Comparison of statistical properties of coolant temperature and neutron flux in a reactor PWR-1000
Using wavelet analysis an existence of regularities in “noise” signals from neutron flux detectors and temperature ones of a design in-pile control system of PWR-1000 reactor was investigated. For the purpose to check the possibility to estimate coolant parameters through neutron flux parameters and...
Збережено в:
| Дата: | 2017 |
|---|---|
| Автори: | , , , , |
| Формат: | Стаття |
| Мова: | English |
| Опубліковано: |
Інститут проблем безпеки атомних електростанцій НАН України
2017
|
| Назва видання: | Проблеми безпеки атомних електростанцій і Чорнобиля |
| Теми: | |
| Онлайн доступ: | http://dspace.nbuv.gov.ua/handle/123456789/159108 |
| Теги: |
Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
|
| Назва журналу: | Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| Цитувати: | Comparison of statistical properties of coolant temperature and neutron flux in a reactor PWR-1000 / A.D. Skorbun, S.M. Stadnyk, V.G. Kotelenets, O.L. Arvaninov, A.V. Drobetskii // Проблеми безпеки атомних електростанцій і Чорнобиля: наук.-техн. зб. — 2017. — Вип. 29. — С. 22-26. — Бібліогр.: 9 назв. — англ. |
Репозитарії
Digital Library of Periodicals of National Academy of Sciences of Ukraine| id |
irk-123456789-159108 |
|---|---|
| record_format |
dspace |
| spelling |
irk-123456789-1591082019-09-24T01:25:27Z Comparison of statistical properties of coolant temperature and neutron flux in a reactor PWR-1000 Skorbun, A.D. Stadnyk, S.M. Kotelenets, V.G. Arvaninov, O.L. Drobetskii, A.V. Проблеми безпеки атомних електростанцій Using wavelet analysis an existence of regularities in “noise” signals from neutron flux detectors and temperature ones of a design in-pile control system of PWR-1000 reactor was investigated. For the purpose to check the possibility to estimate coolant parameters through neutron flux parameters and vice versa, a comparative analysis of statistical characteristics for parallel long regular sets of measurements of coolant temperature and a neutron flux was carried out. It were analyzed neutron signals from three measurement canals with ionization chambers, and from temperature sensors for coolant of the cool and hot branches of the first contour. In both variables a rich spectrum of various long time (tens of minutes and hours) periodicities has been revealed. Some of them are found both in temperatures and in the neutrons flux. It was however revealed also periodicities, which in neutrons and in a coolant do not correlate between themselves. The differences between the pictures of wavelet transform of temperatures of hot and cold branches of the first contour have been shown. In the temperature of hot branch additional more high frequency vibrations, which can be ascribed to development of turbulence, are arisen. The conclusion was drawn, that fluctuations of signals, that have been analyzed, are not random noise, but are the data carrier about neutron subsystem of a reactor. It is suggested to carry out the analysis of fluctuations of signals of a design in-pile control system for early detection of a reactor operating conditions. Із використанням вейвлет-аналізу досліджено наявність закономірностей у «шумових» сигналах детекторів нейтронного потоку і температури теплоносія штатної системи внутрішньореакторного контролю реактора ВВЕР-1000. Виконано порівняльний аналіз довгих рядів вимірювань температури теплоносія і щільності потоку нейтронів. В обох параметрах виявлено багатий спектр різноманітних періодичностей. Деякі з них знайдено як у даних температури, так і потоку нейтронів. Проте в динаміці потоку нейтронів і температури виявлено також періодичності, які не корелюють між собою. Показано відмінності між картинами вейвлет-перетворення температур гарячої і холодної ниток: у температурі гарячої нитки виникають додаткові більш високочастотні коливання, які можна приписати розвитку турбулентності. Зроблено висновок, що проаналізовані флуктуації сигналів не є випадковим шумом, а є носіями інформації про стан нейтронної підсистеми реактора. Пропонується виконувати аналіз флуктуацій сигналів штатної системи контролю для діагностики режимів роботи реактора. С использованием вейвлет-анализа изучено существование закономерностей в «шумовых» сигналах детекторов потока нейтронов и температуры теплоносителя штатной системы внутриреакторного контроля реактора ВВЭР-1000. Выполнен сравнительный анализ длинных рядов измерений температуры теплоносителя и плотности потока нейтронов. В обоих параметрах обнаружен богатый спектр разнообразных периодичностей. Некоторые из них найдены как в температурных данных, так и данных потока нейтронов. Однако в динамике потока нейтронов и температуры обнаружены также периодичности, которые не коррелируют между собой. Показаны отличия между картинами вейвлет-преобразования температур горячей и холодной ниток: в температуре горячей нитки возникают дополнительные более высокочастотные колебания, которые можно приписать развитию турбулентности. Сделан вывод, что проанализированные флуктуации сигналов не являются случайным шумом, а являются носителями информации о состоянии нейтронной подсистемы реактора. Предлагается выполнять анализ флуктуаций сигналов штатной системы внутриреакторного контроля для диагностики режимов работы реактора. 2017 Article Comparison of statistical properties of coolant temperature and neutron flux in a reactor PWR-1000 / A.D. Skorbun, S.M. Stadnyk, V.G. Kotelenets, O.L. Arvaninov, A.V. Drobetskii // Проблеми безпеки атомних електростанцій і Чорнобиля: наук.-техн. зб. — 2017. — Вип. 29. — С. 22-26. — Бібліогр.: 9 назв. — англ. 1813-3584 http://dspace.nbuv.gov.ua/handle/123456789/159108 621.039.564 en Проблеми безпеки атомних електростанцій і Чорнобиля Інститут проблем безпеки атомних електростанцій НАН України |
| institution |
Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| collection |
DSpace DC |
| language |
English |
| topic |
Проблеми безпеки атомних електростанцій Проблеми безпеки атомних електростанцій |
| spellingShingle |
Проблеми безпеки атомних електростанцій Проблеми безпеки атомних електростанцій Skorbun, A.D. Stadnyk, S.M. Kotelenets, V.G. Arvaninov, O.L. Drobetskii, A.V. Comparison of statistical properties of coolant temperature and neutron flux in a reactor PWR-1000 Проблеми безпеки атомних електростанцій і Чорнобиля |
| description |
Using wavelet analysis an existence of regularities in “noise” signals from neutron flux detectors and temperature ones of a design in-pile control system of PWR-1000 reactor was investigated. For the purpose to check the possibility to estimate coolant parameters through neutron flux parameters and vice versa, a comparative analysis of statistical characteristics for parallel long regular sets of measurements of coolant temperature and a neutron flux was carried out. It were analyzed neutron signals from three measurement canals with ionization chambers, and from temperature sensors for coolant of the cool and hot branches of the first contour. In both variables a rich spectrum of various long time (tens of minutes and hours) periodicities has been revealed. Some of them are found both in temperatures and in the neutrons flux. It was however revealed also periodicities, which in neutrons and in a coolant do not correlate between themselves. The differences between the pictures of wavelet transform of temperatures of hot and cold branches of the first contour have been shown. In the temperature of hot branch additional more high frequency vibrations, which can be ascribed to development of turbulence, are arisen. The conclusion was drawn, that fluctuations of signals, that have been analyzed, are not random noise, but are the data carrier about neutron subsystem of a reactor. It is suggested to carry out the analysis of fluctuations of signals of a design in-pile control system for early detection of a reactor operating conditions. |
| format |
Article |
| author |
Skorbun, A.D. Stadnyk, S.M. Kotelenets, V.G. Arvaninov, O.L. Drobetskii, A.V. |
| author_facet |
Skorbun, A.D. Stadnyk, S.M. Kotelenets, V.G. Arvaninov, O.L. Drobetskii, A.V. |
| author_sort |
Skorbun, A.D. |
| title |
Comparison of statistical properties of coolant temperature and neutron flux in a reactor PWR-1000 |
| title_short |
Comparison of statistical properties of coolant temperature and neutron flux in a reactor PWR-1000 |
| title_full |
Comparison of statistical properties of coolant temperature and neutron flux in a reactor PWR-1000 |
| title_fullStr |
Comparison of statistical properties of coolant temperature and neutron flux in a reactor PWR-1000 |
| title_full_unstemmed |
Comparison of statistical properties of coolant temperature and neutron flux in a reactor PWR-1000 |
| title_sort |
comparison of statistical properties of coolant temperature and neutron flux in a reactor pwr-1000 |
| publisher |
Інститут проблем безпеки атомних електростанцій НАН України |
| publishDate |
2017 |
| topic_facet |
Проблеми безпеки атомних електростанцій |
| url |
http://dspace.nbuv.gov.ua/handle/123456789/159108 |
| citation_txt |
Comparison of statistical properties of coolant temperature and neutron flux in a reactor PWR-1000 / A.D. Skorbun, S.M. Stadnyk, V.G. Kotelenets, O.L. Arvaninov, A.V. Drobetskii // Проблеми безпеки атомних електростанцій і Чорнобиля: наук.-техн. зб. — 2017. — Вип. 29. — С. 22-26. — Бібліогр.: 9 назв. — англ. |
| series |
Проблеми безпеки атомних електростанцій і Чорнобиля |
| work_keys_str_mv |
AT skorbunad comparisonofstatisticalpropertiesofcoolanttemperatureandneutronfluxinareactorpwr1000 AT stadnyksm comparisonofstatisticalpropertiesofcoolanttemperatureandneutronfluxinareactorpwr1000 AT kotelenetsvg comparisonofstatisticalpropertiesofcoolanttemperatureandneutronfluxinareactorpwr1000 AT arvaninovol comparisonofstatisticalpropertiesofcoolanttemperatureandneutronfluxinareactorpwr1000 AT drobetskiiav comparisonofstatisticalpropertiesofcoolanttemperatureandneutronfluxinareactorpwr1000 |
| first_indexed |
2025-07-14T11:41:19Z |
| last_indexed |
2025-07-14T11:41:19Z |
| _version_ |
1837622393625903104 |
| fulltext |
22 ISSN 1813-3584 ПРОБЛЕМИ БЕЗПЕКИ АТОМНИХ ЕЛЕКТРОСТАНЦІЙ І ЧОРНОБИЛЯ 2017 ВИП. 29
УДК 621.039.564
A. D. Skorbun
1
, S. M. Stadnyk
1
, V. G. Kotelenets
2
, O. L. Arvaninov
2
, A. V. Drobetskii
2
Institute for Safety Problems of Nuclear Power Plants, NAS of Ukraine, Kirova str., 36a, Chornobyl, 07270, Ukraine
2
Separated department «South-Ukraine NPP» of the state enterprise «NNEGC «Energoatom», Mikolaiv oblast, Yuzh-
nourainsk, 55000, Ukraine
COMPARISON OF STATISTICAL PROPERTIES OF COOLANT TEMPERATURE
AND NEUTRON FLUX IN A REACTOR PWR-1000
Using wavelet analysis an existence of regularities in “noise” signals from neutron flux detectors and tempera-
ture ones of a design in-pile control system of PWR-1000 reactor was investigated. For the purpose to check the possi-
bility to estimate coolant parameters through neutron flux parameters and vice versa, a comparative analysis of statistic-
al characteristics for parallel long regular sets of measurements of coolant temperature and a neutron flux was carried
out. It were analyzed neutron signals from three measurement canals with ionization chambers, and from temperature
sensors for coolant of the cool and hot branches of the first contour. In both variables a rich spectrum of various long
time (tens of minutes and hours) periodicities has been revealed. Some of them are found both in temperatures and in
the neutrons flux. It was however revealed also periodicities, which in neutrons and in a coolant do not correlate be-
tween themselves. The differences between the pictures of wavelet transform of temperatures of hot and cold branches
of the first contour have been shown. In the temperature of hot branch additional more high frequency vibrations, which
can be ascribed to development of turbulence, are arisen. The conclusion was drawn, that fluctuations of signals, that
have been analyzed, are not random noise, but are the data carrier about neutron subsystem of a reactor. It is suggested
to carry out the analysis of fluctuations of signals of a design in-pile control system for early detection of a reactor oper-
ating conditions.
Keywords: nuclear reactor, neutrons, coolant, noise statistics.
Introduction
In the theory of nuclear reactors it is generally known [1], that between the state of a coolant and
neutron field there is nonlinear connection – in dependence on pressure, temperatures and other factors,
probability of moderation of neutrons changes, that in turn causes changes in probability of fission processes
and energy release. Among them boiling of water, that is appearance in it vapor bubbles, especially influ-
ences on the value of such connection. Methodology of detecting of boiling processes in PWR reactors [2, 3]
is based on such connection, where it is suggested to draw conclusion about the presence of boiling on the
increase of dispersion of neutron flux noise.
Thus at now neutron flux fluctuations are connecting with fluctuations of coolant parameters, and
beginning of boiling it is proposed to detect through control the value – relation of dispersion of a neutron
flux noise to its mean value [2]. On the assumption of this, in the given work in-depth statistical analysis of
noises of neutron detectors signals has been carried out with the purposes to reveal other possible parameters
of these signals, which would characterize connection of the neutron subsystem with a coolant.
With that end in view the sets of measurements of coolant temperatures and neutron flux, which
were carried out simultaneously, have been analyzed. These data are received as a result of regular mea-
surements of mentioned parameters with the design devices of the in-pile control system. The task was given
to show, how much the changing of these parameters in time are interconnected.
Because to make purposeful experiments in a reactor is practically impossible, the results, which
were received in a given work, are compared with the results of [2], which are based on special experiments.
In [2] the next assertion were formulated, which can be checked on real data:
1. When boiling appearing, dispersion of neutrons noises is increasing. At this the value of a neutron
flux proper can be not changed.
2. The pointed changes are fixed, when amplitude of coolant temperature fluctuations at input in a
reactor exceed approximately 0.2 °С.
3. The effect (a value of ) is maximal when frequency of fluctuation is in the region of 0.5 Hz (the
result of model calculation with NOSTRA code and comparison with an experiment).
Experimental data and methods of analyzing
In the given work signals from control equipment of the system of in-pile control, which were re-
ceived during the work of the reactor facility PWR-1000 in steady-state conditions at nominal power, were
analyzed. That is for receiving such information it’s not necessary to use the additional special equipment.
© A. D. Skorbun, S. M. Stadnyk, V. G. Kotelenets, O. L. Arvaninov, A. V. Drobetskii, 2017
COMPARISON OF STATISTICAL PROPERTIES
________________________________________________________________________________________________________________________
ISSN 1813-3584 ПРОБЛЕМИ БЕЗПЕКИ АТОМНИХ ЕЛЕКТРОСТАНЦІЙ І ЧОРНОБИЛЯ 2017 ВИП. 29 23
Data. Due to a comparative analysis of changing of coolant temperature and neuron flux was
planned to carry out, than data of measurements from two systems have been analyzed.
1. Signals of an equipment for neutron flux control from three measurement channels 1, 8, 15 (which
in what follow will be named K1, K8, K15), which are shifted on 120
о
relative to each other in a cross-
section of a reactor (Fig. 1 see cover page 2). At nominal power devices for neutron flux detecting (ioniza-
tion chambers) work in a current mode, and each channel performs information acquisition and treatment for
signals from two blocks of neutron flux detecting (BDNF): lower – BDNFl, and upper – BDNFu. These
detecting blocks (for all three channels) are placed approximately on equal height relative to the middle of a
reactor core. Integrated data from BDNFl and BDNFu are the data from one channel. Thus, it is analyzed an
information, which is sufficiently averaged over reactor core volume.
2. Signals from temperature sensors, which are placed at the input and output of a coolant from a
reactor core:
YA21T31 – coolant temperature on the outlet from a reactor the second branch of the main circula-
tion contour (measuring tool – a resistance sensor);
YA22T32 – coolant temperature on the input in a reactor the second branch of the main circulation
contour (measuring tool – a resistance sensor).
Analyzing methods. The main task, on the analogy of [2], was to search correlations between
changes of temperature and neutron fluxes with the aim finding possibilities of an early diagnostics, at the
expense of use special sensitive methods of analyzing. In this work it was supposed, that it is easier to find
correlation, analyzing periodical signals. About existing of such periodicities in many sensor it had known
[4].
For searching periodicities a wavelet analysis was used, with the help of which it is possible not only
to reveal a periodical phenomena, but also clear to fix intervals of time, when they arise and disappear, a
shift of the signals in a time between different detectors and so on [5, 6, 7]. Gaussian wavelets of 2-th and
10-th orders were used as very convenient to reveal periodicities. The result of wavelet transform is a two-
dimensional matrix of coefficients, the value of which will be maximal, when a function, which is analyzed,
will contain corresponding frequencies.
This matrix can be represent graphically, where on horizontal axis of this two-dimensional image the
consecutive number of an array element is drawn (at sampling periodicity in one second it will be simply
time in seconds), and along vertical axis the so called scale of wavelet, which is proportional to period, is
drawn. The values of wavelet coefficients can be drawn in colors or shades of gray. If “noise“ signal during
some time interval contains sinusoidal component, then in image of wavelet coefficients a series of regular
“spots” will appear, the position of which will corresponds to maximums and minimums of this sinusoid.
The data from BDNF were recorded with sampling period Tp = 1 s, and sampling of temperature was
carried out at Tp = 0.67 s. Thus, in contrast to measurement, for instance, vibration frequencies with special
devices, in our analysis the frequencies, greater than 1/Tp , cannot be revealed.
Results and discussion
It turned out, that frequency-temporal structure of fluctuations of analyzed signals is enough rich and
therefore cannot be represent in one image in details. Therefore two different situations were examined sepa-
rately: short-period and long-period processes in the same signals. This gives the possibility, using “micro-
scopic” abilities of wavelet analysis [6], to demonstrate periodicities in a wide range of frequencies.
Long-period processes
Fig. 2 (see cover page 2) shows the results of wavelet transformation of the neutron flux density sig-
nals for all three channels: K1, K8, and K15 in the units of nominal power (% Nnom). In the upper part of
each figure there is a graph of changes in the neutron flux density over time. Two-dimensional picture is the
result of wavelet transformation of this signal. On the horizontal axis, the consecutive number of the element
of a series of measurements is marked. Since the sampling is regular and goes with intervals in one second,
then for neutrons the horizontal axis is a time from the beginning of measurements in seconds. On a vertical
axis the period is set in units of the values of the horizontal axis. Exactly the period of spot series can be
determined using software as the distance between the spots. The duration of the measurements was about 30
000 s, that is, more than 8 hours.
In the behavior of the neutron flux density one can see the following. 1. In addition to fluctuations
from about 99.8 to 101 (% Nnom) there is a trend: the signal gradually decreases, then increases. 2. Also, there
are quite sharps jumps of the signal (shown by the arrows in Fig. 2b) (see cover page 2). The reasons for
A. D. SKORBUN, S. M. STADNYK, V. G. KOTELENETS ET AL.
________________________________________________________________________________________________________________________
ISSN 1813-3584 ПРОБЛЕМИ БЕЗПЕКИ АТОМНИХ ЕЛЕКТРОСТАНЦІЙ І ЧОРНОБИЛЯ 2017 ВИП. 29 24
such behavior of the neutron flux are not discussed at present. 3. For the analysis, we have specially chosen
the time interval at which these features are observed in order to demonstrate the capabilities of the method
for our task: to record both the appearance of features in the neutron flux and their manifestation in the re-
sults of wavelet transformation.
In pictures of wavelet transforms of neutron signals one can see the following. 1. Pictures in general
have the same appearance for all three channels. 2. In all of the pictures there is a series of three large spots
(indicated by the arrows in Fig. 2a) (see cover page 2) This means the existence in the signal a periodical
component with a half-period, which is equal to the distance between two adjacent spots. It is programmati-
cally possible to determine the position of these spots, the distance between them and thus to calculate the
period, which turned out to be approximately 7.2 hours. 3. The position of the spots on the horizontal axis is
the same for all three graphs, that is, within the accuracy of the measurements, time shift is no detected be-
tween them. As a minimum, this means that the field of neutrons in the volume of the reactor core on such
time intervals changes more or less same.
Fig. 3 (see cover page 2) shows the results of wavelet transform for cold branch temperature
(YA22T32 sensor) and hot thread (YA21T31 sensor) of coolant. This makes it possible to directly compare
the pictures of the wavelet transform for neutrons and for temperatures. The first thing to note is that in the
signals for both temperatures and in a neutron signal periodicity of 7.2 hours is presented also (three large
spots in the lower part). Secondly, that in the wavelet transform of the temperature of the hot branch there is
a high-frequency component which is absent both in neutrons data, and in the temperature of the cold branch.
This effect is appeared as black random structure in an upper part of the image. Upper part of the image for
cold water is gray in contrary. And the third thing to note is that the temperature curve of the cold branch
resembles (trend and jumping) the neutron behavior pattern. At the same time, for the temperature of the hot
branch, the jump in the center is barely noticeable, and the trend is not visible to the naked eye, although
wavelet transform clearly marks out these features.
Thus, from the comparison of the results of wavelet analysis of signals from neutron sensors and
temperature sensors, one can conclude that the noise component of these signals contains the same informa-
tion about the periodicity, hidden in them. Particularly clear correlation manifests itself when comparing the
neutron signal and the temperature of the cold branch of the coolant. In turn, this suggests that the fluctua-
tions of the analyzed signals are not random noise, but reflect the dynamics of the corresponding parameter
(neutrons or temperature), since otherwise there would be no correlation found.
In [2] by the calculation method it was shown that the change in the dispersion of the noise neutron
signal is noticeable (statistically with the given reliability) when the coolant temperature changes at the reac-
tor input at 0.2 - 1 °C. This is a fairly noticeable value, since (Fig. 3 see cover page 2), in the normal mode of
operation, at small intervals of time, temperature fluctuations are observed in the region of < 0.1°С for cold
branch, and < 0.4 °С for hot branch. The graphs show, that large-scale 7.2 hours trend gives only 0.1 °C for a
cold branch, and without this trend the value of fluctuations are even smaller. And just the 7.2 hour trend
appears in the form of large spots on a two-dimensional picture. If on the graph for neutrons the trend is ob-
served with the naked eye, then at the temperature, especially hot, it is not so. However, wavelet analysis
easily detects it in all cases. That is, the sensitivity of this method of analysis is much higher. This allows us
to hope that the appearance of boiling can be fixed at much earlier stages.
Consequently, it can be assumed that in this paper it is shown experimentally that there is a correla-
tion between the behavior of the fluctuations in the temperature of the cold branch and the neutrons fluctua-
tions. Thus, it is directly shown that the fluctuations of these signals carry the same information as the
changes in the dispersion of the signals, used in [2].
Short-periodic processes
It must be emphasize that the above shown two-dimensional pictures are only a graphical representa-
tion of the general two-dimensional matrix of coefficients of the wavelet transform, and any parts of this
matrix of the transform can be analyzed separately. That is, all the information shown in Fig. 4 (see cover
page 3) on the cover there was also in Figs. 2 and 3 (see cover page 2), but hidden due to incomparability of
the corresponding coefficients of the wavelet transform.
It was suggested at signals analyzing, that the reactor work in a design mode and boiling is absent.
But it should to notice, that in a reality on a given time instrumental methods for control of boiling presence
are absent. In Figs. 2 and 3 (see cover page 2) we see a correlation in changing of temperature and neutrons
with time. But observed period in 7 hours is too big to detect such a phenomenon as boiling, and to take
COMPARISON OF STATISTICAL PROPERTIES
________________________________________________________________________________________________________________________
ISSN 1813-3584 ПРОБЛЕМИ БЕЗПЕКИ АТОМНИХ ЕЛЕКТРОСТАНЦІЙ І ЧОРНОБИЛЯ 2017 ВИП. 29 25
measures. Therefore, the question arises: can one see something in a shorter period of time and for what pe-
riod of time to calculate, say, ?
To do this, we will analyze our data sets, for example, in the first 10 000 seconds, to avoid, at the
same time, the manifestations of a leap of values in the middle of the sets and to analyze only statistical fluc-
tuations. The results are shown in Fig. 4. (see cover page 3) In these figures, first, it is clearly seen that the
signals of both temperature and neutrons have a rich internal structure and are therefore not accidental in the
statistical sense of the word. Recall that regularly located series of spots mean the occurrence and disappear-
ance of periodicities with sufficiently big amplitude. Such series of spots in Fig. 4 (see cover page 3) exist at
different times and on different scales.
Secondly, although such periodicity is apparent on both graphs (both for neutrons and for tempera-
ture), however, careful consideration of drawings, and careful analysis with software shows that only long
time periodicities are common for the analyzed signals.
But there are other interesting peculiarities. Thus, temperature has clearly apparent spots series with
period about 5700 c (shown by the red rectangle), that is about 47 min. All other spot series for hot branch
have a behavior, more inherent to random data. The cold branch on the scale about 500 sec has a clear peri-
odicity, which correspond to ~ 5 min period (green rectangular). Thus, long time periodicities are revealed
both in neutrons and in temperatures. But more high frequency oscillations, which are revealed in tempera-
tures, are not connected with neutron ones (yellow rectangle).
Thus it is revealed, that in the fluctuations both neutrons and temperatures there are enough noticea-
ble oscillations, which are interconnected only partially. If the temperature fluctuations can still be a priori
associated with the turbulence of the flow, that always exists [8], and may occur regardless of the existence
of the neutron (see [9] and the references therein), then for the occurrence of oscillations in the density of the
neutron flux there should be a nonlinearity, which today is tied to a nonlinear relationship with the parame-
ters of the coolant-moderator. And, consequently, between their changes in time there should be connection,
that is observed only in long time oscillations.
An explanation of why there is no link between the oscillations in the density of the neutron flux and
the temperature (Fig. 4 see cover page 3) can be the finite sensitivity of the characteristics of this process: in
order for the connection manifests, there should be sufficiently large changes in the state of the coolant,
which are estimated in [2] as more than 0.5 °С, which is not so in our experiment. But then, that is in the
absence of a correlation between the behavior of neutrons and the temperature of the coolant, it is necessary
to look for other causes of the oscillation in the neutron field. For example it can be assumed, that in reality
there is a low amplitude chaotic regime, which is normal even at the regular mode of operation of the reactor,
or purely neutron effects like xenon oscillations, but with a much smaller period and amplitude.
Conclusions
1. Using wavelet analysis, it is shown that the fluctuation component of signals from temperature and
neutron fluxes standard sensors, at the regular mode of operation of the PWR-1000, contains a variety of
information about the dynamics of these parameters in time.
2. It is shown that the various patterns found in the analyzed signals are a reflection of the dynamics
of the corresponding parameters, but not the random noise.
3. It was found that changes in the temperature of the cold branch of the first contour, at a magnitude
of these changes of about 0.1 °C, correlate with changes in the density of the neutron flux.
4. It was revealed also numerous irregular periodicities in the signals of both temperature and neu-
tron sensors, which do not correlate with each other.
5. It was shown that the sensitivity of the applied analysis method for detecting changes in the sig-
nals is much higher than the simple estimate of the changes in the magnitude of the dispersion of the signal
fluctuation component. This allows us to offer a wavelet analysis of parallel measurements of the tempera-
ture and the neutron flux density to control the appearance of boiling of the coolant.
REFERENCES
1. Pavlovich V. M. Physics of nuclear reactors / V. M. Pavlovich. – Chornobyl, 2009. – 224 p. (Ukr)
2. Аnalysis of noises of neutron flux, induced by fluctuations of a coolant parameters in a PWR active zone / Yu. M.
Semenchenkov, V. A. Мilto, A. A. Pinegin, B. E. Schumskii //Аtomnaya energiya. – 2007. – Vol. 103. – Iss. 5. –
P. 283 - 286. (Rus)
3. Моdernization of the method of diagnostics of a coolant in a PWR active zone / Yu. M. Semenchenkov, V. A.
Мilto, B. E. Schumskii, D. V. Vorobiova // Аtomnaya energiya. – 2015. – Vol. 118. – Iss. 1. – P. 6 - 10. (Rus)
A. D. SKORBUN, S. M. STADNYK, V. G. KOTELENETS ET AL.
________________________________________________________________________________________________________________________
ISSN 1813-3584 ПРОБЛЕМИ БЕЗПЕКИ АТОМНИХ ЕЛЕКТРОСТАНЦІЙ І ЧОРНОБИЛЯ 2017 ВИП. 29 26
4. Regularities in noises of reactor equipment / A. D. Skorbun, S. M. Stadnik, V. G. Kotelenets, A. V. Drobetskii //
Problemy bezpeky atomnykh electrostantsiy i Chornobylya (Problems of Nuclear Power Plants' Safety and of
Chornobyl). – 2017. – Iss. 28. – P. 10 - 16. (Ukr)
5. Аstafieva N. M. Wavelet analysis: the bases of theory and examples of application / N. M. Аstafieva // UFN. -
1996. - Vol. 166, № 11. - P. 1145 – 1170. (Rus)
6. Kumar P. Wavelet analysis for geophysical application / P. Kumar, E. Foufoula-Georgiou // Reviews of geophys-
ics. – 1997. – Vol. 35. – № 4. – P. 385 – 412.
7. Skorbun A. D. Аnalysis of long-time characteristics of wind velocities in the region of Chornobyl NPP and possi-
bilities of forecasting of aerosols emission from “Shelter” object / A. D. Skorbun, B. I. Ogorodnikov //
Мeteorologia i gydrologia. – 2010. – № 8. – P. 49 - 60. (Rus)
8. Thermal physics of a damages of rector plants / O. O. Кliuchnikov, I. G. Scharaevskii, N. M. Fialko, L. B. Zimin,
N. I. Scharaevskaya. – Chornobyl, 2013. – 528 p. (Rus)
9. Postnikov N. S. Investigation of chaotic oscillations in boiling reactors / N. S. Postnikov // Аtomnaya energiya. –
2009. – Vol. 107. – Iss. 5. – P. 243 - 251. (Rus)
А. Д. Скорбун
1
, С. М. Стадник
1
, В. Г. Котеленець
2
, О. Л. Арванінов
2
, А. В. Дробецький
2
1
Інститут проблем безпеки АЕС НАН України, вул. Кірова, 36а, Чорнобиль, 07270, Україна
2
Відокремлений підрозділ «Южно-Українська АЕС» Державного підприємства «НАЕК «Енергоатом»,
Миколаївська обл., Южноукраїнськ, 55000, Україна
ПОРІВНЯННЯ СТАТИСТИЧНИХ ВЛАСТИВОСТЕЙ ТЕМПЕРАТУРИ ОХОЛОДЖУВАЧА
І НЕЙТРОННОГО ПОТОКУ В РЕАКТОРІ ВВЕР-1000
Із використанням вейвлет-аналізу досліджено наявність закономірностей у «шумових» сигналах детек-
торів нейтронного потоку і температури теплоносія штатної системи внутрішньореакторного контролю реакто-
ра ВВЕР-1000. Виконано порівняльний аналіз довгих рядів вимірювань температури теплоносія і щільності
потоку нейтронів. В обох параметрах виявлено багатий спектр різноманітних періодичностей. Деякі з них знай-
дено як у даних температури, так і потоку нейтронів. Проте в динаміці потоку нейтронів і температури виявле-
но також періодичності, які не корелюють між собою. Показано відмінності між картинами вейвлет-
перетворення температур гарячої і холодної ниток: у температурі гарячої нитки виникають додаткові більш
високочастотні коливання, які можна приписати розвитку турбулентності. Зроблено висновок, що проаналізо-
вані флуктуації сигналів не є випадковим шумом, а є носіями інформації про стан нейтронної підсистеми реак-
тора. Пропонується виконувати аналіз флуктуацій сигналів штатної системи контролю для діагностики режимів
роботи реактора.
Ключові слова: ядерний реактор, нейтрони, теплоносій, статистика.
А. Д. Скорбун
1
, С. Н. Стаднык
1
, В. Г. Котеленец
2
, А. Л. Арванинов
2
, А. В. Дробецкий
2
1
Институт проблем безопасности АЭС НАН Украины, ул. Кирова, 36а, Чернобиль, 07270, Украина
2
Обособленное подразделение «Южно-Украинская АЭС» Государственного предприятия
«НАЭК «Энергоатом», Николаевская обл., Южноукраинск, 55000, Украина
СРАВНЕНИЕ СТАТИСТИЧЕСКИХ СВОЙСТВ ТЕМПЕРАТУРЫ ОХЛАДИТЕЛЯ
И НЕЙТРОННОГО ПОТОКА В РЕАКТОРЕ ВВЭР-1000
С использованием вейвлет-анализа изучено существование закономерностей в «шумовых» сигналах
детекторов потока нейтронов и температуры теплоносителя штатной системы внутриреакторного контроля
реактора ВВЭР-1000. Выполнен сравнительный анализ длинных рядов измерений температуры теплоносителя
и плотности потока нейтронов. В обоих параметрах обнаружен богатый спектр разнообразных периодичностей.
Некоторые из них найдены как в температурных данных, так и данных потока нейтронов. Однако в динамике
потока нейтронов и температуры обнаружены также периодичности, которые не коррелируют между собой.
Показаны отличия между картинами вейвлет-преобразования температур горячей и холодной ниток: в темпера-
туре горячей нитки возникают дополнительные более высокочастотные колебания, которые можно приписать
развитию турбулентности. Сделан вывод, что проанализированные флуктуации сигналов не являются случай-
ным шумом, а являются носителями информации о состоянии нейтронной подсистемы реактора. Предлагается
выполнять анализ флуктуаций сигналов штатной системы внутриреакторного контроля для диагностики режи-
мов работы реактора.
Ключевые слова: ядерный реактор, нейтроны, теплоноситель, статистика.
Надійшла 19.06.2017
Received 19.06.2017
|