Академический Документы
Профессиональный Документы
Культура Документы
ART
explored
...p3
Tackling
TDR
theory
...p4 - 5
Power
problems?
Try mice!
p8
Published by Megger
April 2014
ELECTRICAL
TESTER
Moisture measurement
Introduction
It is well documented that moisture in a
mineral oil cooled and insulated power
transformer has detrimental effects. In fact,
it is said that doubling the moisture content
in the transformer will have the effect of
approximately halving the life of the unit.
Thus the transformer will deliver only half the
expected return on investment and its reliability
will be impacted earlier in its life. This article
looks at some of the methods of determining
the amount of moisture that will affect the
operation of the unit and the subsequent
management of the oil and paper systems.
www.megger.com
Monitoring
Figure 2: Moisture data plotted on a set of inverted equilibrium curves (2.5% - 3%)
ELECTRICAL
TESTER
Contents
Moisture measurement.......................... P1-3
Tom Dalton - Business unit manager: power
transformers, Martec, South Africa
University at Buffalo
collaboration brings benefits for all.....P6
Casey Henry, marketing program manager
Q&A...................................................................P8
The stability factor......................................P8
Damon Mount - power sales manager
Megger Limited
www.megger.com
System approach
The system approach is far more refined
and uses an electrical means to measure
the system rather than trying to measure
one of the components (either oil or air)
and calculating the resultant moisture. This
method is finding greater acceptance and is
improving as the technology matures and gains
momentum. Frequency Domain Spectroscopy
(FDS) or variable frequency dissipation factor
measurement (tan delta) takes both oil and solid
insulation into consideration. The instrument
uses the data measured (dielectric dissipation
factor or tan ) and models this to a known
curve, which then equates to moisture in oil and
moisture in insulation. That is, it takes the whole
insulating system into consideration.
This measurement cannot be done on line and
the transformer will need to be disconnected
from the network. The windings are normally
Management
Many people ask how much moisture should
be allowed or is good practice. To answer this
question three categories of transformer need
to be considered:
New
In repair process
In service
www.megger.com
ELECTRICAL
TESTER
In summary
Keeping transformers dry is the preferred
practice. To do this, a molecular sieve or other
on-line drying technology can be deployed for
continuous drying of the oil and solid insulation
system, thus avoiding the need for major dry
outs when the unit is found to be wet.
Measuring moisture in an operating transformer
is not practical without the oil temperature
being taken along with the sample. However,
for reliable results it is best to have a trained
sampler to take the sample from the
transformer using the correct techniques and
equipment, and transporting the sample reliably
to the laboratory. Oil sampling is still a good
first line defence, but follow up measurements
must be made if the transformer shows signs of
undue moisture increases that are unrelated to
variations resulting from the sample process.
The other methods that are mentioned in this
article are second tier methods and are used
to gather further and more detailed condition
information. However, in many cases these
methods have their own limitations. A good
solution is the frequency domain spectroscopy
technique which has made this uncertainty
a thing of the past. Spectroscopy and data
modelling offer clear advantages in modern
science and laboratory practice, and now FDS
has brought this modern laboratory technology
into the field, complemented by well-proven
modelling techniques that provide very specific
decision support for the user.
Different asset owners have different priorities
for the operation and maintenance of
transformers. Some, for example, focus on
optimisation and return on investment rather
than reliability, whereas for others reliability
is paramount. In reality, however, a balanced
approach is best.
In closing, it is always worth remembering that
a wet transformer is an unreliable transformer.
www.megger.com
ELECTRICAL
TESTER
Power reflectometers
Power reflectometers have at their heart a fast
and expensive flash analogue-to-digital converter.
The latest instruments have sampling rates of up
to 400 MHz, which is at least twice the rate used
in the previous generation of instruments.
It is widely believed that the higher the
sampling rate, the better the instrument,
because higher sampling rates provide
increased resolution. In theory this is true,
but in practice there is little benefit to be had
from further increasing sampling rates, as the
resulting higher resolution is only relevant
at short distances. At larger distances, the
instrument display cannot show the full
measured resolution.
(80m/s)80.000km/s
Res =
= 0.2m
400MHz
Range and resolution
To explore in more detail the relationship
between sampling rate, range and display
resolution, we will examine the situation with
a modern power reflectometers that is among
the best of its type the Teleflex SX or its large
equivalent the Teleflex VX.
Assuming that the measuring pulse travels with
a typical propagation velocity of 80 m/s, with
a sampling rate of 400 MHz the theoretical
resolution will be 0.2 m, as this equation shows:
The lowest range of the Teleflex SX is 20 m.
The display is 1,024 pixels wide so, when the
20 m range is in use, in round terms 20 m =
1,000 pixels. This means that 0.2 m = 10 pixels.
On this range, the full theoretical resolution
of the instrument can, therefore, be used.
However, when the 200 m range is selected,
0.2 m is equivalent to only one pixel, so the
usable resolution is starting to be limited by the
display rather than the sampling rate. At longer
ranges, even if the display is magnified with the
zoom function, the limiting effect of the display
is completely dominant and the theoretical
resolution of the instrument calculated from
the sampling rate alone can never be used.
Triggering
With power reflectometers, a widely used
measuring technique is the arc reflection
method (ARM), where a surge generator is
used to produce an arc at the fault location,
and this arc (while burning, it is low resistive)
reflects the pulse from the reflectometer. For
this method of measurement to work well, the
reflectometer needs a very reliable trigger.
The conventional triggering method is delay
time adjustment, where the instrument user
manually adjusts the delay time between
the power surge that strikes the arc and
the release of the measuring pulse from the
reflectometer. This approach requires the user
to have a high level of skill as the delay time
needed depends on several parameters related
to the ARM oscillation, which is determined
by the resonant circuit formed by the surge
capacitor, the cable capacitance and the cable
inductance. These factors change with cable
length, so its clear that a fixed delay time
is not a reliable solution. The objective is to
trigger the reflectometer when the arc from
the ARM discharge has reached its highest
current value and is burning solidly. One
method of compensating for the difficulty of
triggering at exactly the right moment is to
take several reflectometer recordings of the
same ARM shot. This increases the probability
of capturing at least one useable trace.
A better approach, however, is the U
triggering system, which is also known as one
shot triggering technology. This works by
continuously evaluating the current of the ARM
oscillation. After the current has reached its
peak, the instrument waits for the voltage to fall
by a predetermined amount (U) and then
Range compensation
For both power and telecom reflectometers,
the attenuation and dispersion of the cable
cause problems. Impedance and resistive
losses, cable length and other influences
change the amplitude and shape of the
pulse transmitted by the cable. Attenuation
makes the signal smaller as distance
increases. The effect of attenuation is
shown by the red line in Figure 4. Since
attenuation follows an exponential function,
it can be calculated and compensated
for. The distance-related amplitude
correction provided as part of the ProRange
compensation function is shown in Figure 5.
Dispersion is another factor that affects
pulse shape. Higher signal frequencies are
attenuated more than lower frequencies,
www.megger.com
ELECTRICAL
TESTER
Dead zone
The dead zone is the basis for endless
discussions. It results from the fact that the
transmit pulse, which looks to the instrument
just the same as a reflected pulse except
that it is much larger in amplitude, saturates
the instrument for a certain distance at
the beginning of the trace. Only beyond
this distance do reflected pulses become
visible. Various technologies are available
to compensate for the dead zone, one of
the most effective being the use of a split
transformer. The basic circuit for this type of
compensation is shown in Figure 6.
Introduction
A TDR (time domain reflectometer) uses the
radar principle to identify faults on cables.
The instrument fires a pulse down the cable,
and any change in the impedance of the
cable will result in reflections being sent
back down the cable toward the instrument.
These reflections are measured and displayed
so that a map of the cable is shown. Many
faults are found at terminations, cable joints
and other locations where there has been
disruption to the route of the cable.
Since they work by identifying changes in
impedance, TDRs require two conductors
that run parallel to each other in order
to operate. Any connection, change of
cable type, break in the cable, or fault will
cause a change of impedance. Each type
of change has a different effect on the TDR
display; a positive reflection shows higher
impedance, a negative reflection shows
lower impedance.
Using a TDR
Identify the location of fault by testing from
both ends of the cable.
www.megger.com
Cable type
Construction
Velocity factor %
Power
0.72 to 0.84
216 to 252
Power
XLPE
0.54 to 0.62
162 to 186
Power
EPR
0.45 to 0.57
135 to 171
Twisted pair
Polyethylene
0.64 to 0.67
192 to 201
Twisted pair
PTFE
0.71
213
Twisted pair
Paper
0.72 to 0.88
216 to 264
Telecomms
PIC
0.65 to 0.72
195 to 216
Telecomms
Pulp
0.66 to 0.71
198 to 213
Telecomms
Gel filled
0.58 to 0.68
174 to 204
Telecomms
Coax
0.82 to 0.98
246 to 294
Velocity factor
When using a TDR it is necessary to tell it the
speed of the pulse in the cable. This enables
the TDR to convert the time the reflected
pulses take into distance. Different types of
cable have different velocity factors (VF).
VF is the ratio of the speed of the pulse
in the cable to the speed of light. It may
alternatively be entered as m/s when it
is called velocity of propagation. If you do
not know the VF of the cable under test it
may be possible to test a known length and
adjust the VF until the distance displayed to
the end of the cable is correct.
Open conductor
A large positive trace
Short circuit
A negative trace
Range
Cables splice/joint
A small positive followed by a small negative
Wet splice/water
Short positive/negative trace
Water ingress
A negative trace
ELECTRICAL
TESTER
University at Buffalo
Casey Henry
marketing program manager
software modules for the latest state-ofthe-art relay protection test systems. While
participating in the project, the students work
under the guidance not only of the universitys
tutors, but also of Megger engineers who
have wide and current practical experience in
the relay test field and who act as mentors.
The university benefits by having access to the
latest test equipment and from the up-to-theminute input provided by the mentors, while
the students enjoy both of these benefits
plus the confidence-boosting opportunity to
produce work that will ultimately be tested
and used in a commercial environment. The
project makes additional resources available
www.megger.com
ELECTRICAL
TESTER
www.megger.com
Sweden
Bornholm
Island
ELECTRICAL
TESTER
Q&A
Q:
A:
Test sets capable of working at 15 kV
have been produced in response to requests
from customers who, in turn, are typically
responding to the introduction of standards
like NETA MTS 1997 Table10.1, applicable to
the maximum voltage rating of equipment,
NETA ATS 2007 Section 1.5 for mediumvoltage motors, and IEC 60229 2007 for
electric cables with a special protective
function. All of these call for 15 kV testing in
One of the most widely used testing techniques in the power sector is high-voltage dc insulation resistance
testing. The principles of this type of testing are well known, but there are still aspects that give rise to
questions. Here are the answers to some of the most common of these.
some circumstances. While these standards are
relatively specialised, they do demonstrate a
trend toward higher insulation resistance test
voltages so, in addition to those who actually
need to work to these standards, there are also
some users who are specifying 15 kV test sets
as a form of future proofing.
Q: For users who dont need to comply
with these standards, are there
any other benefits to be gained by
testing at 15 kV?
A:
As the use of 15 kV testing spreads, it is
becoming apparent that testing at this voltage
can detect many faults that are not apparent
when tests are performed at lower voltages.
Typically, these faults include fractured or
crushed insulation, damaged cable outer
sheaths, and insulation that has been degraded
by corrosive or conductive contaminants.
www.megger.com