UTTIIL
LIIT
TYYEEXXPPE
ERRIIM
MEEN
NTTSSH
HO OWWSSDDEESST
TRRUUCCT
TIIV
VEE
N
NAATTUURRE
EO OFFVVL
LFFHHIIP
POOT
TTTEESST
TIIN
NGG
Background
A large U.S. utility wanted to better understand the effects of very-low-frequency
(VLF) on their cable system. The utility wanted to reconcile VLF manufacturer
claims with the IEEE 400 standards. If the standards were examined and the
before and after test results were compared, as indicated below, even if one
defect is brought to failure, other electrical trees can be created and/or
deteriorated to levels that will promote successive failures. The data will confirm
IEEE standards and demonstrate that VLF is a destructive test.
IEEE 400-2001 Standard Guide for Field Testing and Evaluation of the Insulation
of Shielded Power Cable Systems categorized VLF as a withstand test according
to the introduction to VLF testing in section 8.1. Furthermore, the standard
states, “In withstand testing, insulation defects are caused to breakdown (fault) at
the time of testing … the withstand test is considered a destructive test.”
According to the VLF manual and brochure, “A VLF test is a pass/fail stress test
and if a cable or splice defect cannot hold the test voltage, it will fail.” Now,
compare this to the IEEE definition of a withstand test. The purpose of a
withstand test is to breakdown the insulation until failure which would categorize
it as a withstand test method. Furthermore, if you examine Table 1 of IEEE 400
section 8.1, it confirms that VLF was designed to grow electrical trees at the
highest rate possible. At 3 times the operating voltage, a 0.1Hz VLF Test will
grow an electrical tree at 10.9 to 12.6 mm/hr compared to power frequency which
is 2.2 to 5.9 mm/hr. In addition, the VLF test duration is further evidence that the
test is designed to promote existing defects. Hence, VLF by design is used to
grow electrical trees whether or not they are deteriorated to their breakdown
point. The test below was designed to confirm this with actual field installed
cable.
Introduction
This application describes a case study where the result of a VLF HIPOT test
was monitored utilizing the IMCORP partial discharge (PD) diagnostic. The
purpose was to answer the questions such as those posed by IEEE 400 section
8.5 regarding VLF HIPOT:
IMCORP- Power Cable Reliability
179 Middle Turnpike ● Storrs, CT 06268 ● Tel: 860.427.7620
www.imcorptech.com
1. At what voltage level can a defect be detected?
2. Is it possible to miss a defect which will fault when the cable is returned
to service?
3. Does the test aggravate what was a negligible defect so that it will fault
when the cable is returned to service?
Although this experiment is not statistically significant, some implications can be
surmised from the observations made. To address these questions, the following
experiment was devised:
1. PD test was performed on a three phase cable.
2. VLF HIPOT was performed on each phase individually.
3. PD tests were repeated on any surviving phases and the results were
compared to the initial PD test results.
Test setup
According to the manufacturer of a VLF source, a 25kV class cable should be
tested with 0.1Hz VLF at 3xUo for 30 minutes. This may be confusing for some
because 3xUo equals 3x13.2Vrms, or 39.6kVrms. The IEEE guide for VLF
HIPOT field testing (IEEE 400.2) recommends 23kVrms for 25kV cables. During
the experiment, the manufacturer of the VLF source explained that even though
their equipment only produces a maximum of 23kVrms, they claim that only the
peak voltage needs to reach ~40kV to have an effective VLF HIPOT. The
manufacturer made the following statements:
Electrical tree growth rate is 472mils/hr or 12mm/hr in XLPE insulation.
A VLF HIPOT left on for 30 minutes will grow any significant defect to
failure in a 25kV class cable.
If the cable fails within the 30 minute HIPOT test, the cable is ‘bad’.
If the cable survives the 30 minute HIPOT test, the cable is ‘good’.
Only 3 to 4% of cables that test ‘good’ will fail within the next three years.
It is recommended that the 30 minute VLF HIPOT be repeated after a
failed cable is repaired to ensure reliability.
Before VLF
HIPOT
2U0
After VLF
HIPOT
1.3U0
Figure 1
Cable Length (ft)
A Phase Partial Discharge Test Results
Before and After VLF HIPOT
Discussion
This experiment raises a number of interesting questions:
Why didn’t ‘C’ phase fail within 30 minutes?
What would have happened if ‘C’ phase was put back in service after a 30
minute HIPOT which determined it to be a ‘good’ cable?
If ‘A’ phase had so many defect sites, why did the VLF HIPOT determined
it to be ‘good’?
Why didn’t the VLF HIPOT fail the PD sites in ‘A’ phase and why did
additional sites appear?
When you have a cable with multiple joints and cable defects, how do you
know the time it will take the joints or terminations to fail? Are cable
defects the same as accessory defects? Which defect has been
aggravated closer to failure?
IMCORP- Power Cable Reliability
179 Middle Turnpike ● Storrs, CT 06268 ● Tel: 860.427.7620
www.imcorptech.com
In Japan, the VLF HIPOT time is 15 minutes. In Germany, the standard
time is an hour. In the USA, the time varies from 15 to 45 minutes
depending on the utility. Is the withstand time for a VLF HIPOT a
subjective number?
If VLF is harmless. then why did the PD inception voltage decline to a
more severe level?
A recent 3 year study on 30 year old XLPE cable using non destructive PD
testing techniques (similar to a factory standard test) demonstrated that up
to 50% of the cable can have multiple defects (PD sites in cable insulation,
most associated with water trees). IEEE 400.2, the VLF point document,
advises against using a VLF HIPOT on cables with multiple defects and
states that, “When testing cables with extensive water tree damage or
partial discharges in the insulation, VLF withstand testing may not be
conclusive.” How can the operator determine if the cable has multiple
defects sites before conducting a VLF test?
Could it be that different defects have different growth rates, contrary to
the work done in laboratories with ‘water needles’ and extruded materials?
Will these defects be aggravated by the HIPOT withstand only to fail at a
later date?
The above questions are rather rhetorical. The take-home message from this
case is, as the industry claims, the VLF HIPOT might be a good replacement for
the DC HIPOT; however, the user should understand that HIPOT withstand does
not predict future performance and the detrimental effects a HIPOT withstand
test can have on cable system must be considered before application. If you are
interested in determining the actual condition of your cable, off-line partial
discharge diagnostics is the only available proven method.
References
IEEE 400- Standard Guide for Field Testing and Evaluation of the Insulation
of Shielded Power Cable Systems. (For PD testing, IEEE 400-2001).