Predictive maintenance in nuclear power plants through online monitoring
The nuclear power industry is working to reduce generation costs by adopting conditionbased maintenance strategies and automating testing activities. These developments have stimulated great interest in online monitoring (OLM) technologies and new diagnostic and prognostic methods to anticipate,...
Gespeichert in:
| Datum: | 2013 |
|---|---|
| 1. Verfasser: | |
| Format: | Artikel |
| Sprache: | English |
| Veröffentlicht: |
Державне підприємство "Державний науково-технічний центр з ядерної та радіаційної безпеки" Держатомрегулювання України та НАН України
2013
|
| Schriftenreihe: | Ядерна та радіаційна безпека |
| Online Zugang: | http://dspace.nbuv.gov.ua/handle/123456789/97492 |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| Назва журналу: | Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| Zitieren: | Predictive maintenance in nuclear power plants through online monitoring / H.M. Hashemian // Ядерна та радіаційна безпека. — 2013. — № 4. — С. 42-50. — Бібліогр.: 10 назв. — англ. |
Institution
Digital Library of Periodicals of National Academy of Sciences of Ukraine| id |
irk-123456789-97492 |
|---|---|
| record_format |
dspace |
| spelling |
irk-123456789-974922016-03-29T03:02:24Z Predictive maintenance in nuclear power plants through online monitoring Hashemian, H.M. The nuclear power industry is working to reduce generation costs by adopting conditionbased maintenance strategies and automating testing activities. These developments have stimulated great interest in online monitoring (OLM) technologies and new diagnostic and prognostic methods to anticipate, identify, and resolve equipment and process problems and ensure plant safety, efficiency, and immunity to accidents. This paper provides examples of these technologies with particular emphasis on a number of key OLM applications: detecting sensingline blockages, testing the response time of pressure transmitters, monitoring the calibration of pressure transmitters online, cross-calibrating temperature sensors in situ, assessing equipment condition, and performing predictive maintenance of reactor internals. У сфері атомної енергетики проводяться роботи зі зниження витрат на виробництво електроенергії шляхом прийняття стратегій технічного обслуговування за поточним станом обладнання та автоматизації випробувань. Ці роботи викликали великий інтерес до технологій оперативного контролю та нових методів діагностування та прогнозування для виявлення і вирішення проблем, пов’язаних з устаткуванням і технологічними процесами, а також забезпечення безпеки, ефективності та стійкості до аварій. У статті наводяться приклади технологій, заснованих на ряді ключових напрямків застосування оперативного контролю: виявлення блокування вимірювальних ліній, тестування часу відклику датчиків тиску, контроль калібрування перетворювачів тиску, взаємне калібрування температурних датчиків на місці, оцінка стану обладнання та виконання профілактичного обслуговування внутрішньокорпусних пристроїв реактора. В атомной энергетике проводятся работы по снижению затрат на производство электроэнергии путем принятия стратегий технического обслуживания по текущему состоянию оборудования и автоматизации испытаний и опробований. Эти работы вызвали большой интерес к технологиям оперативного контроля и новым методам диагностирования и прогнозирования для выявления и решения проблем, связанных с оборудованием и технологическими процессами, а также обеспечения безопасности, эффективности и устойчивости к авариям. В этой статье приводятся примеры технологий, основанных на ряде ключевых направлений применения оперативного контроля: обнаружение блокировки измерительных линий, тестирование времени отклика датчиков давления, контроль калибровки преобразователей давления, взаимная калибровка температурных датчиков на месте, оценка состояния оборудования и выполнение профилактического обслуживания внутрикорпусных устройств реактора. 2013 Article Predictive maintenance in nuclear power plants through online monitoring / H.M. Hashemian // Ядерна та радіаційна безпека. — 2013. — № 4. — С. 42-50. — Бібліогр.: 10 назв. — англ. 2073-6231 http://dspace.nbuv.gov.ua/handle/123456789/97492 621.039.058 en Ядерна та радіаційна безпека Державне підприємство "Державний науково-технічний центр з ядерної та радіаційної безпеки" Держатомрегулювання України та НАН України |
| institution |
Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| collection |
DSpace DC |
| language |
English |
| description |
The nuclear power industry is working to reduce generation costs by adopting conditionbased
maintenance strategies and automating testing activities. These developments have
stimulated great interest in online monitoring (OLM) technologies and new diagnostic and
prognostic methods to anticipate, identify, and resolve equipment and process problems and
ensure plant safety, efficiency, and immunity to accidents. This paper provides examples of these
technologies with particular emphasis on a number of key OLM applications: detecting sensingline
blockages, testing the response time of pressure transmitters, monitoring the calibration of
pressure transmitters online, cross-calibrating temperature sensors in situ, assessing equipment
condition, and performing predictive maintenance of reactor internals. |
| format |
Article |
| author |
Hashemian, H.M. |
| spellingShingle |
Hashemian, H.M. Predictive maintenance in nuclear power plants through online monitoring Ядерна та радіаційна безпека |
| author_facet |
Hashemian, H.M. |
| author_sort |
Hashemian, H.M. |
| title |
Predictive maintenance in nuclear power plants through online monitoring |
| title_short |
Predictive maintenance in nuclear power plants through online monitoring |
| title_full |
Predictive maintenance in nuclear power plants through online monitoring |
| title_fullStr |
Predictive maintenance in nuclear power plants through online monitoring |
| title_full_unstemmed |
Predictive maintenance in nuclear power plants through online monitoring |
| title_sort |
predictive maintenance in nuclear power plants through online monitoring |
| publisher |
Державне підприємство "Державний науково-технічний центр з ядерної та радіаційної безпеки" Держатомрегулювання України та НАН України |
| publishDate |
2013 |
| url |
http://dspace.nbuv.gov.ua/handle/123456789/97492 |
| citation_txt |
Predictive maintenance in nuclear power plants through online monitoring / H.M. Hashemian // Ядерна та радіаційна безпека. — 2013. — № 4. — С. 42-50. — Бібліогр.: 10 назв. — англ. |
| series |
Ядерна та радіаційна безпека |
| work_keys_str_mv |
AT hashemianhm predictivemaintenanceinnuclearpowerplantsthroughonlinemonitoring |
| first_indexed |
2025-07-07T05:04:52Z |
| last_indexed |
2025-07-07T05:04:52Z |
| _version_ |
1836963273495281664 |
| fulltext |
42 ISSN 2073-6237. Ядерна та радіаційна безпека 4(60).2013
UDC 621.039.058
H. M. Hashemian, Ph.D.
AMS Corporation, AMS Technology Center, 9119 Cross Park
Drive, Knoxville, Tennessee 37923 USA
Predictive maintenance
in nuclear power plants
through online monitoring
The nuclear power industry is working to reduce generation costs by
adopting condition-based maintenance strategies and automating test-
ing activities. These developments have stimulated great interest in on-
line monitoring (OLM) technologies and new diagnostic and prognostic
methods to anticipate, identify, and resolve equipment and process prob-
lems and ensure plant safety, efficiency, and immunity to accidents. This
paper provides examples of these technologies with particular emphasis
on a number of key OLM applications: detecting sensing-line blockages,
testing the response time of pressure transmitters, monitoring the calibra-
tion of pressure transmitters online, cross-calibrating temperature sensors
in situ, assessing equipment condition, and performing predictive mainte-
nance of reactor internals.
K e y w o r d s: Nuclear power plants; Noise analysis; Sensor response
time testing; Sensing-line blockages; Calibration monitoring; Reactor diag-
nostics.
Х. М. Хашеміан
Прогнозне технічне обслуговування АЕС із застосу-
ванням оперативного контролю
У сфері атомної енергетики проводяться роботи зі зниження витрат
на виробництво електроенергії шляхом прийняття стратегій технічного
обслуговування за поточним станом обладнання та автоматизації ви-
пробувань. Ці роботи викликали великий інтерес до технологій опера-
тивного контролю та нових методів діагностування та прогнозування
для виявлення і вирішення проблем, пов’язаних з устаткуванням і тех-
нологічними процесами, а також забезпечення безпеки, ефективності
та стійкості до аварій. У статті наводяться приклади технологій, засно-
ваних на ряді ключових напрямків застосування оперативного контро-
лю: виявлення блокування вимірювальних ліній, тестування часу від-
клику датчиків тиску, контроль калібрування перетворювачів тиску,
взаємне калібрування температурних датчиків на місці, оцінка стану
обладнання та виконання профілактичного обслуговування внутріш-
ньокорпусних пристроїв реактора.
К л ю ч о в і с л о в а: атомні електростанції, аналіз перешкод, тесту-
вання часу відклику датчиків, блокування вимірювальних ліній, контр-
оль калібрування, діагностика реактора.
© H. M. Hashemian, 2013
T
he term online monitoring (OLM) describes methods,
such as the noise analysis technique, for evaluating
the health and reliability of nuclear plant sensors,
processes, and equipment from data acquired while
the plant is operating. Although OLM technologies
typically apply to all types of nuclear power reactors, this paper
uses pressurized water reactors (PWRs) as the reference plant
since they are the type most commonly used in the Western
hemisphere. To control the PWR plant and protect its safety,
several different kinds of sensors are employed to measure
the process parameters (see Table 1). Figure 1 shows a simplified
schematic of the primary loop of a PWR plant and its important
sensors. Depending on the plant design and manufacturer,
a PWR plant has 2 to 4 primary coolant loops; however,
Russian PWRs (called VVERs or WWERs) have up to 6 loops.
The normal output of these sensors can be used both to establish
the health and condition of the plant and sometimes to verify
the performance of the sensors themselves.
Table 1. Typical Population of Important
Sensors in Pressurized Water Plants
Sensor Measurement
Typical
Population
in a Reactor
RTDs (a)
CETs (b)
Pressure
Transmitters (c)
Neutron
Detectors (d)
Temperature
Temperature
Pressure, Level, and Flow
Neutron Flux
16 e 60
50 e 100
500 e 2500
10 e 20
(a) Resistance Temperature Detectors
(b) Core-Exit Thermocouples
(c) Including Differential-Pressure Transmitters
(d) Ex-core and Some In-core Neutron Detectors
Figure 2 shows the output of a process sensor as a function
of time during plant operation. Normally, while the plant is
operating, the output of the sensor will have a steady-state
value that corresponds to the process parameter indicated by
the sensor. This steady-state value is often referred to as the
static component or DC value. Figure 2 also shows a magnified
portion of the sensor’s output signal to illustrate that, in addition
to the static component, a small fluctuating signal is naturally
present on the sensor output. The fluctuating signal, which
is known as the signal’s dynamic or AC component, derives
from inherent fluctuations in the process parameter as a result
of turbulence, random flux, random heat transfer, vibration,
and other effects. It has long been known that the condition
of a nuclear power plant can be effectively monitored by
analyzing these small fluctuations in the process variables, such
as reactivity coefficients, vibration amplitudes, and response
times, around their stationary value. This technique, commonly
known as noise analysis, noise diagnostics, or reactor diagnostics,
makes it possible to discover the abnormal state of the system,
which registers either as a shift of these parameters into non-
permitted regions or the appearance of a changed structure of the
noise signatures, usually the frequency spectra. The advantage
of the noise analysis technique is that it non-intrusively measures
process variables under normal operation without requiring any
external perturbation.
One idiosyncrasy of the noise analysis technique is that
a change in the measured signal characteristics may be caused
either by a change in the transfer properties of the system or
by a change in the driving force; that is, the fluctuation of the
parameter that induces the measured noise. Hence, performing
ISSN 2073-6237. Ядерна та радіаційна безпека 4(60).2013 43
Predictive maintenance in nuclear power plants through online monitoring
a proper diagnosis requires sufficient expert knowledge to choose
the appropriate model on which the diagnostic algorithm is
based. The noise analysis technique also involves an additional
ambiguity: deteriorating sensor characteristics can change
the measured noise signature. In the Three Mile Island accident
in 1979, for example, a role in the accident sequence was played
by a failed sensor and the control room personnel’s inability
to realize its failure. Sensor malfunction, or just de-calibration,
can also occur under much less dramatic circumstances, through
fouling, drift, response time degradation, and aging. As it turns
out, noise analysis can be used even for sensor health analysis,
by differentiating between sensor degradation/failure and system
malfunction/anomaly.
Because the static (DC) and dynamic (AC) components
of the sensor output each contain different information about
the process being measured, they can be used for a wide
range of monitoring applications. For example, applications
that monitor for gradual changes in the process over the fuel
cycle, such as sensor calibration monitoring, make use of the
static component. In contrast, applications that monitor fast-
changing events, such as core barrel motion, use the information
in the dynamic component, which provides signal bandwidth
information. Figure 3 illustrates how existing data from
process sensors is used for these applications. Note that in this
figure, the static data is analyzed using empirical and physical
modeling and averaging techniques involving multiple signals,
while dynamic data analysis entails time domain and frequency
domain analysis based on single signals or pairs of signals. For
example, the dynamic response time of a nuclear plant pressure
transmitter is identified by Fast Fourier Transform (FFT) of the
noise signal. The FFT yields the auto power spectral density
(APSD) of the noise data from which the transmitter response
time is calculated. In applications where pairs of signals are
used (e.g., core barrel vibration measurements), the cross power
spectral density (CPSD), the phase, and the coherence data are
calculated to distinguish the vibration characteristics of various
constituents of the reactor internal.
The types of OLM applications used in nuclear power
plants are in large part determined by the sampling rates
available for data acquisition. Static OLM applications, such
as resistance temperature detector (RTD) cross-calibration
and online calibration monitoring of pressure transmitters,
typically require sampling rates up to 1 Hz, while dynamic
OLM applications such as sensor response time testing use
data sampled in the 1 kHz range. Other, high-frequency
OLM applications, such as measuring the vibration of rotating
equipment and monitoring loose parts, may use data sampled at
up to 100 kHz. Figure 4 shows examples of OLM applications
Figure 1. Primary Loop
of a Pressurized Water
Reactor (PWR)
Figure 2. Normal Output of a Process Sensor with
Illustration of the DC and AC Components of the Output
44 ISSN 2073-6237. Ядерна та радіаційна безпека 4(60).2013
H. M. Hashemian, Ph.D.
that can be used in nuclear power plants, with their range
of data sampling frequency. For low-frequency (DC) data
analysis, averaging techniques are typically used for redundant
sensors, and empirical and physical modeling techniques are
used for non-redundant sensors.
Because I&C sensors that measure temperature, pressure, level,
flow, and neutron flux up to data sampling frequencies of around
1 kHz represent the majority of measurement devices in nuclear
power plants, focusing this paper on the OLM applications that
monitor these sensors will show to best advantage the potential
benefits of OLM for nuclear plants. Other OLM applications,
such as measuring the vibration of rotating equipment and
monitoring loose parts, which primarily rely on high-frequency
acquisition of data from accelerometers, are not discussed in this
paper because they don’t acquire data from the existing process
sensors of the plant.
OLM APPLICATIONS IN NUCLEAR POWER PLANTS
The success of the noise analysis technique in nuclear power
plant applications stimulated the industry to examine the feasibility
of implementing an online monitoring (OLM) system that
incorporates the noise analysis technique in both the current and
next generation of nuclear reactors for the purpose of dynamically
testing sensors, measuring the vibration of reactor internals, and
performing a variety of diagnostic applications. This OLM system
will also give plants the capability to verify the calibration of pressure,
level, and flow transmitters as well as RTDs and thermocouples.
The system will have built-in signal validation, noise analysis, and
OLM algorithms that will enable nuclear power plants to check for:
(1) calibration and response time of process instruments; (2) identify
sensing-line blockages; (3) monitor the reactor coolant flow, and (4)
alert the reactor operator of excessive vibration of reactor internals.
Figure 4. Online Monitoring
Applications Versus
Sampling Frequency
Figure 3. Online Monitoring
Application of Static and Dynamic
Data Analysis in this Paper
ISSN 2073-6237. Ядерна та радіаційна безпека 4(60).2013 45
Predictive maintenance in nuclear power plants through online monitoring
Such an OLM system can provide plants with the information
they need to evaluate I&C sensors by providing applications that
identify drifting instruments, alert plant personnel of unusual
process conditions, and predict impending failures of plant
equipment. Moreover, operating nuclear power plants can use
OLM technologies to improve their efficiency. For example,
nuclear power plants are required to calibrate important I&C
instruments once every fuel cycle. This requirement dates back
40 years to when commercial nuclear power plants were first
licensed to begin operation. Based on calibration data accumulated
over these four decades, it has been determined that the calibration
of some instruments, such as pressure transmitters, does not drift
enough to warrant calibrating all transmitters as often as once
every fuel cycle. OLM allows calibration efforts to be focused on
the instruments that have drifted out of tolerance, thereby saving
plants a significant amount of the time and manpower.
Online Detection of Sensing-line Blockages
Chief among applications of noise analysis in nuclear power
plants is detecting sensing- line blockages. Sensing-lines (also called
impulse lines) are small diameter tubes that bring the pressure
signal from the process to the pressure sensor. Depending on the
application and the type of plant, pressure sensing-lines can be
as long as 300 meters or as short as 10 meters. The isolation valves,
root valves, and bends along their length make them susceptible
to blockages from residues in the reactor coolant, failure of the
isolation valves, and other problems. Sensing- line blockages are
a recurring problem in PWRs, boiling water reactors (BWRs),
and essentially all water-cooled nuclear power plants. They are
an inherent phenomenon that causes the sensing-lines of nuclear
plant pressure transmitters to clog up with sludge, boron,
magnetite, and other contaminants. Typically, nuclear plants
purge the important sensing-lines with nitrogen or backfill the lines
periodically to clear any blockages. This procedure is, of course,
time consuming and radiation intensive, and more importantly,
not always effective in eliminating blockages. Furthermore, with
the exception of noise analysis, no way exists to know ahead
of time which sensing-lines may be blocked. Also, unless the noise
analysis technique is used, it’s not possible after purging or back
filling a sensing-line to verify that the line has been cleared.
Figure 5 shows the cutaway of a partially blocked sensing-
line of a nuclear power plant pressure transmitter. It’s clear from
the figure that this blockage can reduce the flow path in this
sensing-line by about 40 percent. A blockage like this hampers
the dynamic response of the pressure sensor at the end of the
sensing-line. In particular, depending on the design characteristics
of the pressure transmitter, a sensing-line blockage like this can
cause the response time of the affected pressure transmitter
to increase by an order of magnitude. The degree of increase
in the dynamic response depends on the “compliance” of the
pressure transmitter. Compliance is a pressure transmitter
design parameter that relates to the physical displacement of the
sensing element of the transmitter per unit of input pressure.
Some transmitters, such as those with sensing elements made
of “bellows,” have a large compliance and are therefore
affected strongly by sensing-line blockages. On the other hand,
transmitters with sensing elements made of stiff diaphragms have
smaller compliances and are therefore less affected by sensing-
line blockages.
The effect of compliance on the dynamic response of a pressure
transmitter was revealed in a research project performed by
the author for the U.S. Nuclear Regulatory Commission (NRC)
in the early 1990s (Hashemian, 1993). The goal of the project was
to characterize the effects of normal aging on the performance
of nuclear plant pressure transmitters by illustrating the effect
of compliance on the response time of representative nuclear-
grade pressure transmitters from three manufacturers: Barton,
Foxboro, and Rosemount (see Figure 6). The data in Figure 6
is from laboratory tests measuring the response time of the
transmitters using a pressure ramp signal.
A significantly blocked sensing-line can render the pressure
sensor essentially useless or even dangerous. The danger here
is that, due to a total blockage, the operating pressure may get
locked in the transmitter and cause its indication to appear
normal. Then, when the pressure changes, the transmitter will
not respond and will continue to show the locked-in pressure,
which will confuse the reactor operators and potentially pose
a risk to the safety of the plant.
If a blocked pressure transmitter happens to be a part
of a redundant safety channel, it can trip the plant during
a transient. More specifically, the indication of a blocked
transmitter will obviously not match the other redundant
channels, creating a mismatch that could trigger a reactor trip.
Figure 5. Photograph of a Nuclear Plant
Sensing-line with a Partial Blockage
Figure 6. Research Results on the Effect
of Sensing-line Blockages onResponse Time
of Nuclear Plant Pressure Transmitters
46 ISSN 2073-6237. Ядерна та радіаційна безпека 4(60).2013
H. M. Hashemian, Ph.D.
In fact, this problem has occurred in France where partial
blockages in flow transmitters caused two French PWRs to trip
during load flowing episodes in the mid-1980s (Meuwisse and
Puyal, 1987).
Some sensing-line blockages are so severe that the sensing-
line has to be drilled to clear the blockage. This type of problem
is the reason why measuring the response time of pressure
transmitters is so important and why it is so surprising that even
today, some nuclear power plants measure the response time
of their safety-related pressure transmitters using conventional
procedures that exclude the sensing-lines. These plants typically
use a hydraulic pressure generator to input a pressure signal to the
transmitter and measure its response time. In doing this, the sensor
is isolated from the sensing-lines. The research work documented
in the U.S. Nuclear Regulatory Commission report NUREG/
CR-5851 uncovered this flaw in the maintenance of nuclear plant
pressure transmitters. As a result, many plants have recognized
that they must measure the response time of both their pressure
transmitters and their sensing-lines. These plants have accordingly
switched to the noise analysis procedure to verify the dynamic
characteristics of their pressure sensing systems.
Response Time Testing of Pressure Transmitters
Pressure, level, and flow transmitters in nuclear power plants
behave like filters on the natural plant fluctuations that are
presented to their inputs. That is, if one assumes that the input
to the transmitter exhibits wide-band frequency characteristics
(which is typically the case for nuclear power plant fluctuations),
information about the sensor itself can be inferred by measuring
the transmitter output. This is the basis of the noise analysis
technique that is used to determine the dynamic response
of pressure, level, and flow transmitters in nuclear power plants
(Thie, 1981).
The noise analysis technique is used to remotely measure
sensor response time from the control room area while the plant
is online. These measurements do not require the sensors to be
disconnected from the plant instrumentation or removed from
service for the tests. That is, the tests are passive and do not
cause any disturbance to plant operation. This reduces test time
and helps to reduce radiation exposure of the test personnel who
would otherwise have to enter the reactor containment to make
the response time measurements.
Dynamic response analysis is performed in the frequency
domain and/or time domain, and is based on the assumption
that the dynamic characteristics of the transmitters are linear
and that the input noise signal (i.e., the process fluctuations)
has proper spectral characteristics. Frequency domain and time
domain analyses are two different methods for determining
the response time of transmitters. It is usually helpful to analyze
the data with both methods and average the results, excluding
any outliers.
In frequency domain analysis, the APSD of the signal
is generated first, usually using an FFT algorithm. After
the APSD is obtained, a mathematical function (model) that
is appropriate for the transmitter under test is fit to the APSD
to yield the model parameters. These parameters are then used
to calculate the dynamic response of the transmitter. The dynamic
response of the transmitter can then be analyzed to determine
the response time of the transmitter in situ. Figure 7 shows
an example of process noise that enters a pressure transmitter
and is subsequently filtered by the transmitter. The response
time of the transmitter can be inferred from the APSD with
the proper analysis tools.
Under normal plant conditions, the APSDs of nuclear plant
pressure transmitters have characteristic shapes that can be
baselined and compared with the APSDs of similar transmitters
operating under the same process conditions. Figure 8 shows
examples of a few typical nuclear plant APSDs for steam
generator level, reactor water clean-up flow, and pressurizer
pressure transmitters.
Through laboratory experiments, the noise analysis
technique was validated for in-situ response time testing
of pressure transmitters. This validation work involved directly
measuring response time and then using the noise analysis
technique to compare the ramp input signals with the response
time results. Table 2 shows representative results of this
validation work for seven different transmitters from various
manufacturers of nuclear-grade pressure transmitters. For each
transmitter, the results of the direct measurement of response
time (ramp test) were compared with the results of the noise
analysis test; the difference between the two results is shown
in Table 2. The details of the validation of the noise analysis
technique for response time testing of nuclear plant pressure
transmitters are documented in a comprehensive report, “Long
Term Performance and Aging Characteristics of Nuclear Plant
Pressure Transmitters,” published by the NRC in March 1993
as NUREG/CR-5851 (Hashemian, 1993).
Online Calibration Monitoring of Pressure Transmitters
Online calibration monitoring refers to monitoring the normal
output of nuclear plant pressure transmitters during plant
operation and then comparing this data with an estimate of the
process parameter that the transmitter is measuring. At most
plants, the plant computer contains all the data with an estimate
of the process parameter that the transmitter is measuring.
At most plants, the plant computer contains all the data that
is needed to verify the calibration of pressure transmitters,
including data from plant startup and shutdown periods used
to verify the calibration of instruments over their operating
range. Using the online calibration method, transmitter outputs
are monitored during process operation to identify drift. If drift
is identified and is significant, the transmitter is scheduled for
a calibration during an ensuing outage. On the other hand, if
the transmitter drift is insignificant, no calibration is performed
for typically as long as eight years. This eight year period is
based on a two year operating cycle and a redundancy level
Figure 7. Example of a Pressure
Transmitter Filtering
the Process Noise
ISSN 2073-6237. Ядерна та радіаційна безпека 4(60).2013 47
Predictive maintenance in nuclear power plants through online monitoring
of four transmitters. In this application, OLM is not a substitute
for traditional calibration of pressure transmitters; rather,
it is a means for determining when to schedule a traditional
calibration for a pressure transmitter.
Reviews of the calibration histories of process instruments
in nuclear power plants have shown that high-quality instruments,
such as nuclear-grade pressure transmitters, typically maintain
their calibration for more than a fuel cycle of about two years
and do not, therefore, need to be calibrated as often (Hashemian,
1995, Hashemian 1998). The validity of the OLM approach
in verifying the calibration of nuclear plant pressure transmitters
was researched in the mid-1990s using both theoretical work
and laboratory experiments and in-plant trials. The results
are documented in a comprehensive NRC report published
in November 1995 as NUREG/CR-6343: “Online Testing
of Calibration of Process Instrumentation Channels in Nuclear
Power Plants” (Hashemian, 1995).
To perform online calibration monitoring, the output
of redundant sensors is averaged. The average value, called
the process estimate, is then used as a reference to determine
the deviation of each sensor from the average of the redundant
sensors and to identify the outliers. For non-redundant sensors,
averaging obviously cannot be used to determine a reference
value. Therefore, if there is insufficient instrument redundancy,
the process estimate for calibration monitoring is determined by
analytical modeling of the process. Both empirical and physical
modeling techniques are used in this application, although
empirical models are preferred because they can be adapted
to various processes and different operational envelopes.
In particular, empirical modeling techniques involving
neural networks have been researched for online calibration
monitoring applications in nuclear power plants (as documented
in NUREG/CR-6343). Although neural networks are effective,
the nuclear industry does not currently favor this application
Figure 8. Examples of Auto Power Spectral Densities
of Nuclear Plant Pressure Transmitters
Table 2. Representative Results of Laboratory Validation
of Noise Analysis Technique for Response Time
Testing of Nuclear-grade Pressure Transmitters
Number
Response Time (s)
Ramp Test Noise Analysis Difference
Barton
1 0.05 0.09 0.04
2 0.17 0.20 0.03
3 0.17 0.25 0.08
4 0.12 0.15 0.03
5 0.12 0.20 0.08
6 0.11 0.15 0.04
7 0.12 0.18 0.06
Foxboro
1 0.13 0.16 0.03
2 0.21 0.18 -0.03
3 0.16 0.13 -0.03
4 0.09 0.12 0.03
5 0.29 0.30 0.01
6 0.25 0.15 -0.10
7 0.28 0.25 -0.03
Rosemount
1 0.05 0.06 0.01
2 0.32 0.28 -0.04
3 0.07 0.05 -0.02
4 0.10 0.07 -0.03
5 0.11 0.08 -0.03
6 0.09 0.08 -0.01
7 0.09 0.09 0.00
Other manufacturers
1 0.15 0.15 0.00
2 0.21 0.18 -0.03
3 0.02 0.08 0.06
4 0.03 0.07 0.04
5 0.08 0.11 0.03
6 0.15 0.27 0.12
7 0.33 0.37 0.04
48 ISSN 2073-6237. Ядерна та радіаційна безпека 4(60).2013
H. M. Hashemian, Ph.D.
because of difficulties in determining the uncertainty of their
results. As such, other methods, such as averaging or analytical
modeling techniques, have been developed for monitoring
the calibration of pressure transmitters.
In-situ Cross Calibration of Temperature Sensors
PWR plants often employ 20 to 40 RTDs to measure the fluid
temperature in the reactor coolant system. The temperatures
measured by the RTDs are used by the plant operators for process
control and to assess the operational status and safety of the plant.
As such, the calibration of the RTDs is normally evaluated at least
once every refueling cycle. Each RTD must meet specific accuracy
requirements in order for the plant to continue to produce power
according to its design specifications. There are also about 50 core-
exit thermocouples (CETs) in PWRs to provide an additional way
to monitor reactor coolant temperature. Accuracy for CETs is
not as important as for RTDs because CETs are used mostly for
temperature monitoring. Nevertheless, CETs are sometimes cross
calibrated against RTDs to ensure that their output is reliable.
In each loop of a PWR plant and for each core quadrant,
redundant RTDs and CETs are used to minimize the probability
of failure of any one RTD or CET affecting the safety of the
plant. This redundancy of temperature sensors is the basis for
a method of evaluating the calibration of RTDs and CETs called
cross calibration. In cross calibration, redundant temperature
measurements are averaged to produce an estimate of the true
process temperature. The result of the averaging is referred
to as the process estimate. The measurements of each individual
RTD and CET are then subtracted from the process estimate
to produce the cross- calibration results in terms of the deviation
of each RTD from the average of all redundant RTDs (less any
outliers). If the deviations from the process estimate of an RTD
or CET are within acceptable limits, the sensor is considered
in calibration. However, if the deviation exceeds the acceptance
limits, the sensor is considered out of calibration and its use for
plant operation is evaluated.
Traditionally, cross-calibration data has been acquired using
data acquisition equipment that is temporarily connected to test
points in the plant instrumentation cabinets. The traditional
cross-calibration method, while highly accurate, causes the plant
to lose indication when the data is being acquired, and costs
the plant time during shutdown and/or startup to restore
the temperature indications. Now, with new and more advanced
plant computers, RTD and CET measurements can be collected
in the plant computer, which provides a centralized location for
monitoring and storing the measurements. Using online data
from the plant computer for cross calibration can save plants
startup and shutdown time, while producing results that are
comparable to the traditional method.
Equipment Condition Assessment
Static analysis methods may be used for other purposes besides
evaluating the health of individual sensors as in online cross-
calibration and transmitter calibration monitoring. Equipment
condition assessment (ECA) applications take the idea of online
calibration monitoring a step further by monitoring for abnormal
behavior in a group of sensors. An example of ECA is illustrated
in Figure 9, which shows a simplified diagram of a typical
chemical and volume control system (CVCS) in a PWR.
The primary functions of a typical CVCS in a PWR are:
1. Controlling the volume of primary coolant in the reactor
coolant system (RCS)
2. Controlling chemistry and boron concentration in the RCS
3. Supplying seal water to the reactor coolant pumps (RCPs)
Several transmitters are typically used to monitor various
parameters related to the operation of the CVCS. Figure 9
highlights the normal operation of a few of these parameters:
1. Charging Flow — measures the flow rate of the coolant
being provided from the volume control tank (VCT) to the RCS
and RCP seals
2. Reactor Coolant Pump Seal Injection Flow — measures
the flow rate of the coolant provided to the RCP seals
3. Seal Return Flow — measures the flow rate of the coolant
returned to the VCT from the RCP seal injection
4. Letdown Flow — measures the flow rate of the reactor
coolant as it leaves the RCS and enters the VCT
During normal operation, the measurements of these
parameters will fluctuate slightly, but should remain at
a consistent relative level. However, in abnormal conditions
such as a RCP seal leak, some parameters may exhibit upward or
downward trends, indicating a problem in the plant. For example,
Figure 10 shows the four flow signals mentioned above during
normal operation of a PWR plant (the actual flow rates are
scaled to simplify this example). As Figure 10 shows, the flows
remain at relatively constant rates relative to one another.
Figure 11 shows how these flow signals may appear at the onset
of a RCP seal leak in this PWR plant. In this example, the onset
of the RCP seal leak is first indicated by a downward trend
in the seal return flow measured at time T1. This is followed by
an increase in charging pump flow at time T2 as the charging
pump compensates for the loss of coolant due to the RCP seal leak.
Of course, an abnormal trend in an individual parameter such
as seal return flow could mean that the sensor is degrading; however,
abnormalities in related parameters that occur close together in time
are more likely to indicate the onset of a system or equipment
problem. Early warning of these types of failures is thus the key
benefit of ECA. More specifically, early warning of impending
equipment failures can provide nuclear plants with increased plant
safety through early recognition of equipment failures and reduced
downtime stemming from timely repair of affected equipment.
Predictive Maintenance of Reactor Internals
A research project published in NUREG/CR-5501 (June
1998), “Advanced Instrumentation and Maintenance Technologies
for Nuclear Power Plants,” investigated such OLM applications
as noise analysis for measuring the vibration of reactor internals
and other components such as RCPs (Hashemian, 1998). Typically,
vibration sensors (e.g., accelerometers) are located on the top
and bottom of the reactor vessel to sound an alarm in case
the main components of the reactor system vibrate excessively.
However, neutron detectors have proved to be more sensitive than
accelerometers in measuring the vibration of the reactor vessel and
its internals. This is because the frequency of vibration of reactor
internals is normally below 30 Hz, which is easier to resolve using
neutron detectors than accelerometers. Accelerometers are more
suited for monitoring higher-frequency vibrations.
Figure 12 shows the APSD of the neutron signal from an ex-
core neutron detector (NI-42) in a PWR plant. This APSD
contains the vibration signatures (i.e., amplitude and frequency)
of the reactor components, including the reactor vessel, core
barrel, fuel assemblies, thermal shield, and so on. It even contains,
at 25 Hz, the signature of the RCP rotating at 1,500 revolutions
ISSN 2073-6237. Ядерна та радіаційна безпека 4(60).2013 49
Predictive maintenance in nuclear power plants through online monitoring
per minute, which corresponds to 25 Hz. These signatures can
be trended to identify the onset of aging degradation, which can
damage the reactor internals. The neutron detection approach
has been recognized as a predictive maintenance tool that can
help plants guard against vibration-induced mishaps that may
be encountered as plants age, making them more vulnerable
to challenges to their structural integrity.
Over the last ten years, numerous plants have begun
programs to measure reactor internal vibration using neutron
noise analysis and then trend the results so as to identify changes
and signs of degradation. Table 3 shows average values for
the resonant frequency of vibration of reactor internals of PWR
plants. The resonant frequency of the RCP also shows up on the
neutron noise signal, as shown in Table 3, at 25 Hz corresponding
to 1500 RPM (revolutions per minute).
The results in Table 3 are the average of neutron noise
measurements made by the authors and others in 15 PWR plants
around the world representing ABB, Westinghouse, Babcock
and Wilcox (B&W), Areva (i.e., Framatome and Siemens),
and Mitsubishi Heavy Industries (MHI) plants. The details are
presented in NUREG/CR 5501 (Hashemian, 1998).
Summary and Conclusions
Over the past 40 years, an array of techniques has been
developed for equipment and process condition monitoring.
Because of regulatory constraints, cost of implementation, and
other factors, these techniques have been used in nuclear power
plants mostly on an “as-needed” basis rather than for routine
condition monitoring applications. Now, with the advent of fast
data acquisition technologies and proliferation of computers
and advanced data processing algorithms and software packages,
condition monitoring can be performed routinely and efficiently
using dedicated equipment installed at the plants.
This paper reviewed a class of condition monitoring technologies
that depend on data from existing process sensors during all modes
of plant operation including startup, normal operating periods,
and shutdown conditions. The data may be sampled continuously
or periodically depending on the application. The steady-state
(DC) component of the data is analyzed to identify slowly
developing anomalies such as calibration changes in process
sensors. The fluctuating (AC) component of the data is analyzed
to determine such parameters as the response time of pressure
Figure 9. Simplified Diagram of Chemical and Volume Control System Components
Figure 10. Normal Operation of Chemical
and Volume Control System Flow Parameters
Figure 11. Chemical and Volume Control System Flow
Parameters at the Onset of a Reactor Coolant Pump Seal Leak
Time (Hours)Time
0 2 4 6 8 10 12
50 ISSN 2073-6237. Ядерна та радіаційна безпека 4(60).2013
H. M. Hashemian, Ph.D.
sensors or to measure the vibrational characteristics of reactor
internals, check for blockages within the reactor coolant system,
identify flow anomalies, and provide other diagnostics.
The AC and DC data acquisition and signal processing
techniques described in this paper can be integrated together
to provide an online monitoring (OLM) system for nuclear power
plants. This paper introduced the key applications of this system
together with the requirements for implementing it in nuclear
power plants. Such OLM systems should be built into the design
of the next generation of reactors to contribute to optimized
plant maintenance by providing automated measurements,
condition monitoring, and diagnostics. In fact, an OLM system
is currently under development by the author and his colleagues
at AMS Corporation for the Small Modular Reactors (SMRs)
currently under design and development in the United States.
One such reactor, referred to as mPower, is currently slated
to be built by Babcock and Wilcox (B&W) Company and will
be constructed in Oak Ridge, Tennessee. The SMR plant will
belong to Tennessee Valley Authority (TVA) which is a Federal
Utility in the United States. TVA already owns and operates six
nuclear power plants, and is in the process of completing its
seventh conventional PWR. The SMR plant to be built in Oak
Ridge, Tennessee will consist of two to four modules of about
200 MWe power, to be completed by 2022.
As for the current generation of reactors, they should be
retrofitted with OLM systems as utilities begin to appreciate their
benefits and as regulators realize the added benefits of OLM
to nuclear reactor safety.
References
1. Arndt, S. A., and D.W. Miller, 1991. The use of radiation
detection noise for signal validation and sensor degradation monitoring.
Proceedings of the Symposium on Nuclear Reactor Surveillance and
Diagnostics (Gatlinburg, Tennessee, USA), SMORN VI 2,66.01 —
66.12.
2. U.S. Nuclear Regulatory Commission. Long term performance
and aging characteristics of nuclear plant pressure transmitters.
NUREG/CR-5851, 1993.
3. U.S. Nuclear Regulatory Commission. Online testing
of calibration of process instrumentation channels in nuclear power
plants. NUREG/CR-6343, 1995.
4. U.S. Nuclear Regulatory Commission. Advanced instrumentation
and maintenance technologies for nuclear power plants. NUREG/CR-
5501, 1998.
5. Lillis, D., 2008. Oral presentation at AMS, November 2008,
Knoxville, Tennessee, USA.
6. Meuwisse, C., and C. Puyal, 1987. Surveillance of neutron and
thermodynamic sensors in EDF nuclear power plants. Paper presented
at the CORECH Meeting, Chatou, France.
7. Montanari, G. C., 2008. Envisaging links between fundamental
research in electrical insulation and electrical asset management. IEEE
Electrical Insulation Magazine, 24(6).
8. NRC Project no. 669, July 2000. “Safety Evaluation by the Office
of Nuclear Reactor Regulation: Application of Online Performance
Monitoring to Extend Calibration Intervals of Instrument Channel
Calibrations Required by the Technical Specifications — EPRI Topical
Report (TR) 104965 Online Monitoring of Instrument Channel
Performance.” U.S. Nuclear Regulatory Commission: Washington,
D.C.
9. Nuclear Energy Institute, 2008. Affordable Energy to Support
Economic Growth. NEI, Washington, DC.
10. Thie, J.A., 1981. Power reactor noise. American Nuclear Society,
La Grange Park, Illinois, USA.
Отримано 21.10.2013.
Figure 12. Auto Power Spectral
Density Containing Vibration
Signatures of Reactor Internals
Table 3. Typical Frequencies of Motion of Reactor
Internals at Pressurized Water Reactor Plants
Reactor Component
Average Resonant
Frequency (Hz)
Fuel Assembly 3.0
Core Barrel Beam Mode 9.7
Core Barrel Shell Mode 23.1
Thermal Shield 12.5
Reactor Vessel 18.5
Reactor Coolant Pump 25.0
Source: NUREG/CR-5501
|