On-site diagnosis for cast-resin transformers

Transcription

On-site diagnosis for cast-resin transformers
ON-SITE DIAGNOSIS FOR CAST-RESIN
TRANSFORMERS
SUMMARY
B Louyot
FRANCE - TRANSFO, France
Cast-resin transformers have been widely used for a
long time. These transformers have the advantage of
easy installation and improved fire behaviour in case of
class F1.
To improve even more reliability during operation, an
inspection of the transformers may be carried out: onsite
partial discharges measurements are a mean to
undertake this inspection.
ON-SITE PARTIAL DISCHARGES
MEASUREMENTS PURPOSE
During operation, transformers may be severely
stressed: overloading, overvoltages, pollution,
conductive particles deposit…
Installation conditions may be more or less severe:
closely cables, high connection resistance…
Some of these installation and operating conditions may
lead to deterioration and PD inception.
ON-SITE MEASUREMENT TECHNIQUES
PD measurements are difficult ones, but are well
mastered when they are carried out in workshops after
manufacturing.
The challenge is to carry out these measurements on site
where the transformers are installed and operating.
There are some difficulties: lack of accessibility around
the transformer, ambient disturbances.
To take into account these difficulties, several kinds of
detection and localisation may be used:
Œ
Detection in the Ultrasonic frequency range of the
acoustic waves created by PD
Œ
Detection of the electromagnetic waves created by
PD.
Œ
Electrical measurement with portable instrument
All of these kinds of measurement may have advantages
and disadvantages. They may be used together to give
more pieces of information.
ON-SITE MEASUREMENT EXAMPLES
For onsite measurements, these 3 kinds of
measurements may be used alone or together. It depends
on installation conditions, space around the transformer,
switching off opportunities.
First example:
The aim of the measurement was to check transformer
electrical ageing.
Measurements with US detector and electrical portable
instrument were used.
Ultrasonic noises coming from the magnetic core were
detected: laminates magnetostriction gives noises which
frequency range may go up to several ten kHz.
Second example:
The upper part of the transformer coils was dusty. The
aim was to check if this deposit could be harmful by
creating partial discharges.
Measurements with ultrasonic detector and detection of
electromagnetic waves were used:
In this case, onsite measurements showed that this dust
deposit did not bring measurable partial discharges.
Third example:
The aim was to check transformer installation.
Measurements with US detector and electrical portable
instrument were used.
In this case, an abnormality was detected: a cable too
close from one of the coils. Ultrasonic detection
pinpointed this source. Because of rather high level
noise, it was not possible to detect these partial
discharges by electrical means.
CONCLUSIONS
The aim of the above examples was to check the
operating conditions.
Some bad operating conditions may be detected, such
as: metal or conducting particles in high electrical field,
small clearances between all feeding cables and the
surface of the transformer coils.
These examples showed feasibility of undertaking an
inspection of transformers. Depending on the case, this
inspection may be more or less sensitive. Carrying out
three kinds of measurements should allow undertaking
transformer inspection in many cases.
These onsite measurement techniques are a mean to
develop diagnosis and help users in improving
preventive maintenance.
DIAGNOSTIC SUR SITE POUR
TRANSFORMATEURS ENROBES
RESUME
B Louyot
FRANCE - TRANSFO, France
L’utilisation des transformateurs enrobés est maintenant
largement répandue. Ces transformateurs présentent
l’avantage de la simplicité d’installation et d’un
comportement au feu amélioré avec la classe F1.
Pour augmenter la fiabilité en service du matériel, il est
intéressant de pouvoir faire un état des lieux des
transformateurs en exploitation. Les mesures de
décharges partielles sur site sont un moyen prometteur
pour faire cet état des lieux.
INTERET DES MESURES DE DECHARGES
PARTIELLES SUR SITE
En exploitation, les transformateurs sont soumis à
diverses contraintes : surcharges, surtensions, pollution,
dépôt de particules éventuellement conductrices, … Les
conditions d’installation peuvent être plus ou moins
contraignantes : proximité des câbles d’alimentation,
mauvais contacts au niveau des raccordements, …
Certaines de ces conditions d’installation et
d’exploitation peuvent conduire à des détériorations et à
l’apparition de décharges partielles.
conditions d’installations, de la place disponible autour
du transformateur, des possibilités de consignation.
Premier exemple :
Le but des mesures était d’évaluer le vieillissement
électrique du transformateur.
Les mesures par détecteur ultrasons et les mesures
électriques par couplage galvanique ont été utilisées.
Une émission ultrasonore en provenance du circuit
magnétique a été détectée : la magnétostriction des tôles
engendre des bruits dont le spectre de fréquence peut
monter jusqu’à plusieurs dizaines de kHz.
Deuxième exemple :
La partie supérieure du transformateur était
poussiéreuse. Le but de l’inspection était d’estimer le
danger de ces poussières. Les mesures par détecteur
ultrason et la détection des ondes électromagnétiques
ont été utilisées : les mesures ont montré que le dépôt de
poussières n’apportait pas de décharges partielles
mesurables.
LES TECHNIQUES DE MESURES SUR SITE
Troisième exemple :
Les mesures de décharges partielles sont des mesures
délicates mais maîtrisées quand elles sont effectuées en
usine en fin de chaîne de fabrication des transformateurs
enrobés.
La problématique est donc d’effectuer de telles mesures
sur les sites d’installation et d’exploitation des
transformateurs. Les difficultés sont les suivantes :
manque d’accessibilité autour du transformateur,
perturbations ambiantes rayonnées ou transmises dans la
gamme de fréquences du phénomène de décharges
partielles.
Compte tenu de ces difficultés, plusieurs types de
détection et de localisation peuvent être utilisées :
Œ
Détection des ondes acoustiques générés par les
décharges partielles, dans le domaine des ultrasons
Œ
Détection des ondes électromagnétiques générés
par les décharges partielles
Œ
Mesure électrique avec matériel portable
Chacun de ces types de mesure a ses avantages et
inconvénients. Tous ces types peuvent être
complémentaires.
Le but était de vérifier l’installation du transformateur.
Les mesures par détecteur ultrasons et les mesures
électriques par couplage galvanique ont été utilisées.
Un défaut a été détecté : un câble passait trop près d’une
des bobines. La détection ultrasonique l’a très bien mis
en évidence alors que la mesure électrique n’était pas
sensible du fait d’un bruit de fond important.
CONCLUSIONS
Le but des essais cités était de vérifier les conditions
d’exploitation des transformateurs.
Certaines conditions d’exploitation anormales peuvent
être détectées : particules métalliques ou conductrices
dans un champ électrique élevé, distances insuffisantes
entre câbles et surface des bobines.
Ces exemples montrent la faisabilité d’inspection de
transformateurs. Suivant les cas, ce diagnostic peut être
plus ou moins sensible. La mise en œuvre de trois types
de mesures différentes devrait permettre d’entreprendre
un diagnostic dans beaucoup de cas.
EXEMPLES DE MESURES SUR SITE
Pour les mesures sur site, toutes ou partie des types de
mesures peuvent être utilisées, en fonction des
Ces techniques de mesures sur site permettent de
développer le diagnostic et d’aider les exploitants dans
l’amélioration de la maintenance préventive.
ON-SITE DIAGNOSIS FOR CAST-RESIN
TRANSFORMERS
B Louyot
FRANCE - TRANSFO, France
Cast-resin transformers have been widely used for a
long time. These transformers have the advantage of
easy installation and improved fire behaviour in case of
class F1.
Climatic and environmental classification tests E2 and
C2 show these transformers have a good behaviour
under severe operating conditions. Experience shows
their good reliability.
To improve even more this reliability during operation,
an inspection of the transformers may be carried out:
on-site partial discharges measurements are a mean to
undertake this inspection.
These onsite diagnoses could show abnormal situations
such as: too stressed operating conditions, too severe
installation conditions, abnormal ageing of the
transformer, …
All these events may create partial discharges. In this
case, transformer life-time may be shortened, depending
on partial discharge harmfulness.
ON-SITE PARTIAL DISCHARGE
MEASUREMENT PURPOSE
Partial discharge measurements were measured on HV
terminals according to IEC 270, during an inducedvoltage test up to 1.5 Um (Um = 24 kV for this
transformer). Results (see table 1) show that for some
conductive path, partial discharges appear for voltages
lower than rated voltage. These partial discharges are
located on the surface of the coil on the conductive path
border. With time, they may lead to electrical
breakdown.
TABLE 1 – Partial Discharge measurements
Voltage
(kV)
Partial discharge (pC)
<5
<5
55
10
<5
<5
425
20
Without
With two
With one large
patch
patches
patch 170 mm
30x30 mm
height, 30 mm
wide
Examples of conductive patches on outer side of HV
cast-resin coil:
After transformer manufacturing, several quality control
tests are carried out. Among these, checking that they
are partial discharge free. No manufacturing defect nor
design defect should bring partial discharge inception at
voltages around rated voltage. If there is such defect, its
harmfulness will lead to transformer collapse, more or
less rapidly.
During operation, transformers may be more or less
severely stressed: overloading, overvoltages, pollution,
conductive particles deposit…
Installation conditions may be more or less severe:
closely cables, high connection resistance, …
Some of these installation and operating conditions may
lead to deterioration and partial discharge inception.
EXAMPLES OF PARTICULAR OPERATING
CONDITIONS
With two patches
30x30 mm
With one large patch 170
mm height, 30 mm wide
Conductive particles deposit
Examples of ageing under partial discharges
Laboratory experiments have been carried out to
simulate the effect of conductive particles deposit on
cast-resin HV coils. For these experiments, silver lack
was used to simulate conductive patches on outer side
of coil, for a 400 kVA, 20 kV transformer.
Behaviour of electromechanical materials
Laboratory experiments have been carried out to assess
the behaviour of electromechanical materials operating
with partial discharges. Partial discharges can be
measured continuously during operation until
breakdown. But measurement analysis and
interpretation vs. life duration are still in progress.
FIGURE 1 – partial discharges during operation for
a 30 kV rotating machine bar
Life duration for enamelled wire
Life duration under electrical field was assessed for
enamelled wires. Voltage was applied between the two
wires of a twisted pair as shown on photograph. The
twisted pairs may be cast or not.
For cast-resin twisted pairs, no partial discharge
occurred before breakdown.
For bare twisted pairs, partial discharges appeared
around 800 V. In this case, life duration under partial
discharges vs electrical field could be assessed.
FIGURE 2 – partial discharges measurement on
bare twisted pair
FIGURE 3 – life duration under partial discharges
vs electrical field
100000
10000
Time
Failure
time to failure (h)
1000
100
10
1
5
7
9
electrical field (kV/mm)
11
13
ON-SITE MEASUREMENT TECHNIQUES
Partial discharge measurements are difficult ones, but
are well mastered when they are carried out in
workshops after manufacturing.
The challenge is to carry out these measurements on site
where the transformers are installed and operating.
There are some difficulties:
Œ
Lack of accessibility around the transformer
Œ
Ambient disturbances
To take into account these difficulties, several kinds of
detection and localisation may be used:
Œ
Detection in the ultrasonic frequency range of the
acoustic waves created by partial discharges
Œ
Detection of the electromagnetic waves created by
partial discharges.
Œ
Electrical measurement
All of these kinds of measurement may have advantages
and disadvantages. They may be used together to give
more pieces of information.
Detection in the ultrasonic frequency range of the
acoustic waves created by partial discharge
Discharges in air or on solid insulation material surfaces
create acoustic pressure waves that may be detected by
an ultrasonic receiver. Within ultrasonic frequency
range, detection sensitivity is rather good and noise
level rather low.
Measuring system is made of an ultrasonic sensor, a
parabolic reflector. The ultrasonic signals are converted
into audible signals via an earphone.
Detection level is around some 10 picoCoulombs or so.
Partial discharge searching is made around the
transformer with a clearance according to safety rules.
This kind of detection method does not allow detecting
internal partial discharge within the windings, due to
bad propagation within polymeric materials.
This method detects as well other acoustic wave sources
than partial discharges: for example magnetic laminates
vibrations. It may as well be disturbed by wave
reflections.
This method is a very good one to pinpoint ultrasonic
sources in air.
Detection of the electromagnetic waves created by
partial discharges
Electromagnetic transients created by partial discharges
may be picked-up and measured by means of capacitive
or inductive sensors.
Searching may be made with the apparatus around the
transformer with a clearance according to safety rules.
Sensitivity is nevertheless rather poor.
If switching off the transformer is possible, the
apparatus may be placed closer to the transformer on an
insulated tripod. Sensitivity is then fine. Detection and
localisation within the transformer is then possible.
Electrical measurement
When electrical measurement is possible, it must be
carried out according to IEC 726 standard, dealing with
dry-transformers. It needs a lot of materials. For
example, for a 3-phase transformer: a 3-phase supply, 3
capacitors, and partial discharge measuring device. It is
the only way to carry out measurements that give
reliable results. Unfortunately, there are often
disturbances in electrical areas and noise level is rather
high.
In case such a measurement is not possible, some
portable device may be used: it is connected to one of
the transformer terminals and isolated from ground or
any metallic part. The measured signal is transmitted by
means of an optic fibre and processed. This method
which allows measurement on terminals, may detect
internal and external partial discharges and is a sensitive
one. But it may give erroneous results: measurements
conditions are not at all the same than according to IEC
726, concerning potential to earth for example. Thus,
one must consider that this method gives only
indication.
ON-SITE MEASUREMENT EXAMPLES
For onsite measurements, these 3 kinds of
measurements may be used alone or together. It depends
on installation conditions, space around the transformer,
switching off opportunities.
First example:
The aim of the measurement was to check transformer
electrical ageing.
Measurements with US detector and electrical portable
instrument were used:
Œ
first, searching with the ultrasonic detector showed
noises in the upper part of the transformers, near
the magnetic core.
Œ
Then, electrical measurements with galvanic
coupling were carried out: they showed there was
no partial discharges above the noise level (3 pC).
It seemed that the detected ultrasonic noises were
coming from the magnetic core: laminates
magnetostriction gives noises which frequency range
may go up to several ten kHz.
Second example:
The upper part of the transformer coils was dusty. The
aim was to check if this deposit could be harmful by
creating partial discharges.
Measurements with ultrasonic detector and detection of
electromagnetic waves were used:
Œ
ultrasonic noises were detected but only near the
magnetic core.
Œ
For detection of electromagnetic waves, the
instrument was put on an isolated tripod very near
the transformer, during interruption of high voltage.
It was placed where the ultrasonic noises had been
located. In these conditions, detection level was
relatively low. These last measurements showed
there were no partial discharge above the noise
level (15 pC).
In this case, onsite measurements showed that this dust
deposit did not bring measurable partial discharges.
Third example:
The aim was to check transformer installation.
Measurements with US detector and electrical portable
instrument were used:
Œ
First, searching with the ultrasonic detector showed
noises in the upper part of the transformers, near
the magnetic core. Noise was also detected between
one of the transformer coil and a LV cable situated
along the coil, 10 cm from its surface.
Œ
Then, electrical measurements with galvanic
coupling were carried out: they showed there was
no partial discharges above the noise level (50 pC).
In this case, an abnormality was detected: a cable too
close from one of the coils. Ultrasonic detection
pinpointed this source. Because of rather high level
noise, it was not possible to detect these partial
discharges by electrical means.
CONCLUSIONS
Partial discharge measurements on site are not
contractual for transformers. Their aim is to check
operating conditions.
Some bad operating conditions may be detected, such
as: metal or conducting particles in high electrical field,
small clearances between all feeding cables and the
surface of the transformer coils.
The above examples showed feasibility of undertaking
an inspection of transformers. Depending on the case,
this inspection may be more or less sensitive. Carrying
out three kinds of measurements should allow
undertaking transformer inspection in many cases.
These onsite measurement techniques are a mean to
develop diagnosis and help users in improving
preventive maintenance.
References:
Kreuger F.H, 1989, “partial discharge detection in
high-voltage equipment”, Butterworth & Co,
London, UK
.
IEC 60270, 1981, “partial discharge measurement,
Geneva, Switzerland
.
Bartnikas R and McMahon E.J., 1984,
“Engineering dielectrics: corona measurement and
interpretation”, volume 1,ASTM, Philadelphia,
USA
.