Xlpe .I HZ VLF

Session 27
Preventive Maintenance Testing of

Primary XLPE Underground Distribution Cables
and the Application of the .I Hz VLF Test Method
By Henning Oetjen, HDW Electronics, Inc.
Summarv
VLF testing is being considered as the safe alternative to DC Hipot testing of XLPE and
mixed cables because of the known problem of induced space charges, which will cause
premature failures in this type of cable. The IEEE work group 400.2 is currently working
on a recommendation regarding the VLF test methodology. The paper compares
different VLF technologies and discusses the correlation between field test data and test
parameters.
Introduction
Deregulation is having a big impact on the utility industry in the US in many ways.
Increasing the System reliability has become one of the key objectives in the current
environment. With regard to underground distribution cable systems consideration must
be given to three components since they will effect the number of outages and their
duration (Saidi / Saifi). These three components are restoration and specifically the
employed fault locating methods and techniques, preventive maintenance and predictive
maintenance test methods. Fault locating methods must be included in the considera-
tions since they not only impact the SAID1 but will also have a pronounced effect on the
preventive and predictive maintenance data.
Preventive Maintenance Testinq
DC Hipot testing was initially used to test oil filled transformers and was (and still is)
used to test PlLC cables. However when applied to XLPE cables numerous studies
have established beyond any doubt that DC Hipot testing will cause premature failure in
this type of cable due to the formation of space charges. Because of the continuing
replacement and substitution of PlLC cables by XLPE cables the .1 Hz AC VLF test
method was developed in late 80's to replace the DC Hipot test. This change in test
philosophy is also documented by the fact that in the US as well as in Europe new test
specifications are generated by the respective Standard Comittees (IEEE and IEC),
which eliminate the use of DC Hipot testing for XLPE cables.
Any preventive maintenance test is primarily a Go / no Go test. The .1 Hz VLF test is no

exception in this regard. The DC Hipot test fits the same characterization, however it
offers in addition also some limited diagnostic information with regard to the cable
insulation by means of its leakage current. Even today's VLF test instruments cannot
provide this information the capability to measure the leakage current will become
available in the very near future in conjunction with the .1 Hz cos rectangular VLF
technology. This feature is of particular interest since it allows to correlate previously
gathered DC Hipot leakage data with new VLF test leakage data, provided the .1 Hz cos
rectangular VLF technology is employed.
02000 IEEE.
0-7803-6625-5/00$10.00 3 54
VLF Technoloqies
The difference in the available technologies is reflected in the wave shape of the .1 Hz
AC voltage. Aside from this difference there are also common aspects to them. Both
show a significantly accelerated growth rate for electrical trees when compared to that at
power frequency. VLF testers using a .1 Hz AC frequency are substantially smaller in
weight and size when compared to 60 Hz equipment. For a given test voltage and cable
test capacity the ratio between the 60 Hz and the .1Hz cos rectangular is 20 to 1 and
between the .1 Hz sinusoidal and the .1 Hz cos rectangular still 5 to 1.
This is directly related to the fact that the full charging energy for each half wave must be
provided respectively dissipated twice within each cycle. The .1 Hz cos rectangular
technology utilizes a patented energy recovery system, which recuperates approx. 90 YO
of the required charging energy between half waves.
The difference in wave shape has a significant effect on the rate of change of the
potential. The sinusoidal .1 Hz wave takes 5 seconds to accomplish the change in
polarity compared to the cos rectangular .1Hz wave, which takes only 5 milli-seconds
(average), thus matching the 8 milli-seconds of the 60 Hz power frequency very closely.
The rate of change is relevant in regards to the electrical stress level and can be
compared to the ramp-up speed during a high potential test,
Another difference must be taken into account when comparing the test voltages for both
technologies. In the case of the .1 Hz cos rectangular wave the RMS value and the peak
value of the test voltage are identical (typically applied test voltage = 3 times Vo , where
Vo represents the phase to ground voltage). The peak value of the equivalent test
voltage for the sinusoidal wave equals 4.2 times Voor a 40% higher test voltage.
The above-mentioned 5 to 1 weight ratio between the sinusoidal and the cos rectangular
technology makes it less practical to build sinusoidal test sets for a large cable capaci-
tance. Therefore it is common to offer sinusoidal test sets with lower than .1 Hz test fre-
quencies, i.e. .02 Hz, .05 Hz and even .01 Hz in order to be able to test cables with a
large capacitance (2 times respectively 5 or 10 times when compared to the capacitance
@ .1 Hz). When this solution is applied the duration of the test must be increased in-
versely proportional to the reduction in frequency in order to maintain comparability
between test results which are gathered on a number of cables at different test
frequencies. As mentioned above the growth rate of the electrical trees is frequency
dependent. Published test results show a significantly slower (10 times) growth rate for
.01Hz compared to .1 Hz.
Test Parameters versus Meaninq of Test Results
Since the late 80's when the .1 Hz cos rectangular technology was first introduced
thousands of cables have been tested by applying a one hour test duration and a test
voltage of 3 Vo. A cable, which passes this test protocol, will statistically not fail within the
next 2-3 years. A change in test parameters (voltage and or time) will produce test
results, which require a new interpretation, which itself will require a substantial and new
database with field test results.
355
On the other side one cannot assume that different test methods, which apply the same
or similar test parameters, but cause different phenomena within the cable insulation, will
have the same meaning with regard to their test results
10 Year Field Experience Report
In the last part of the paper a brief summary of the field test data is provided, which had
been collected by a single customer over a 10 year period, using the cos rectangular .1
Hz VLF technology. This customer is one of several hundred customers worldwide, who
is using this technology and has tested approx. 2,200 miles of cables in his distribution
system. Like many other customers he has PILC, XLPE and mixed cables in the system.
The most significant results are as follows:
1. a total of 120 faults occurred

2. during the 60 minute test Q 3V0 66% of all breakdowns occurred during the first 10
minutes, 75 Yoof all breakdowns occurred during the first 30 minutes
3. approx. 10% of all faults occurred between 30 and 40 minutes
4. approx. 10% of all faults occurred between 50 and 60 minutes
5. statistically no faults occur after 60 minutes, this also confirms, that no new faults in
service aged cables are created because of the prescribed test parameters.
6. PlLC cables have an average fault rate of one per 20 miles
XLPE cables one per 14 miles
7. All faulty joints failed regardless of their design within first 20 minutes (88% within
first 10 minutes)
8. Faults in PlLC cables occur within the specified time inspite of a lower voltage level
when compared to the level in the DC Hipot test; this qualifies the cos rectangular .1
Hz VLF test for mixed cable installations
9. Existing water trees (XLPE cables) frequently fail shortly before the completion of the
60 minutes test; therefore it is important to adhere to the 60 minutes test duration
356

Xlpe .I HZ VLF

Uploaded by

Copyright:

Available Formats

Xlpe .I HZ VLF

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Xlpe .I HZ VLF

Uploaded by

Copyright:

Available Formats

Session 27

Preventive Maintenance Testing of

Preventive Maintenance Testinq

Any preventive maintenance test is primarily a Go / no Go test. The .1 Hz VLF test is no

Test Parameters versus Meaninq of Test Results

10 Year Field Experience Report

1. a total of 120 faults occurred

You might also like