation is required, a user delivers the set value of CR or F to the processing part 2. Thereafter, void repeat is carried out, and the midway convergence value K1 of the characteristic value is delivered to a characteristic value storing part 4 every void repeat. The void repeat is ended three times, and the final convergence value is estimated by a characteristic value calculating part 5 by use of the characteristic value midway convergence values K2, K3 of the second and third void repeats stored in the housing part 4.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a reactor core performance calculation device for calculating or monitoring a reactor power distribution and the like.

[0002]

2. Description of the Related Art Conventionally, for monitoring and predicting the core of a nuclear reactor,
TIP (Traversing In-Core Probe), LPRM (Local P
ower range monitor) and other nuclear neutron counters are used, and at the same time, nuclear thermohydraulic characteristics such as eigenvalue of reactor, three-dimensional power distribution, neutron flux distribution, margin from thermal limit value, etc. In order to obtain it, a 3D core simulator based on a physical model is provided. As a physical model, neutrons are classified into high-speed group, out-heat group, heat group, etc. by energy, and a diffusion equation that follows the neutron flux of each group is derived, and then so-called modified one-group diffusion equation is solved. Nuclear calculation models are often used.

The diffusion equation is usually solved by dividing the fuel assembly, which has several hundreds in the horizontal direction, into one lattice and dividing it into a lattice of ten to twenty nodes in the axial direction and discretizing it. By the way, this nuclear calculation model includes a nuclear constant determined by the void distribution, burnup distribution, presence of control rods, etc. in the core as a coefficient. For example, when the control rod is operated, the nuclear constant changes depending on the presence or absence of the control rod, the neutron flux distribution changes, the output distribution changes, and the void distribution in the reactor changes, so the nuclear constant changes. The change in the void distribution due to the change in the output distribution is calculated by the thermal-hydraulic calculation model. Therefore, in order to find a consistent neutron flux distribution and void distribution, the variables on the grid are replaced many times and the nuclear calculation and thermal-hydraulic calculation are repeated, so-called void iteration is performed.

Conventionally, when predicting a core state at a planned point, first, actual data of the current core state (output, flow rate, control rod position, etc.) is taken in, and flow rate and control are performed in accordance with a target output level. Set the change width of the bar position. This setting determines an approximate value according to the output of the reactivity, the flow rate, the sensitivity with respect to the control rod position, and the presence or absence of a thermal margin, which are obtained in advance.

Next, by performing three-dimensional core calculation in this state, the reactor eigenvalue, neutron flux distribution, power distribution, thermal margin, etc. are obtained. This three-dimensional core calculation takes time because it performs the void iteration, which is the repetition of the above-mentioned nuclear calculation and thermal-hydraulic calculation.

By the way, in order to keep the reactor critical at the planned point, it is necessary to set the reactor eigenvalue to 1.0 in consideration of the error of the three-dimensional diffusion model. In this way, the initial setting is a rough state, so normally the reactor eigenvalue does not become 1.0, and the reactivity (that is, the deviation from the eigenvalue of 1.0) and the sensitivity of the core state are corrected again. Then, set another state and recalculate. This calculation is repeated until the state at the prediction point becomes critical. That is, many trial calculations are required to find a critical state. If the thermal margin in the state where the criticality can be finally satisfied satisfies the constraint condition, it is adopted as the planned point.

[0007]

As described above, in the predictive calculation, first, several trial calculations are required to search the state where the reactor becomes critical. For this reason, it is necessary to perform three-dimensional core calculation repeatedly with time-consuming void iterations, and there is a problem that the prediction calculation time is very long.

In order to reduce the time required for such a search,
(1) A method to find the eigenvalue in the set state well and reduce the number of void iterations in 3D nuclear thermal-hydraulic calculation, (2) A method to find and search the critical core state from the eigenvalue obtained in the set state. A method of reducing the number of times can be considered.

An object of the present invention is to provide a reactor core performance calculation device that executes prediction calculation in a short time.

[0010]

In order to achieve the above object, the present invention provides a means for storing the intermediate convergence value of eigenvalues that can be calculated for each void iteration, and the eigenvalues of the eigenvalues as a function of the stored intermediate convergence values. It is configured to provide a means for predicting and calculating the final convergence result.

[0011]

That is, conventionally, one trial calculation is completed by repeating the iterations until the power distribution and the void distribution converge as a result of the void iteration which conventionally requires repeated calculation of multivariables in the three-dimensional core. However, what is needed as a result of trial calculation for critical state search is the eigenvalue, which is a single core value, and the result of three-dimensional quantity (power distribution and void distribution)
Is unnecessary.

By the way, the eigenvalue is estimated for each void iteration in the trial calculation, and the iteration is repeated to converge to the final converged value. The eigenvalues of the core tend to converge extremely monotonically, as compared to the amount that locally changes drastically with each iteration of voids, such as the core power distribution and void distribution. Therefore, it is possible to obtain the eigenvalues with sufficient accuracy as a trial calculation by using the eigenvalues several times in the early stages of void iteration and then estimating the final convergence value by finding the change after that by function approximation. .

According to the present invention, the three-dimensional core calculation is performed by providing a means for predicting the final convergent value of the eigenvalue as a function of the intermediate convergent value (core single point value) of the eigenvalues obtained by the first three void iterations. There is no need to converge to the end, and one trial calculation can be completed. As a result, the number of void iterations required for each trial calculation can be significantly reduced and the total calculation time can be shortened.

FIG. 2 shows in detail the details of the calculation time which can be shortened by the present invention.

A search for a critical state is performed by trial calculation T _(i)
Until the eigenvalue k _(i) obtained as a result becomes k (= 1.0), the core state is changed based on the sensitivity with the reactivity (that is, the deviation from 1.0 of the eigenvalue) and repeated. (Fig. 3, middle row). The void iteration U _(ji) is repeated for each conventional trial calculation T _(i) . When predicting a particularly large state change, the number of iterations is about 10 times (upper part of FIG. 3).

However, in the present invention, the number of void iterations, which was required 10 times to converge one trial calculation, is terminated by three times, and the convergence value is k for each void iteration.
Save _(ji) . Then, by calculating this three times a function of the eigenvalues G _(i), to estimate the T eigenvalue convergence value k of _{(i) (i).} This eigenvalue estimation value may be used to change the setting of the core state. Function of eigenvalue at one core G _(i)
The time required for calculation can be neglected compared to the calculation time for one void iteration required for three-dimensional core calculation, so the calculation time required for one trial calculation can be reduced to about 1/3 (lower part of Fig. 3). .

In order to know the final distribution, after determining the flow rate and the control rod position by trial calculation using this means, each distribution and eigenvalue may be obtained until the void iteration converges only once.

[0018]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A first embodiment of the operation plan creating apparatus of the present invention will be described below with reference to FIG. In FIG. 1, 1 is a core performance calculation device, 2 is an input processing unit, 3 is each thermal-hydraulic calculation unit, 3
a is a nuclear calculation unit, 3b is a thermal-hydraulic calculation unit, 4 is a eigenvalue intermediate convergence value storage unit, 5 is a final eigenvalue convergence value calculation unit, 10 is a reactor, and 11 is a core. The operation of each calculation unit will be described in detail below.

At the time of normal monitoring calculation, the current control rod position CR, the flow rate F, and unillustrated nuclear instrumentation (TIP, LPRM, etc.) of the reactor core 11 are measured data.
The reading value φm and the like are read by the input processing unit 2. In the nuclear thermal-hydraulic calculation unit 3, the neutron flux distribution φ, the output distribution P, etc. are calculated in the nuclear calculation unit 3a, and the void distribution V is calculated in the thermal-hydraulic calculation unit 3b.
Etc. are calculated with each other's output results as the other's input.

FIG. 3 shows details of the operation of the nuclear thermal-hydraulic calculation unit 3. That is, in the nuclear calculation unit 3a, the coefficient of the diffusion equation is corrected using the measured data φm under the condition that the reactor is critical (the eigenvalue is 1). Next, diffusion calculation is performed to obtain the neutron flux distribution and the eigenvalue k _j . Further, the output distribution P is calculated from the neutron flux distribution and transferred to the thermal-hydraulic calculation unit 3b. The thermal-hydraulic calculation unit 3b calculates the flow distribution, pressure loss, etc., and determines the void distribution V. Using the void distribution V, the coefficient of the diffusion equation is corrected, and the nuclear thermal-hydraulic calculation is repeated (void repetition).

Now, when the prediction calculation becomes necessary, the user delivers the set value of CR or F to the input processing unit 2. The whole procedure of the prediction calculation is shown in FIG. After the state is set in the input / output processing unit 2, the nuclear thermal-hydraulic calculation unit 3 repeats the void.
Further, as shown in FIG. 3, the intermediate convergence value k _j of the eigenvalue is passed to the eigenvalue storage unit 4 for each void iteration j. This k
_j is a value in the middle of the void iteration, and is still different from the final converged value. In this device, this void iteration is aborted in three times, and the eigenvalue calculation unit 5 uses the eigenvalue intermediate convergence values k ₂ and k ₃ stored in the eigenvalue storage unit 4 for the second and third void iterations to obtain the final convergence value. To estimate.

FIG. 5 shows how the eigenvalue is estimated. When k ₂ and k ₃ are transferred from the eigenvalue storage unit 4 to the eigenvalue calculation unit 5, the coefficient of the function G defined inside the eigenvalue calculation unit 5 is determined. In FIG. 5, as a function of k ₂ and k ₃ of the eigenvalue calculator 5,

[0023]

## EQU1 ## A case where G (x) = x / (ax + b) (Equation 1) is selected is shown. Where x is the number of void iterations, a,
b is a constant, and can be obtained by substituting k ₂ and k ₃ for x = ₂ and ₃ in Equation 1. When the coefficient is determined, the calculated value of the final convergence value is immediately obtained as a value 1 / a (a is a function of k ₂ and k ₃ ) when x is infinite.

The state in which the final convergence value of void iteration is often expressed by a function such as equation 1 is shown below. FIG. 6 shows transitions of eigenvalues during void repetition when the void repetition is continued until various distributions and eigenvalues finally converge with respect to various state changes. It can be seen that when the state changes significantly, around 10 void iterations are required to converge. Further, it can be seen that the eigenvalue convergence value at the first iteration of voids may be significantly affected by the initial state as when the control rod is inserted, but monotonically increases and decreases after the second iteration. In this embodiment, in order to reduce the number of void repetitions as much as possible, the second and third convergence values k ₂ and k ₃ are used. Here, the results when the equation 1 is selected as the function of k ₂ and k ₃ of the eigenvalue calculation unit 5 are shown by black circles. The solid line is the transition of the eigenvalue when the void repetition is actually repeated. For each state change, the final convergence value of void repetition (value at the right end of the solid line) and the estimated value (black circle) according to Equation 1 are in good agreement. Therefore, the user can use this estimated value to estimate the core state that is actually critical from the relationship between the eigenvalue, the control rod position, the flow rate, and the like. In the present invention, three-dimensional multivariable iterative calculation is performed only three times, and since the calculation time of Equation 1 is negligible, this trial calculation is performed as compared with the case where the eigenvalue at the end of void iteration is finally used. The time required for can be reduced to about 1/3.

As described above, one trial calculation is completed, and the final eigenvalue estimation value k estimated to be finally obtained when the void iteration is repeated is obtained. If this calculated value is not 1, the sensitivities (output, flow rate, control rod position) to various reactivities are corrected, the critical state is reset, and the next trial calculation is repeated. This calculated value is 1
Then, finally, the output distribution, the thermal margin, etc. are calculated by calculating only once until the void iterations converge.

According to the present embodiment, the number of void iterations required for one trial calculation can be reduced to about 1/3, and the calculation time can be shortened.

[0027]

According to the present invention, it is possible to reduce the time required for three-dimensional nuclear thermal-hydraulics calculation in trial calculation for adjusting the flow rate and control rod, and to shorten the entire prediction calculation time.

[Brief description of drawings]

FIG. 1 is a block diagram of a reactor core performance calculation device of the present invention.

FIG. 2 is an explanatory diagram showing a breakdown of time reduction of the reactor core performance calculation device of the present invention.

FIG. 3 is a flowchart of nuclear thermal hydraulic calculation.

FIG. 4 is a flowchart showing a prediction calculation procedure of the present invention.

FIG. 5 is a flowchart of estimating an eigenvalue final convergence value.

FIG. 6 is an explanatory diagram showing eigenvalue estimation results in various core states.

[Explanation of symbols]

1 ... Reactor core performance calculation device, 2 ... Input processing unit, 3 ... Nuclear thermal-hydraulic calculation unit, 3a ... Nuclear calculation unit, 3b ... Thermal-hydraulic calculation unit,
4 ... Eigenvalue storage unit, 5 ... Eigenvalue calculation unit, 10 ... Reactor,
11 ... Reactor.

Claims (3)

[Claims] 1. A reactor, which receives heat output of a nuclear reactor, core flow rate, control rod pattern, and count value of an in-core neutron detector as input, and a void distribution obtained as a result of thermal-hydraulic characteristic calculation as input. Means for calculating nuclear characteristics such as eigenvalues, three-dimensional neutron flux distribution, and power distribution, and the result of the nuclear characteristics calculation,
The obtained output distribution is used as an input, and means for calculating void distribution, thermal-hydraulic characteristics such as a margin from a thermal limit value, and the above-mentioned nuclear characteristics and thermal-hydraulic characteristics are repeatedly calculated. In the reactor performance calculation device that determines the state,
It is characterized by further comprising means for storing a convergent value of eigenvalues for each iteration, and means for calculating a final convergent value of eigenvalues by using a function of convergent values in the iterative calculation of a plurality of eigenvalues. Reactor core performance calculator. 2. The nuclear reactor core performance calculation device according to claim 1, wherein the means for storing the convergent value of the eigenvalue stores the convergent value of the second or third iteration. 3. The reactor core performance calculation according to claim 1, wherein the function of the intermediate convergence value of the iterative calculation of the plurality of eigenvalues is a function of the intermediate convergence value of the second and third eigenvalues of the iterative calculation. apparatus.