What is the purpose of VLF testing?

What is VLF?
VLF is the abbreviation, and commonly used name, for Very Low Frequency. VLF AC hipots produce a frequency output of
0.1 Hz and lower rather than the conventional 50 Hz or 60 Hz. VLF technology was developed for a specific reason, to more
easily high voltage field test certain loads of high capacitance normally requiring very high current and power to test at the
traditional power frequencies of 50 Hz. or 60 Hz.

Why 0.1 Hz?
When using AC high voltage to test a load, the lower the frequency of the applied voltage the lower the current and power
required. The easiest and most economical way to AC high voltage field test high capacitance loads, like cables and
motors/generators, is to use a VLF AC hipot. A frequency of 0.1 Hz. was selected long ago as the standard frequency to be
used for VLF testing. World standards permit, and most VLF instruments deliver, output frequencies of 0.1 Hz. - 0.01 Hz.
For cable testing, 0.1 Hz – 0.01 Hz. is permitted, although 0.1 Hz. and 0.05 Hz. are preferred for use. For motor and
generator coil testing, 0.1 Hz. is required. For some other testing using VLF, explained later 0.1 Hz. is required.

The Math behind VLF Testing
There is no mystery to VLF technology. It conforms to basic physics and established principles of electricity. The most
elemental equation governing the laws of electricity is Ohm's Law: I = V/R. The higher the denominator R (resistance), the
lower the I (current) when a V (voltage) is applied. In AC circuits, the load is mostly capacitive. That means that the R of a
circuit is directly proportional to its operating frequency. R is the Capacitive Reactance, Xc, and calculated using 1/𝜔C,
𝜔 = 2𝜋f and C is the fixed capacitance of the load. Xc (capacitive reactance) = 1/2πfC.
Example: A 15 kV cable 5,000' (1500 m) long has approximately 0.5 uF of capacitance. At 60 Hz. the capacitive reactance is
5300 ohms. At a common test voltage of 22 kVac rms, it would require a power supply rated for 4.2 amps, or 91 kVA. A
very large and expensive test set; obviously not very practical for field use.
At 0.1 Hz. the capacitive reactance is 3.2 megohms. The same 22 kV rms test would draw 7 mA, or .154 kVA, 600 times less
than at 60 Hz. At 0.01 Hz, a cable 6000 times longer can be tested than at 60 Hz.
Put another way, at 60 Hz. a cable must be charged to its test voltage every 4.2 milliseconds: 0 – 90° (the peak) of the
waveform. At 0.1 Hz, 0 – 90° takes 2.5 seconds, permitting 600x more time to charge the cable, requiring a far less powerful
voltage source.

Is 0.1 Hz still AC?
Yes. The wave shape of the HVI VLF design is sinusoidal with polarity reversals every half cycle, only at a slower rate than
60 Hz. Frequencies as low as 0.01 Hz are recognized in the IEEE 400.2-2013 standard as useable for cable testing, although
typically 0.1 Hz. – 0.05 Hz. are preferred. VLF is not DC, where a monopolar negative polarity is applied to the load for long
periods of time, which causes space charges to develop within the insulation through a polarizing dipole effect on the
molecules and a stored energy to develop within the load. Again, VLF is an AC (alternating current) voltage, sinusoidal with
conventional polarity reversals every half cycle.

Where is VLF used?
VLF testing is principally used for two applications: AC field testing medium and high voltage cable and testing rotating
machinery - motor and generator coils. These two applications are defined and sanctioned by several world standards,
including IEEE 400.2-2013 for cable testing and IEEE 433-2012 for testing rotating machinery. VLF can be used for testing
other high capacitance loads like large insulators, arrestors, bus duct, etc. but no recognized standards exist.
One of the best applications for the use of VLF is to check installation quality of cable and accessories, like splices and
terminations. Many in service failures are due to damage to cable during its installation, improper workmanship, faulty
materials, etc. These in service failures can be prevented through VLF Acceptance testing after installation and after every
cable fault repaired. For motor and generator coil testing, producers of these products must factory test their coils with
50/60 Hz. power frequency AC Dielectric test sets, or hipots. However, rewind and repair shops and field maintenance
testing can use VLF and should due to the size, weight, and price advantage over the use of prohibitively large, heavy,
expensive, and difficult to set up 50/60 Hz. hipots.

What VLF hipots are available?
High Voltage, Inc. produces VLF hipots that produce from 30 kVac up to 200 kVac with load ratings from 0.4 uF to 50 uF,
the equivalent capacitance ratings of cable approximately 3000' (914 m) to 40 miles (64 km) in length and can test the
largest of generator coils. Both manual controlled, conventional oil cooled designs with analog controls (knobs, meters,
switches, etc.), and fully programmable, automatic, wireless, PC controlled solid state designs are available

How do you do the test?
The test is very simple. With the cable to be tested isolated from any voltage source, connect the high voltage output lead
of the VLF to the conductor and a common ground to the shield. Like any hipot, apply the test voltage for the required
duration. A basic withstand test is that easy. Other diagnostic tests, yet to be described, are a little more complicated.
What's the test voltage and for how long?
The IEEE 400.2-2013 standard offers precise test voltages for medium voltage and some high voltage cable. (Generally
however, the test voltages are approximately 1.7 - 3 times (1.7 Uo - 3 Uo) the normal line-to-ground voltage for 30 – 60
minutes, with the multiple number depending on the voltage rating and thickness of the cable insulaiton.) A chart of the
test voltages is below. The European standard mandates 3 Uo (3 times the normal line-to–ground voltage) for 60 minutes
for any cable. Other countries have also written standards for VLF cable testing.
For a 15 kV cable, the Maintenance test is usually performed at 22 kVac peak and the Acceptance test voltage is 30 kVac
peak. A 35 kV cable is Maintenance tested at 47 kVac peak and Acceptance tested at 62 kVac. The standard includes cables
rated to 69 kV, although 200 kVac peak VLF units are available for testing cables rated up to 150 kVac.



Different VLF units output different waveforms. What's best?
All HVI VLF units, and most others, produce a sine wave output. The original German designs, which are still offered, do not
produce a sine wave output. They produce a trapezoidal, or cosine-rectangular, waveform. The cosine-rectangular
waveform works well to VLF hipot cable; however, it is not as usable as a sinusoidal design as a voltage source for Tan Delta
and Partial Discharge testing. For a VLF unit to be used for diagnostic testing, either Tan Delta of Partial Discharge, it should
produce a sine wave. The IEEE recognizes the sine wave output as advantageous and mandates it when VLF is used for
testing rotating machinery. Stick with a sine wave design to keep your future diagnostic testing options open.

Is the VLF test destructive?
VLF hipoting is not destructive to good insulation and does not lead to premature failures like DC voltage testing. Using
VLF does not cause degradation of good insulation nor aggravate defects too small to be triggered into PD under the test
voltage. It does cause existing cable defects that are severe enough to be triggered into partial discharge under the test
voltage, to break through, or fail, during the test. Minor defects that are not triggered into PD from the test voltage are
unaffected. If a cable can't hold 2 – 3 times normal operating voltage, it is not a reliable cable. Cause failure at the defect
location during a controlled outage or prior to energizing newly installed or repaired cables, find the fault, make the repair,
and be left with a good cable. It is AC voltage; what the cable is designed for and experiences during service. Cable is
factory tested with AC voltage at levels far higher than the field test.
But my cable might fail during the test
Precisely, that is the point of AC Withstand testing. It is not a diagnostic test. It is an AC stress test, or proof or pressure
test. There are no leakage current readings to measure. A cable either holds the test voltage or fails. If this method of
testing is not acceptable, there are diagnostic tests that can be performed that nearly eliminate the chance of cable failure
during the test. These tests allow the user to learn something of the cable insulation rather than possibly cause a failure
during the test. See the Tan Delta and Partial Discharge sections of this FAQ

Who endorses VLF?
Nearly every applicable engineering body in the world, cable producers, and the hundreds of utilities worldwide that use
the over 5000 VLF units shipped by HVI and others over the past 20 years. EPRI, IEEE, IEC, CEA, VDE, other countries
engineering organizations, nearly every cable manufacturer, and many utilities throughout the world have embraced the
effectiveness of VLF testing. German VLF test standards (DIN-VDE Standard 0276-620 & 0276-1001) have existed for many
years and the IEEE has released an updated VLF specific cable testing standard - IEEE 400.2-2013. IEEE 433-2012 for VLF
testing of rotating machinery has existed for over 40 years, originally released in 1974.

What are the alternatives to VLF Withstand testing cables?
Not many, when you consider the available technologies and weigh the costs, effectiveness, ease of use, and other factors.
A 50/60 Hz. power frequency AC Withstand test is not usually an option, as described earlier. Certainly DC should no longer
be used since it leads to future cable failures and tells little about the cables insulation and accessories quality. One can
reduce the DC test voltage, but then the test is even less meaningful. A 5 kVdc megohmmeter IR test on a 15 kV cable that
operates at over 10 kVac peak stress when in service reveals little or nothing about a cable's quality, nor does a 24 hour no
load on-line soak test. There are several "diagnostic" tests possible but many are experimental, esoteric in their theory and
design, and not economically feasible; except for off-line Tan Delta and Partial Discharge testing, described below.

VLF Diagnostic testing methods
There are times when a go/no-go, or pass/fail Withstand test is not suitable. We would rather learn something about the
health of the cable system without risk of failure during the test; a test that measures the quality, or integrity, of the
insulation and its expected life is preferred. A VLF hipot can be, and often is, used as a variable voltage source to apply an
overvoltage to the test object while various measurements and observations are made about the insulation quality.
Two common methods of using VLF technology to perform off-line elevated voltage diagnostic, or non-destructive, testing
on cables and rotating machinery are Tangent Delta and Partial Discharge. Both use a VLF instrument to apply a variable
voltage to the load while diagnostic measurements are taken. Based on these measurements, the voltage can be raised up
to perhaps 1.7 Uo to 2.0 Uo the normal operating voltage, but only for very brief periods, perhaps only seconds. From the
data gathered, judgments can be made as to the integrity and expected life of the insulation.
Tan Delta, or Tan δ, Dissipation Factor (analogous to power factor), testing is performed to provide an overall assessment
of the integrity of the insulation, usually compared to the ideal or to many other tests to prioritize where replacement,
repair, or rejuvenation money should be spent. This method works well, is easy to perform with minimal training, and is
relatively easy to interpret results. TD testing is a very common test performed by many.
Partial Discharge testing exposes specific locations and severity of troublesome electrical discharges along a cable path or
the overall electrical noise (PD) within a coil. PD testing is more difficult to perform and interpret than Tan Delta and is
more expensive. However, both are valid but reveal very different data sets. PD testing is commonly performed on cables,
substation apparatus, and other gear but at 50/60 Hz. PD testing cables and generator coils using 0.1 Hz. is a newer

technology but no less possible or reliable than at power frequency.
The surest way to weed out bad cables and accessories is to perform a simple, over voltage AC hipot test, just like we do
with vacuum bottles, arrestors, hot sticks, switchgear bus, insulators, etc. Yes the cable may fail under test if it has a severe
defect, but that's the point of the test. If a cable can't withstand 2 - 3 times its normal voltage for 30+ minutes, it's bound
to soon fail. Cause failure when convenient to repair, rather than waiting for an in-service failure to occur at the worst
possible time. If preferred, VLF voltage sources can be used to perform non-destructive diagnostic testing, like the Tan
Delta/Power Factor and Partial Discharge methods described earlier.

Summary
Very Low Frequency technology is readily available and has been used for decades to AC voltage test cables and rotating
machinery in the field that was never before practical, nor even possible in many cases. IEEE, IEC, and other Standards exist
to define the testing and interpretation of results. Overvoltage AC Withstand and Diagnostic testing methods are readily
available in voltages up to 200 kVac peak from several vendors worldwide.

Source :

Practical Applications of Very Low Frequency (VLF) Testing

Note 1

If the operating voltage is a voltage class lower than the rated voltage of the cable, it is recommended that the maintenance test voltages should be those corresponding to the operating voltage class.
Note 2

The maintenance voltage is about 75% of the acceptance test voltage magnitude.
Note 3

Some existing test sets have a maximum voltage that is up to 5% below the values listed in the table. These test sets are acceptable to be used. However, there is a risk that the cable may be "understated" due to a combination of lower test voltage and allowed uncertainty of the measuring circuit.

VLF ac voltage testing methods utilize ac signals at frequencies in the range of 0.01 Hz to 1 Hz. The most commonly used, commercially available VLF ac voltage test frequency is 0.1 Hz. VLF ac test voltages with cosinerectangular and the sinusoidal wave shapes are most commonly used. While other wave shapes are available for testing of cable systems, recommended test voltage levels have not been established.


5.1.1 VLF ac withstand voltage test parameters

The purpose of a withstand test is to verify the integrity of the cable under test. If the test cable has a defect severe enough at the withstand test voltage, an electrical tree will initiate and grow in the insulation. Inception of an electrical tree and channel growth time are functions of several factors including test voltage, source frequency and amplitude, and the geometry of the defect. For an electrical tree from the tip of a needle in PE insulation in laboratory conditions to completely penetrate the insulation during the test duration, VLF ac voltage test levels and testing time durations have been established for the two most commonly used test voltage sources, the cosine-rectangular and the sinusoidal wave shapes. However, the time to failure will vary according to the type of insulation such as PE, paper, and rubber. Thus the electrical tree growth rate is not the same for all materials and defects.

The voltage levels (installation and acceptance) are based on the most used, worldwide practices of from less than 2 U0 to 3U0, where U0 is the rated rms phase to ground voltage, for cables rated between 5 kV and 69 kV. The maintenance test level is about 75% of the acceptance test level. One can reduce the test voltage by another 20% if the voltage is applied for longer times (Bach [B2]; Baur, Mohaupt, and Schlick, [B6]; Krefter [B27]). Evidence (Hernandez-Mejía, et al. [B21]) indicates that increasing the voltage above 3U0 to compensate for reduced test cycles (time) does not replicate performance either on test or in service as compared to the lower voltage, longer time tests.

Table 3 lists voltage levels for VLF withstand testing of shielded power cable systems using cosine-rectangular and sinusoidal waveforms (Bach [B2]; Eager, et al. [B9]; Krefter [B27]; Moh [B28]). For a sinusoidal waveform the rms is 0.707 of the peak value, assuming the harmonic distortion is less than 5%. The rms and peak values of the cosine-rectangular waveform are assumed to be equal. It should be noted that terminations may need to be added to avoid flashover for installation tests on cables rated above 35 kV. Regarding the test times:

The recommended minimum testing time for a simple withstand test on aged cable circuits is 30 min at 0.1 Hz (Goodwin, Oetjen, and Peschel [B13]). If a circuit is considered as important, e.g., feeder circuits, then consideration should be given to extending the testing time to 60 min at 0.1 Hz (Hampton, et al. [B19].

The recommended minimum testing time for an installation and/or acceptance withstand test on new cable circuits is 60 min at 0.1 Hz.

A test time within the range 15–30 min may be considered if the monitored characteristic remains stable for at least 15 min and no failure occurs. It should be noted that the recommended test time for a withstand test is 30 min.

What is a Very Low Frequency Withstand Test?

A Very Low Frequency (VLF) Withstand Test is an AC Withstand Test usually carried out at a frequency between 0.1 Hz and 0.01 Hz. This kind of test is suited for testing high capacitance loads such as cable and rotating machinery. This is a pass fail test i.e. bad cable will fail during testing rather than in service.

The Theory

Basic electrical theory: Xc = 1/(2*pi*f*C)

Where:

Xc = Capacitive reactance (Ω)

f = Frequency (Hz)

C = Capacitance (F)

It can be seen that capacitive reactance, which is the resistance across the power supply output, is inversely proportional to frequency. By reducing the frequency the capacitive reactance is increased.

More basic electrical theory: I = V/R

Where:

I = Current (A)

V = Voltage (V)

R = Resistance (Ω)

The lower the frequency, the higher the capacitive reactance (Xc). The higher the Xc the lower the current and power needed to apply a voltage.

Example:

A length of power cable with capacitance of 1 µF needs to be tested at 34 kV peak.

Power frequency testing (50 Hz):

Using the above formulae it can be calculated that the capacitive reactance would be 3.183 kΩ so the required current would be 10.68 A. Therefore, to test a 1 µF cable at a frequency of 50 Hz the test set would have to be able to give 363 kVA. When you consider that a 30 kV, 40 kVA AC test set can greatly exceed 620 kg it puts into perspective the size an AC test set would have to be to produce 363 kVA.

VLF Testing (0.1 Hz)

Testing the same length of cable at 0.1 Hz would generate 1.59 MΩ of capacitive reactance and the required current would be 21 mA. Therefore testing the same length of cable at 0.1 Hz would require 0.714 kVA, which is 500 times less than at a power frequency of 50 Hz.


Most test sets have a frequency range of 0.1 Hz to 0.01 Hz and the most desirable test sets will select the optimum frequency based upon the capacitance of the cable. The capacitance of the cable is dependent upon its construction materials and its length. As cable length increases so does the capacitance so dropping the frequency allows kilometres of cable to be tested.

Does this difference in test frequency range have any effect on test results?

This was reported on in the National Electric Energy Testing, Research and Applications Center (NEETRAC) report "Estimating the Impact of VLF Frequency on Effectiveness of VLF Withstand Diagnostics" by N. Hampton et al (2014). This report concluded "... there is no distinguishable difference between failure rates on test for the common VLF test frequencies of 0.05 Hz and 0.1 Hz, from data obtained through laboratory and field tests, and all insulation types."

What size VLF test set is required?

Acceptance test voltages are generally 2.5 – 3 times the line to ground system voltage. A list of field test voltages from 5 kV to 69 kV can be found in IEEE 400.2.

How long should a VLF test last?

According to IEEE 400.2 VLF tests should last between 15 and 60 minutes with a recommended minimum duration of 30 minutes. This length of time hasn't just been plucked out of the air; there is some theory behind it.

According to IEEE 400, when testing at 3 times the line to ground system voltage the tree growth rate of XLPE at a 0.1 Hz sinusoidal test voltage is 10.9 – 12.6 mm/h.

A 15 kV 133% cable has an insulation thickness of 5.9 mm. Therefore in a 30 minute test nearly all defects will grow to failure.

What is a Tree?

At stressed points in insulation where there are protrusions, voids or contaminants a phenomenon called treeing may occur. Treeing is the preferred name due to the branch like structure of these trees. There are two types of tree effect:

    Water tree: Produced by electro-oxidation fuelled by electrical stress and water ingress within insulation. Water trees do not emit partial discharges so PD testing cannot be used to test for the presence of water trees. During service conditions the growth of water trees is extremely slow taking years to completely penetrate the insulation. Like protrusions, voids and contaminants, water trees act as stress enhancements. They can increase the local electrical field and also create local mechanical stresses. If these electrical and/or mechanical stresses are high enough an electrical tree can initiate.
    Electrical tree: Micro-channels of carbonisation or non-insulation found within insulation that is irreversibly damaged. Electrical treeing will lead to a completed failure path and failure of insulation relatively quickly. Below is an electrical tree that is growing from a water tree.


Why not just test at DC?

DC has been used to test paper insulated lead covered (PILC) cable for many years. Therefore it is understandable that DC continued to be used as solid dielectric cable became prevalent. Unfortunately the first solid dielectric cable started to prematurely fail 15 – 20 years after being installed. After extensive testing and research it was found that solid dielectric cable is prone to develop water trees. DC testing at high voltages creates negative space charges within these trees, as highlighted below.

When the testing is complete and the DC is removed these trapped negative space charges remain. When the AC is reapplied a high difference in potential exists at points in the insulation where these trapped negative space charges are found. These areas that are already suffering from heightened electrical and mechanical stresses are prone to turn into electrical trees. Once this point has been reached the cable will fail because the damage is irreversible.

DC test sets cannot be used as a voltage source for Tan Delta or Partial Discharge testing. These diagnostic tests are usually carried out at a frequency of 0.1 Hz

source: You are not allowed to view links. Register or Login