Progress in Nuclear Energy. 1985, Vol. 15, pp. 121-134
(X)7t~6530/85 $(I.(XI + .5(
Printed in Great Britain.
REPORT FROM THE I.W.G.F.R. MEETING BOLOGNA, ITALY (OCTOBER 8-10, 1984) M. MOTrA (EDITOR)
E.N.E.A., Bologna, Italy
INTRODUCTION
The Specialists' Meeting on "Methods and Tools to Detect Thermal Noise in Fast Reactors" was held in Bologna on 8-10 october 1984. The meeting was hosted by th~ ENEA and was sponsored by the IAEA on the recommendation of the International Working Group on Fast Reactors. 17 participants attended the meeting from France, the Federal Republic of Germany, Italy, Japan, the United Kingdom, Joint Research Centre of CEC and from IAEA. The meeting was presided over by Prof. Mario Motta of Italy. The purpose of the meeting was to review and discuss methods and tools for temperature noise detection and related analysis as a potential means for detecting local blockages in fuel and blanket subassemblies
and other faults in LMFBR.
The meeting was divided into four technical sessions as follows: i.
National
review
presentations
on application
purposes
and
research
activities
for
thermal noise detection. 2.
Detection
instruments
and electronic
equipment
for temperature
measurements
in fast
reactors. 3. Physical models. 4. Signal processing techniques. During the meeting, papers were presented by the participants on behalf of their countries and organizations (Ref. /1/-/12/). Each presentation was followed by an open discussion on the subject covered in the paper and general discussion was held at the end of each session. Afver the formal technical sessions were completed s final discussion session was held and general conclusions and recommendations were reached by consensus. Session summaries, general conclusions and recommendations of the official participants (see Appendix i) are given below.
121
22
M. MOTTA
SESSION
I
National review presentations on application purposes and research activities for thermal noise detection
Papers presented by experts from France, the Federal Republic of Germany, Italy, Japan and the United Kingdom indicated that the thermal noise detection and analysis methods in LMFBR are being developed mainly for subassembly surveillance under normal and fault conditions. Detection of coolant flow disturbances and blockages, diagnostics of other malfunctions in subassemblies (due to swelling, deformation, etc.) are the major objectives for application of thermal noise detection in fast reactors. Additional application areas for temperature noise studies include estimates of time constants for instrumentation and transfer functions for reactor phenomena (France and Italy), as well as evaluation of temperature fluctuation impact on mechanical structures
(France).
These areas are also recognized to be important (Federal Republic of Germany, the United Kingdom).
other
by
specialists
of
countries
Instrumentation, physical models and signal processing techniques are being developed for the above purposes and are being tested in out-of-pile experimental facilities used in different countries included subassembly mock-ups (both in sodium and water loops) and unbloeked and blocked electrically heated rod bundles (up to 169 pin bundles). In pile performance tests are carried out in experimental reactors (KNK-II in the Federal Republic of Germany, Joyo in Japan, Phenix in France) and planned for prototype and commercial size plants. French Super-Phenix-I
reactor
is equipped with
which will be able to monitor hourly
temperature
the ANABEL
system for
fluctuations
core surveillance
in the whole core through
eight analogue channels. In addition the use of operational thermocouples for approximately 60 subassemblies is under consideration for temperature fluctuation measurements in SNR300 in Germany. Future investigations are expected to be devoted to the improvement of the sensitivity of methods and sensors. Further theoretical work is envisaged to extend any application of temperature noise detection methods to more realistic conditions. It was agreed by the participants, that two important areas which had not yet been adequately studied were (i) the transport of temperature information between the subassembly outlet and the sensor location, and (2) the optimization of signal processing strategies to obtain reliable discrimination with appropriate response. These problems need to be resolved before the use of temperature noise monitoring can be envisaged in future commercial reactors.
SESSION
Detection reactors.
II
instruments
and
electronic
equipment
Each of the following four countries presented fluctuation monitoring at the S/A outlet.
for
a
temperature
particular
measurements
device
f or
in
fast
temperature
FRANCE emphasized the performance of stainless-steel/sodium thermoelectric junction (used and studied in Japan and FRG too). Developments on mean and fluctuating behaviour were presented. It was shown that a twinned probe (associating a type K and a stainless-steel electrode) could be used for transfer function estimation of the type K sensor.
I.W.G.F.R. meeting
123
JAPAN presented the sensor used in the Joyo experimental reactor and also to be placed at the exit of a few tens of S/As on the MONJU power plant,
It consists
flowmeter used for both temperature and flowrate measurements.
of an eddy current
Sensitivity of this device
vs temperature is shown to be as high as for a standard type K thermocouple. A review of temperature time constants (vs diameter and for different technologies) also presented.
was
F.R.G. has developed a multipurpose probe consisting of a pair of permanent magnetic flowmeters and a pair of 3-wire thermocouples. The sensor is now being operated in KNKII. Standard and transit time analysis of the flowrate meter and thermocouple signals enables the simultaneous determination and on-line comparison of the mean and fluctuating components of flow and temperature. The transfer functions between type-K TC's and SS/Na-junctions have been compared using a simple TC model. In addition the influence of the TC itself on the measurements has been estimated by replacing the TC with sodium. Results, presented for a 0.5 mm diameter type K TC, show very limited frequency response, whereas SS/Na-TC indicates undistorted measurement.
U.K.
presented
an
evaluation
of
a clustered
probe consisting
of 6 mineral
insulated,
coaxial Chromel/Alumel thermocouples. An estimate of the response time of such clusters was obtained and the effect of the presence of gas in the sodium on thermoeouple performance was investigated. The time constant of the clustered thermocouples was found to be better than 50 ms and it was noted that the observed electrical background noise levels would permit the detection of attenuated temperature noise components up to 20 Hz for subassembly fault detection.
Tables 1-5 give an overview of temperature core equipment in the different participating countries.
SESSION III
Physical modelling JAPAN presented
a physical
model which
predicted
RMS
values
of temperature
fluctuation
for 37 and 91 pin bundles. The calculation showed that the RMS temperature fluctuation increases when there are many pins in a bundle, and that there is possibility of detecting the occurence of a blockage in power reactors.
F.R.G. (KfK) made a detailed presentation of physical modelling of the mean and fluctuating temperature fields, and investigated the difference between water and sodium coolant based on eddy diffusivity. Good agreement was observed for axial distance X/d 20. The k-value was introduced as the ratio of RMS noise to axial temperature increase and found to be a sensitive indicator of flow disturbance.
24
M. MOTTA
U.K. has developed the computer code STATEN to predict fluctuation profiles based on a Monte-Carlo technique.
the temperature
and temperature
The effect of heat transfer between fluid particles is considered. Verification of the method was obtained experiments in sodium.
by the use of pipe-flow
and divergent
jet flow
SESSION IV
Signal processing techniques During this Session 3 papers were presented concerned with Signal processing techniques. The first paper, in cooperation between ENEA and Politecnico of Milano potential of a new statistical technique called the Bootstrap method. This technique single sample distributions.
exploited
the
allows the determination of the accuracy of measured quantities from a of independent observations without any assumption of the type of
The method has been used for determination of the accuracy of thermocouple time constants but it can be used also for other quantities, /or example blockage detection values as k-value, skewness and kurtosis. It is proposed that mock-up experiments could be used to validate this method. The Monte-Carlo turbulence and heat transfer model STATEN developed at the Berkeley Nuclear Labs. of the Central Electricity Generating Board, has been used to generate test data to enable the resolution of automatic decision making systems to be determined against different levels of background signals due to temperature tilts across the subassembly. Such data has been supplied as input to a decision making algorithm Learning Network (ALN).
based on an Adaptive
The present evaluation of an ALN has indicated that under the current modelling assumptions, it is possible to automatically identify the presence of blockages ( 6 % ) irrespective of location and temperature tilt (0-50°C) across the subassembly. This has indicated that such a technique has the potential of simplifying core surveillance by enabling the use of a single detection algorithm for all S/A locations. The sensitivity of the blockage detection failure probability to variations in the overall decision time has also been investigated using this approach. The third paper,
from ENEA,
dealt with the use of standard
statistics and functions as
failure indicators in the evaluation of the SILOE-experiments RSI. (One defected preirradiated SPXl-pin). The RMS-value of temperature fluctuations is comparable to the DND method in detecting cladding failure. Additional method.
multi-pin
fuel
element
tests
are
planned
for
further
evaluation
of
this
CONCLUSIONS AND RECOMMENDATIONS
The work that has been performed on rigs, experimental and prototype reactors, physical models and signal processing developments has provided tools which enable the potential usefulness of temperature fluctuation monitoring to be evaluated in experimental situations.
I.W.G.F.R. meeting
125
With the development of commercial breeder reactors the emphasis should now be direct towards the further evaluation of phenomena which must be considered for their safety or operational consequences. This work should develop the phenomenological description of the initiation and evolution of cooling anomalies, as already started. For anomalies which are potentially detectable by temperature fluctuation monitoring there is a requirement to investigate the efficiency, reliability and design consequences of the proposed system and to compare them with the benfeit of other surveillance systems. Additionally, the influence of design and operational constraints (flow mixing at the outlet, power tilts, load following) should be included in the consideration. The
participants
of
the
meeting
recommend
that
further
work
be
carried
out
on
the
following 3 areas: 1 -
2
-
Further developments and justification of current tools to cope with realistic subassembly geometries including transport of temperature information to the sensor and optimisation of signal processing strategies. The
specification
of
realistic
subassembly
phenomena
that
should
and
could
be
monitored by temperature fluctuations. 3 - The efficiency and reliability of temperature noise monitoring for commercial plants.
The status of the work should be discussed at specialists' IWGFR at the appropriate time.
meetings sponsored by the
26
M. MOTrA
REFERENCES
/l/
G. Cantoni, R. Giovannini, M. Piani "Temperature Noise Analysis on a LMFBR defective
fuel pin
irradiated in a sodium
loop"
121 D. Firth and B. O. Wey "Review of temperature noise studies in the UK"
131 R. Giovannini, M. Motta "The activity for the development of the thermal noise detection and analysis at the Italian Committee for Nuclear and Alternative Sources of Energy (ENEA)"
141
J.P. Girard "Temperature-Fluctuation measurements and their use in the french LMFBR program"
151 G. Hughes, R.S. Overton and B.O. Wey "STATEN
Predictions
in
a
CDFR
Subassembly
and
the
Use
of
Simulated
Outlet
Temperature to Test Decision-making Techniques"
161
L. Krebs, G. Weinkotz "Survey on the german
research programme
for blockage
detection within
LMFBR fuel
elements by measurements and analysis of temperature fluctuations" /7/ L. Krebs, K. Bremhorst "Verification of the extended gradient diffusion model by measurements of the mean and fluctuating
temperature fields
in sodium
flow downstream of a multi- Bore jet
block /8/ M. Marseguerra, C.M. Porceddu Cilione "Accuracy evaluation of thermocouple time constants via the bootstrap method"
lel
K. Nakamoto, N. Ohyma, H. Rindo, M. Kamei "Research and Development Activities on Temperature Measurement for LMFBRs"
11Ol R.S. Overton, B.O. Wey, G. Hughes "The temperature Noise Simulation Code STATEN and its Experimental Verification"
1111 T. Sekiya, T. Takeda, K. Nakamoto, K. Kawano, N. Ohyma "A Study
on Detection
of Temperature
Fluctuation Caused by Local Flow Blockage in
37-Pin Bundle" /12/ B.O. Wey, C.J. Brown "CDFR Coaxial Thermocouple Response Measurements at Nigh Reynolds Numbers"
I.W.G.F.R. meeting APPENDIX
127
I
INTERNATIONAL
ATOMIC
ENERGY
AGENCY
- INTERNATIONAL
Organized REACTOR Specialists'
Meeting
-
PHYSICS
WORKING
GROUP OF FAST REACTORS
by the ENEA
AND NOISE
ANALYSIS
LABORATORY
METHODS
AND TOOLS TO DETECT
THERMAL
October
8-10
- Italy
1984 - Bologna
NOISE
IN FAST REACTORS
PARTICIPANTS FEDERAL
REPUBLIC
OF GERMANY
KREBS
L.
T. TAKEDA
Kernoforschungszentrum
Dep.
Institut
fuer R e a k t o r b a u e l m e n t e
OSAKA U N I V E R S I T Y
POSTFACH
3640,D-7500
Yamadaoka
KARLSRUNE
FRANCE J.P.
UNITED
GIRARD
C.E.N.
of Nuclear
Engineering
2-l,Suita,
Osaka
565
KINGDOM
D. FIRTH
- CADARACHE
UKAEA
DRNR/SPCI/LSPC
RISLEY
BP 1 - F-13115
SAIN
PAUL LEZ DURANCE
ITALY
NUCLEAR
WARRINGTON B.O.
LABORATORIES
WAS 6AT
WEY
CEGB A. FEDERICO
Berkeley
ENEA-TERM
BERKELEY,
Str.
Anguillarese,Km
Nuclear
Labs.
GLOS GLIS-9PB
1,300
CASACCIA-ROMA
Commission
of the European
Communities R. GIOVANNINI ENEA
-
TIB/FICS/ANALRUM
Via M a z z i n i , 2
Z. D I A M A N T I D I S
- BOLOGNA
Joint
Research
Centre
Ispra Establ. M.
MARSEGUERRA
Istituto
1-21020
Ingegneria
- ISPRA
Nucleare
Politecnico
di Milano
Via Ponzio,
16 - MILANO
I.A.E.A. V.F.
EFIMENKO
M. MOTTA
International
ENEA
Agency
-
TIB/FICS
Via Mazzini,
2 - BOLOGNA
Wagramerstrasse, A-1400
C.M.
PORCEDDU
ENEA
-
CILIONE
TIB/FICS/ANALRUM
Via Mazzini,
2 - BOLOGNA
R. TAVONI ENEA
-
VEL
Via Dell'Arcoveggio,
56/23
-
BOLOGNA
JAPAN T. SEKIYA Dep.
of Nuclear
Engineering
OSAKA U N I V E R S I T Y Yamadaoka
2-1,
Suita,
Osaka
Atomic
B66
VIENNA
Energy
B PO Box
i00
28
M. MOlTA
OBSERVERS ITALY P. AZZONI ENEA - TIB/FICS/ANALRUM Via Mazzini, 2
M. PIANI ENEA - TIB/FICS/ANALRUM
BOLOGNA
BOLOGNA
B.
-
2
F. CASALI ENEA - TIB/FICS Via Mazzini, 2 BOLOGNA G. CANTONI ENEA - TIB/FICS/ANALRUM Via Mazzini, BOLOGNA
2
V. SANTJUST Istituto Ingegneria Nucleate
BENZI
TIB Via Mazzini, BOLOGNA ENEA
Via Mazzini,
2
Politecnico di Milano Via Ponzio,16 - M I L A N O A. VIGO ENEA-VEL/SIST/PRONOC Str. Anguillarese, km 1,300 CASACCIA - BOMA
COUNTRY FRANCE
NAME OF POWER PLA~IT
TYPE OF PROBE
FREQUENCY CUT-OFF OR TIME CONSTANT
8
8
I
I
I
DISTANCE BETWEEN SENSOR AND THE SUBASSEMBLY EXIT
I
MEAN TEMPERATURE: SAFETY OR CONTRCL~
F
|
!.i X~
g
6~
~u!~om "~I'~l'~'A~'I
TEMPERATURE FLUCTUATIOPI : SAFETY, CONTROL OR EXPERIMENTAL?
TABLE 2
..L
3~
NUMBER OF INSTRUMENTED S/AS
COMMENTS o
o
MEAN TEMPERATURE FLUCTUAT]Or] PART
z
m ul (name o f system)
K N K I[
Probe i n Sodium K
K and SS/Na s p e c i a l instrumentation d e v i c e
% 1 . 5 Hz
%30 mm
Safetys experimen_ tal KASUMOS and MIRA
Experlmental
32
K
32 (operation TC
©
k : ~ 7 HZ SS/Na 20 Hz
1-2 spec.instr.
>
Additional SNR
300
Thimble K
~0.5
Hz
30 ml.
Safety
Experimental
250
Max 60
SS/Na!
instrumentation
a t few S / A - o u t l e t s is under llscussion.
J d
~t is considered
SNR 2
Thimble K
0.8
llz
~ 3 0 mm
Safety
Safety
(?)
to combine the
temperature fluc t u a t i o n method (acoustic, KASUMOS, DND) to Dne conu~on s u r v e i l l a n c e system for SNR2.
ITALY,
COUNTRY
NAME OF POWER PLANT
TYPE OF PROBE
FREQUENCY CUT-OFF OR TIME CONSTANT
DISTANCE BETWEEN SENSOR AND THE SUBASSEMBLY EXIT
MEAN TEMPERATURE: SAFETY OH C~NTRCL?
7
TEMPERATURE FLUCTUAT[O~I: SAFETY, CONTROL OR EXPEEIMENTAL?
ro
Z
o
COUNTRY
JAPAN
NAME OF POWERPLANT
TYPE OF PROBE
P
8
J,,
o
FEEOUENCY CUT-OFF
OR TIME CONSTANT o
8
~
|
|
DISTANCE BETWEEN SENSOR AND THE SUBASSEMBLY EXIT
MEAN TEMPERATURE: SAFETY OR CONTROLq
,7 TEMPERATURE FLUCTUATION: SAFETY, CONTROL OR EXPERIMENTAL?
g.
-7-. g
~
vJ-LoIAI "]AI
~
COUNTRY
U.K.
NAME OF POWER PLANT
~E TYPE OF
PROBE
FREQUENCY CUT-OFF OR TIME CONSTANT
w
DISTANCE BETWEEN SENSOR AND THE SUBASSEMBLY EXIT
MEAN TEMPERATURE: SAFETY CN C~;TRC!. ~
?
Z~ TEMPERATURE FLUCTUAT[O~i: SAFETY, CONTROL OR EXPERIMENTAL?
>
Zg
=
134
M. MOTrA
-DISCUSSION-
BURTON
There is considerable interest in thermal fluctuations not only to detect anomalies in the nuclear core but also because of their possibie e f f e c t on s t r u c t u r a l integrity. Would i t t h e r e f o r e be f i n a l y t o c o n s i der a broadening of the examination ? Particularly to extend the worth of conclusion 1. T h a t i s , t o s t u d y t h e way t h e r m a l f l u c t u a t i o n s are transported to load c a r r y i n g members of t h e r e a c t o r c i r c u i t .
MOTTA
In my personal opinion it should be useful this broadening of examination. Anyway, the general problem of the transport of the thermal signal from the source to the sensor has to be better investigated because of the sensitivity to the distance and structures. I agree that the problem could be considered in the next meeting of I/IEA on this special topic.