You are on page 1of 7

265

CABLE TESTING AND CABLE FAULT LOCATING


WITH
MINIMUM RISK TO GOOD CABLE
Hans R. Gnerlich, Ph.D.

HDW Electronics, Inc.


5897 Colony Drive
Bethlehem, PA 18017

ABSTRACT

The introduction of cables with solid dielectric insulation and of modem


splicing technology has imposed new standards and restrictions on cable
testing and cable fault locating. The arbitrary use of high voltages and
energies during DC,AC and impulse testing of in-service power cables
with solid dielectric insulation frequently programs the cables with
defects which become faults after the cables are returned to service. The
inadequacy of DC testing in determining the cables' AC breakdown
smngth, the danger of programming the cables with faults when DC and
impulse testing at unnecessarily high voltages, ar~dthe advantages of very
low frequency (VLF) testing and arc reflection methods (ARM)in
locating cable defects and faults are presented.

INTRODUCTION

To guarantee a customer's uninterrupted electrical service, many utilities


use various cable testing and fault locating procedures and devices
without realizing or understanding the potential danger of damaging their
good cable with inadequate techniques and equipment. The only expla-
nation for this is a lack of awareness of present-day methods, which will
produce results faster, cheaper, more realiably and with less risk to good
cable.

I i
CABLE TESTING

The HV test of an in-service cable is a GO -- NO GO test. The voltage


is applied for several minutes to an hour and one waits for a leakage
current rise, partial discharge, breakthrough, flashover, or nothing to
happen. The cable is declared good or bad depending on the outcome of
this test. However, the suitability and effectiveness [l] of such a test
depends on several criteria:

A. Is it possible to detect a defect and at what voltage level?

B. Is it possible to miss a potential cable fault, and will it


become a problem when the cable is returned to service?

C. Is there danger of generating a process which would


become a fault when the cable is returned to service?

In the late seventies, it became apparent that DC testing and fault locating
at unnecessarily high DC voltages generated cable defects which caused
cables to fail prematurely after being returned to service, while cables not
subjected to any DC voltages enjoyed a longer lifetime.

The excellent work of Groenefeld [2] and von Olshausen [3] proved that
due to space charge buildup around defects in PE and XLPE insulation:

1. DC testing is inadequate in determining the AC breakdown


strength of defective in-service power cables with solid
dielectric insulation.

2. DC testing, as well as thumping at unnecessarily high


voltages, will pre-program in-service power cables with
solid dielectric insulation for future faults.

In contrast, Very Low Frequency (VLF) testing recognizes existing


defects and faults without jeopardizing the cable's 60 Hz reliability.

But why is DC testing dangerous to the integrity of the cable's insulation?


267

When a DC voltage is applied to a cable, a space charge will develop


around a defect. It is homopolar and decreases in concentration with
distance from the defect. The electric field strength, i.e. the potential
gradient, between defect and immediate area in front of the defect
decreases dramatically compared to a space-charge-free area. The space
charge is shielding the defect. To obtain onset of partial discharge or
breakthrough at the defect, a much higher externally applied DC voltage
is required.

The space charge surrounding a defect in an insulator, is trapped and


cannot change rapidly. When a traveling wave, transient, or externally
applied AC voltage changes the defect's polarity with respect to the
surrounding space charge, a large potential gradient results. The electric
field strength becomes so great that most often it exceeds the breakdown
field strength of the material. Partial discharge occurs and a new cluster
of electrical trees is grown on every voltage flashover.

It is important to understand that damage to the cable is most often done


at the end of the step voltage test when a hard grounding takes place or
when AC is switched in before the cable had time to dissipate the space
charges.

The viciousness of the DC test lies in the coverup of possible cable


problem areas and in the deterioration of cable weak spots, causing the
cable to fail sometime, often months, after being returned to service.

With the unsuitability of DC testing for extruded in-service power cables


established, 0.1 Hz,60 Hz,and resonance testing found renewed atten-
tion. A thorough investigation of advantages and shortcomings of 60 (50)
Hz and 0.1 Hz testing was conducted by Groenefeld [2].

For PE and XLPE insulation, Groenefeld determined that at a constant


test voltage level partial discharge starts approximately ten thousand
times faster at 60 (50) Hz than at 0.1 Hz and that breakthrough of a cable
defect happens much faster at 0.1 Hz than at 60 (50) Hz. He also found
that raising the test voltage level increases the size of a defect at 60 (50)
Hz without breakthrough occurring, but it exponentially decreases the
testing (breakthrough) time at 0.1 Hz.
268

Realizing the implications of these findings, Lefeldt developed 141 and


patented [5] a VLF cable test set. It generates aO.l Hz pulse wave which
changes polarity sinusoidally. The sinusoidal transitions in the 60 Hz
range quickly initiate a partial discharge at a defect which the 0.1 Hz pulse
wave develops into a breakthrough channel at optimum speed. Due to
sinusoidal transitions between the HV pulses, traveling waves are not
generated, and due to continous polarity changes, dangerous space
charges cannot develop. Cables can be tested as required or in preventive
maintenance programs, identified faults can be repaired immediately and
no new defects will have been initiated during the:testing process. When
a cable passes the VLF test, it can be returned to service without
reservation.

At this time, the VLF cable test set seems to be the only device which
permits testing of in-service cables with solid dielectric insulation effi-
ciently without adversely affecting the cable's life expectancy.
CABLE FAULT LOCATING

Many utilities are experienced in locating short and open circuits on


primary or secondary cables. Proper locating [6] of high resistance or
intermittent cable faults which are the majority of the faults, is considered
tedious, unreliable, and time consuming. Therefore, reclosing, burning,
and thumping at extremely high voltage and energy levels, in order to
generate a short or open circuit, are frequently used in ignorance of cable
properties. The danger of activating dormant faults or generating new
faults by these improper locating methods is ignored. Management is, for
the most part, not aware that fault locating is costing the utility hundreds
of thousands of dollars in prematurely failed cable and its replacement.

The introduction of TDRs (cable radar) for cable fault locating, about 35
years ago, has had a lasting impact on all forms of fault locating practices.
TDRs make it possible to actually see into a cable to locate faults. pulses
of short time duration are transmitted through the cable. On the
horizontal and vertical axes of a cathode ray tube (CRT) screen, the
elapsed time of a pulse traveling the entire length of a cable and the pulse
reflections produced by the deviations from the homogenous structure of
the cable are displayed, respectively. Any reflecting surface, cable start,
joints, splices, transformers, faults, changes in cable type, as well as cable
269

end, are shown in time sequence.


When the propagation velocity of a pulse through a cable is known, the
distance between cable start and any discontinuity or irregularity can be
determined from the reflection - time display. In addition, the use of
state-of-the-art electronics makes it possible to provide a digital readout
of the distance to the fault, as well as cable length.

As atractive as the TDR method is, it has limitations. Cable faults


between center conductor and shield with resistance values of greater
than ten times the characteristic cable impedance have small fault reflec-
tions and cannot be distinguished from reflections of naturally occufing
irregularities.

With an impulse generator, high resistance or intermittent cable faults can


temporarily be converted into faults having resistance values equal to or
less than the characteristic impedance.

Combining a TDR and an impulse generator (thumper) allows locating


any type of cable fault. In HDW Electronics' Arc Reflection Method
(ARM),an impulse generator and a TDR are simultaneously connected
to a cable having a high resistance or intermittent fault. An energy
separation filter isolates the TDR from the HV impulses and ensures that
the high frequency test pulses sent into the cable by the TDR are not short
circuited by the impulse generator.
During the first phase of the measurement, the TDR pulses are not
reflected by the high resistance or intermittent fault and only cable start,
joints, splices, transformers, irregularities, and cable end are clearly
visible.
In the second phase, the impulse generator is switched on. The impulse
amplitude is only made sufficiently high to break down the fault and
generate arcing at the fault location.

Any hit of the arc by a TDR pulse will produce a reflection, hence the
name "Arc Reflection Method", and an image of the temporary low
resistance fault, a negative blip, will indicate the fault location on the
CRT. Once arcing ceases, the fault reverts back to its high resistance
state.
270

A real time comparison of the cable with and without HV applied is


observed. During the intervals between arcing, when the impulse
generator is in the charge mode, the reflected image of the cable, start to
end, is displayed with all inherent cable "landmarks". During arcing, the
high resistance fault is converted to a low resistance state and the negative
blip is overlaid on the low voltage display. In this real time display the
fault location is easily determined, not only as a distance in meters (feet)
from the beginning or end of the cable, but also in relation to the other
landmark reflection points.

ARM permits locating of faults in power cables at lowest possible HV


levels quickly, accurately, and reliably with minimum risk to good cable.
It is unique in that it combines previously used methods while eliminating
their limitations. Application of ARM is not limited by cable length or
type, nor by operating voltage range. Component voltage specifications
and cable integrity are the only considerations to be made when using
ARM for a particular fault locating problem.

SUMMARY

By establishing cable fault locating and cable testing guidelines for


cables with solid dielectricinsulation, which incorporate current research
and available technology, utilities can realize substantialsavings in cable
replacement, cable rehabilitation, and manpower budgets.

REFERENCES

[l] Gnerlich, Hans R. "Effects of High Voltage Testing and Fault


Locating on Power Cables with Solid Dielecritc Insulation",
High Voltage Measurements, Test, and Applications
Workshop, Electrical Insulation Research Center, University
of Connecticut, December 7 - 8, 1988.

[2] Groenefeld, Peter, "Elektrische Pruefung von Kunststoffen mit


Spannungen Unterschiedlicher Form", Doctoral Dissertation,
University of Hanover, W-Germany, 1986.
27 1

[31 von Olshausen, Rainer "Verfahren zur Kabelpruefung:


Gleichspannung - und VLF - Pruefung and Masse - und
Kunststoffkabeln", Seminar on Cable Testing, E e l ,
W-Germany, December, 1987.

[4] Lefeldt, Ekkehard, "VLF - Pruefanlage der HDW Elektronik,


GmbH', Elektizitaetswirtschaft,JG 86(87), H 13, pp 581-583.

[5] Lefeldt, Ekkehard, "Testing Device for Voltage Testing of


Cables and Cable Sets", United States Patent, Number
4,825,171, April 25, 1989.

[6] Gnerlich, Hans R., "Cable Fault Locating with Minimum Risk
to Good Cable", 1lth Annual NETA Technical Conference,
March 1989.

You might also like