Power Cable Failure Modes and Fault Location Methods, Practices and Strategies

Earle C. (Rusty) Bascom, III, Martin J. von Herrmann Dr. Tiebin (Tom) Zhao
Senior Member Member Senior Member
Electrical Consulting Engineers, P.C. The VON Corporation Electric Power Research Institute
Schenectady, New York Birmingham, Alabama Charlotte, North Carolina

Abstract  During the last 140 years, insulated cables have

been applied for power transmission and distribution, as well as
signaling and other purposes. While the basic functions of power
cables remain unchanged – providing a conductor to carry
current and a dielectric insulation to support the line-to-ground
voltage – the materials and characteristics used in power cables
have evolved. Cables experience faults, and some methods to find Figure 1: Illustration of major cable components (conductor, insulation)
those faults have changed as technologies have developed.
Classical bridge techniques are still used, often with improved Joints (splices) join sections of cable together and must
null detectors, but modern methods such as time domain restore the cable conducting path(s) and insulating
reflectometry and faulted circuit indicators provide novel characteristics, and terminations must transition the cable to
methods for better accuracy and faster location. This paper other equipment by maintaining conducting path(s) and
summarizes recent work in both areas and provides background transition insulating properties to other equipment.
on cable failure modes and typical fault location methods. When either the conductor or insulation is disrupted in its
I. INTRODUCTION basic function, the cable system is understood to have
experienced a fault that must be located.
Dating back to Thomas Edison’s first cable installations in
the 1870s in New York City, power cable fault location has B. Cable System Installation Characteristics
been a necessary part of operating and maintaining Cable circuit characteristics must be well understood to
underground cable systems as failures do occur, and owners evaluate applicable fault location methods. Some fault
and operators of cable systems are continually looking for location methods benefit from comparing faulted to unfaulted
improved methods – either in terms of accuracy or speed – to cables or require a return conductor. Therefore, the number of
perform fault location. Many times, the most successful fault cables installed (1, 2 or 3 phases, presence of separate neutral,
location efforts are achieved with a mix of classical fault etc) must be understood. The circuit length must be known as
location methods and new technologies, and the most equipment may be limited by the length that may be tested.
knowledgeable technicians keep as many options available as Sometimes tests applied from both ends may be helpful.
possible. Fortunately, there has been much progress in fault Cable splices have different impedance characteristics than
location technology from the early “cut and try” and “re- the cable itself so these may affect fault location test results,
fusing” methods that resulted in multiple splices and additional either by displaying a variation in the cable characteristic,
cable damage. blocking a portion of a test signal, or indicating that multiple
As with many skills, attrition of personnel has meant that cable types or constructions are connected in series, possibly
much of this expert knowledge has left the industry. from a previous cable repair. Variations in cable construction
Organizations have put together detailed references to assist affect resistance, capacitance and propagation velocity for test
the utility industry in retaining this knowledge, and suppliers signals that may be applied to the cable. Consulting utility or
of fault location equipment incorporate novel features in their owner records may help in evaluating these conditions.
equipment to aid field technicians in locating cable faults. This Cable construction can have an impact on fault location. A
paper summarizes methods, practices and strategies commonly concentric shield is helpful for some methods (e.g., TDR) but
applied to fault location efforts [1,2,3,4]. may block others (tone or tracer sets), though a jacket fault can
II. CABLE SYSTEM CHARACTERISTICS be located on cables with a metallic shield, sheath or
concentric neutral. The metallic shield may also be used as a
A. Fundamental Cable Functions
return conductor for some fault location methods. Inconsistent
The two main components of power cables are the characteristics (corroded, different cable types, etc.) can also
conductor and insulation, with other cable features (sheath, affect fault location.
jacket, pipe, armor, etc.) serving to maintain the functions of Pinpointing a fault on cables in conduits or cable pipes can
these basic components. The conductors (at the core, or be more difficult because the conduit/pipe blocks some of the
concentric in the metallic shield, sheath or neutral) carry signal or effectively creates higher impedance for some fault
current and generally have a resistance that is consistent and location methods. Acoustic or earth gradient methods can be
proportional along the length of the circuit unless there is an more difficult to interpret because the apparent source (sound
unintended interruption (break). or signal) is translated to the ends of a conduit segment which
The cable insulation (dielectric insulator) supports the may be some distance away from the actual fault location.
nominal line-to-ground voltage (and phase-to-phase voltage in Direct buried cables may provide better results for tracer
multi-core cables) and should withstand transients, impulses or methods. Increased burial depth affects successful application
surges that may be experienced by the cable during normal of tracer methods.
operation.

978-1-4799-3656-4/14/$31.00 ©2014 IEEE

 Pipe lines, other cables or any other metallic system could IV. FAILURE MODES
disperse measured signals that may complicate fault locating Cable circuit failures, including splices and terminations,
efforts. can result from one or more causes that may be categorized
Terminal fault location methods have generally equal based on the mode in which the failure occurs.
effectiveness on direct buried, cables in conduit or pipe,
submarine or aerial cables, particularly if the cables have a A. Manufacturing Failure Modes
concentric metallic shield. Manufacturing defects occur in the factory when the cable
Documentation of plan and profile information, including or major accessories are made. Factory acceptance tests are
route maps, are often much more comprehensive for intended to detect these defects before the cable is shipped and
transmission versus distribution cable systems, so simply installed. However, the cable may pass these tests and still
locating the cable and route may be easier for transmission experience an in-service failure that must be located. Most of
although circuit lengths may be substantially longer. the manufacturing defects relate to the insulation system on the
Transmission circuits are often networked so while a failure on cable and are summarized here:
these system types may affect system load flows, individual
customers are less likely to be affected as compared to  Voids – Locations in the extruded insulation that are not
distribution circuit outages. Restoration time for transmission completely filled with extrudate. Ionization occurs at the
cables is usually measured in days to weeks while distribution voids that can lead to electrical trees and failure.
cable outages are often completed in hours to days.  Contaminants – Impurities or particles in the extrudate
create sites for ionization that weakens the dielectric
C. Cable System Bonding Considerations
strength of the insulating layer.
Sheath bonding on cable systems can affect the application  Protrusions – These are locations where the conductor
of fault location methods. Fully bonded (multiple-point shield or insulation shield protrude into the insulation,
bonded) cables have the cable concentric neutral/shield/sheath resulting in areas of high electrical stress.
grounded at the terminals and usually at intermediate splice  Eccentricity – If the conductor is not coaxial to the
points. This means that the neutral, shield and sheath are insulation, portions of the insulation wall thickness are
continuous and remain concentric to the phase conductor(s). reduced potentially causing higher electrical stress. The
Transmission cables have cross-bonding or single-point main concern with high eccentricity is successfully
bonding connections that often must be reconfigured to interfacing the cable to accessories (splices, terminations).
perform fault location, or the fault location effort must be  Dry Cable – Paper-insulated cables are taped with dry
focused on individual sections rather than the entire circuit. kraft paper or laminated paper-polypropylene tapes that are
III. FAULT TYPES then saturated with dielectric liquid for varying durations
in large impregnation tanks. If the impregnation time is
A. Shunt Fault
insufficient or pressures are not adequately controlled,
This fault type occurs when there is a breakdown in the incomplete saturation may occur that leaves dry spots.
insulation between the conductor and the concentric shield or Cables may also develop dry spots due to a combination of
between the concentric shield, neutral or sheath and ground. long transit times, high ambient temperatures or failure to
Shunt faults may be either “high” or “low” impedance, which rotate and maintain the reels while in storage. Dry cable
is usually determined if the fault impedance is, respectively, experiences waxing when exposed to voltage that can
greater than or less than the surge impedance. Bolted or short eventually lead to dendrites, tracking and failure.
circuited faults have resistances less than 5 Ohms. Other faults  Taping Defects – If cable butt gaps align between
may be characterized by varying impedance levels as a successive taping layers creating a “registration”, these
function of voltage and classified as “non-linear” faults; some may form voids in the insulation where higher electrical
faults (usually at transmission voltages) may require in excess stress can develop. Also, taping tensions may be outside
of 30kV to create an arc, exceeding most commercially specifications such that wrinkles can develop when the
available fault locators. cables are bent during handling, installation or thermal-
B. Series Fault mechanical bending while in service.
A series fault is a break or discontinuity (cut cables or open B. Construction and Installation Failure Modes
conductors) in the phase conductor (as shown) or the Commissioning tests should detect most installation related
concentric metallic shield, sheath or neutral. A series fault that cable and accessory failures, but some tests are completed
results from neutral corrosion, a deteriorated splice or successfully and still result in in-service failures.
termination, a burned conductor or a fault where water has Additionally, some cables are not tested prior to being placed
saturated the fault, the fault may display “non-linear” in service. Various factors can affect cable system installations
characteristics in that the impedance may change depending on that result in construction or installation related failures as
the applied test voltage. In some cases, a shunt fault may summarized here:
convert to a series fault as a result of excessive fault current.
 Poor Installation Environment – For all cable types but
particularly transmission cables, assembling major
accessories (joints/splices, terminations) where the
environment is not adequately controlled can result in
Figure 2: Shunt (through resistance) and series (conductor break) faults cable circuit failures. Moisture, including high humidity,
 or airborne contaminants can affect the successful sectional area of the shield may result in excessive heating
installation. as the cross-sectional area is reduced. The pipe on pipe-
 Duct (Conduit) or Pipe Integrity – In some cases, the type cable or metallic sheath on self-contained or paper-
conduits or pipes in which the cables are installed may be lead cable contain the dielectric liquid that would
in poor condition, either from factors related to the initial deteriorate in the presence of moisture or loss of pressure.
installation or due to deterioration over time. Passing a  Fluid Migration – Cable systems using impregnated paper
mandrel through the conduit or pipe before pulling cable insulation may experience migration of the dielectric fluid
normally detects problems. A jacket test on transmission resulting in dry cable as discussed earlier.
or distribution cable may pass even though the jacket has  Treeing – Electrical trees, sometimes started as water
been damaged because the conduit essentially provides trees, form in insulation and bridge the dielectric gap
additional insulation to local ground. between the conductor and surrounding shield/neutral.
 Defective Seal or Improper Assembly – Heat shrink  Hydraulic Pressure – Extreme dielectric liquid pressure
materials on splices may not be uniformly heated so there can damage pipe-type and self-contained cables. High
is inadequate or incomplete sealing to the cable jacket to pressure may cause mechanical failure of components,
prevent increases of moisture or contaminants. General while low pressure can allow ionization and eventual
problems with assembly being done inappropriately can tracking in the insulation.
lead to failures.
V. FAULT LOCATION STRATEGIES
 Other External Damage – Damage to the cable may
occur from handling of reels, improperly applied or Fault location involves contact with potentially energized
installed backfills, directional drilling that hits a cable, etc., equipment or using test equipment that may produce a voltage
all of which may fail a cable. or current, either of which may be sufficient to injure
personnel. As with any activity, safety should be a priority.
C. Mechanical Failure Modes The fundamental consideration for fault location is to first
Mechanical failure modes are those that generally result verify that a failure has occurred in the cable itself. This can
from external factors or movement that somehow disrupts the be indicated by the presence of blown fuses at either or both
cable system and results in a failure. Mechanical failure ends of a cable section or directional relays indicating the
modes include: failure is within the cable; faulted circuit indicators may also
point to an issue within the cable. Basic verification may
 Thermal-Mechanical Bending (TMB) – This results from
include a resistance test to show an open or short in the
chronic movement of cables due to load or other induced
insulation or a simple voltage withstand test with AC (VLF,
(solar, ambient air) factor that causes the thermal
50/60Hz) or DC voltage.
expansion and contraction of the cable system with
associated cable movement. Forces from cables in A. Terminal Methods and Pre-Location
conduits or pipes are translated to joints where mechanical As the name implies, terminal methods attempt to locate a
restraints (clamps, anchors, etc.) are insufficient. Some cable failure from the end connections of a cable either
cable sheath materials (e.g., lead) are susceptible to metal through terminations, elbows or other types of disconnects that
fatigue that either disrupts the electrical path or permits provide access to the cable conductor and/or the concentric
moisture ingress to the cable. metallic shield, neutral or sheath. When successful, the
 Vibration – This generally affects cables suspended on or methods only locate the cable fault approximately, so they are
under bridges where the cable sustains damage from often considered “pre-location” techniques. Common terminal
fatigue or the support system fails in some manner methods and their applications are listed in Table I.
resulting in adverse bending. Sources of vibration vary but
can include wind, vehicle traffic or precipitation. TABLE I
METHODS FOR CABLE FAULT PRE-LOCATION (TERMINAL TECHNIQUES)
 Physical Damage – This type of damage can result from Method Applicable Fault Type
various sources including “dig-ins”, particularly for direct- Bridge Measurement Low- or High-Resistance Shunt Faults
buried cables, vandalism, varmints, moisture ingress from Burner Techniques High Resistance, Very High Voltage Flashover Faults
improper or insufficient seals or poor handling during Arc Reflection High Resistance Faults
storage, transportation or installation. Current Impulse Method High Resistance Faults
Voltage Decay Method Very High Voltage Flashover Faults
D. Operational Failure Modes
B. Tracer Methods and Pinpointing (Precision Locating)
Operational failures are those that occur after
commissioning or commissioning tests that may ultimately be After pre-locating the cable fault, the need for more
attributed to one of the failure mechanisms previously accurate fault location must be assessed. For example, if the
discussed. fault location is localized to an area between manholes in a
 Excessive Heating – Exceeding cable temperature limits duct bank, no further fault location may be required since the
can cause cable insulation damage or result in excessive entire manhole-to-manhole section could be replaced.
radial (shield/sheath damage) or longitudinal expansion However, if additional accuracy or confirmation is needed, say
(thermal mechanical bending-induced damage). for a direct buried cable or for a longer cable circuit, a
 Corrosion –Corrosion of the concentric neutral or metallic precision fault location method must be applied to pinpoint the
shield sheath or the cable pipe can affect the ongoing location. Tracer methods are applied for pinpointing and most
operation of the cable system. Reduction in the cross- commonly include the following:
 Modulated Direct Current can be difficult. This can be addressed by using a “burner” to
 Modulated Alternating Current lower the fault impedance (particularly on paper-insulated
 Audio Frequency cables, but at a cost of severe cable damage) or by coupling a
 Impulse (Capacitive Discharge or “Thumping”) TDR to a surge generator/thumper. The latter method
 Earth Gradient temporarily creates a lower impedance fault during which the
TDR can obtain a low impedance reflection (e.g., the “arc
C. Specific Fault Location Strategies reflection method”). Modern digital equipment will store this
Depending on the type of cable system, certain fault “faulted” trace, and allow comparison to the normal TDR
location strategies may be applied. These are summarized in trace. State-of-the-art commercially-available equipment
Table II. incorporates computer algorithms that aid the user by
TABLE II - FAULT LOCATION STRATEGIES automatically comparing the traces and indicating the
Cable Pre-Location Pinpointing approximate distance to the fault.
System Type
Distribution ● Arc Reflection ● Capacitive Discharge
(Isolated) (“Thumping”) with acoustic or
magnetic aids
Distribution ● Fusing (repeated
(Sectionalizing) application of fault current)
● Faulted Circuit Indicators -- Not Applicable --
● Arc Reflection
Network1 ● Arc Reflection may be ● Tone Tracing
used to exclude sections of ● Impulse Methods
network containing a fault ● Modulated A.C.
(testing from multiple ● Modulated D.C.
connections points ● Magnetic Detection
enhances effectiveness) ● Acoustic Detection
Transmission or● Bridge Methods ● Capacitive Discharge
Submarine ● Arc Reflection3 (“Thumping”) with acoustic or
● Voltage Decay (high flash- magnetic aids
over voltage) ● Impulse2
● Submarine pinpointing is very Figure 3: TDR test set coupled to a thumper (arc reflection)
difficult Current Impulse Method – The “current impulse
1
Historical local practices tend to dominate. Pre-location tends to be limited method” evaluates the traveling wave generated in the cable
due to complexity of applying methods to the system (presences of “T’s” or (instead of an external pulse) when the impulse generator
taps, transformers, etc.)
2
Consideration needed for capacitance of cable versus impulse generator size.
temporarily breaks down the cable fault. This technique has
3
May be limited due to long cable length or presence of cross-bonding
the advantage that more energy is used to create the traveling
shield/sheath connections. waves so longer cables may be tested, although interpreting
results is much more difficult, making this a seldom-used
VI. DESCRIPTION OF FAULT LOCATION METHODS method.
The following sections summarize the basic characteristics, Electrical Bridges – Bridge methods use an electrical
theories and application of various terminal (pre-locating) and bridge with known resistances and a null detector
tracer (pinpointing) fault location methods and equipment. (galvanometer) to determine the unknown distance to a fault
by using auxiliary cable (an un-faulted parallel cable works
A. Descriptions of Terminal (Pre-locating) Methods
well); see Fig. 4. With the known resistances, RA and RB
TDR - Time domain reflectometry (TDR, pulse-reflection (shown in Fig. 4), the distance to the fault, X, can be
method, pulse-echo method or cable radar) works well on determined.
short circuit faults (<150 Ohms) or series faults. After a pulse
is applied to the cable, reflections return to the test equipment
that characterizes the variations in impedance along the cable
which is viewed on an oscilloscope. Major deflections (cable
end, splices, faults) can be detected; upward deflections X
RA
 L  D
indicate increased impedances while downward deflections R A  RB
indicate decreased impedances to ground. TDR may be used Figure 4: Murray Bridge method for fault location.
to locate a fault by comparing the distance to the “end” of the Several other bridge techniques have been developed over
cable, possibly using an undamaged parallel phase, to the the years and consider the availability of other faulted or un-
faulted cable. The indicated distance may be different or a faulted cables, or where the cables may be of dis-similar types
deviation in the faulted trace as compared to a good phase may or per-unit resistances; detailed descriptions are available in
show the distance to the fault. A “three stake” or literature [1, 2].
“comparative” method can be used to stake the distance
indicated from TDR tests from both cable ends (two stakes); B. Descriptions of Tracer (Pinpointing) Methods
pinpointing to find the fault location (third stake) can be done Moduled AC/DC – Modulated direct current (DC) or
between the first two stakes. alternating current (AC) can be used to trace the portions of
Arc Reflection Method – The output voltage of a TDR is cable that have continuity to the location where the source is
no more than 50V typically, so locating high-impedance faults applied. The modulation makes the signal unique for the
technician to locate the fault by either checking the current to [2] E.C. Bascom, III, D.W. Von Dollen, H.W. Ng,
ground or by finding a voltage drop along the conductor “Computerized Underground Cable Fault Location
(usually limited to the cable sheath). The method is limited to Expertise”, Transactions of the IEEE Transmission &
low-resistance faults, possibly requiring a burner to condition Distribution Conferences, location, 1994.
the fault, and large power output of the test equipment that [3] R.L. Amonte, “URD Cable Fault Locations for the 1990s”,
potentially could further damage the cable. presented at 42nd Annual Power Distribution Conference,
Tone Tracing – Similar to modulated AC or DC, tone Austin, TX, October 24, 1989.
tracing applies a tone frequency to the cable’s metallic [4] Underground Cable Fault Location Reference and
shield/neutral/sheath that can be detected up to the point of the Application Guide, EPRI Report 1023060, Palo Alto, CA.
fault but not beyond. The method is limited to situations with [5] IEEE 1511-2004, “Guide for Investigating and Analyzing
low fault resistance and requires access to the cable to check Shielded Power Cable Failures on Systems Rated 5kV
for the tone; this may be limited to intermediate manholes or Through 46kV”.
splice boxes. [6] Underground Systems Reference Book, Edison Electric
Capacitor Discharge (“Thumping”) – In Institute, Transmission and Distribution Committee, 1957.
this technique, a D.C. power source charges a
capacitor until a mechanical spark gap or IX. BIOGRAPHIES
electronic trigger discharges the stored Earle C. (Rusty) Bascom, III (M’1989-SM’) holds B.S. and
capacitive energy into the cable causing a M.E. degrees in Electric Power Engineering from Rensselaer
momentary breakdown at the cable fault which Polytechnic Institute and an M.B.A. from the University of
may be detected acoustically with (Fig. 5) or New York at Albany. Mr. Bascom began his career in 1990
without amplification or sometimes with a working with underground cable systems at Power
radiofrequency or magnetic detector. This Technologies, Inc. and, later, Power Delivery Consultants, Inc.
method is simple to operate and easy to before founding Electrical Consulting Engineers, P.C. in 2010
understand, but there is a tendency to use where as a Principal Engineer he continues to provide
excessive voltage to increase the audible engineering services in the analysis, design, specification,
“thump”. Prolonged application can damage Fig. 5 – Acoustic installation oversight and operational characteristics of
the cable in the vicinity of the failure or amplifier underground and submarine transmission and distribution
elsewhere due to the repeated impulse. Thumping is probably cable systems. He is a Senior Member of the IEEE, Power &
the most common fault location technique applied and for Energy Society, a Voting Member of the Insulated Conductors
some situations is the only method applied. Effective pre- Committee and Standards Association, a member of CIGRÉ,
locating is the best way to avoid damage from prolonged and a registered professional engineer in NY, FL and TX. He
applications of thumping. can be contacted at [email protected].
Earth Gradient – When there is a fault to ground (most Martin J. von Herrmann (M'1996) holds a BSEE from the
common with a jacket fault or unshielded distribution cable), a Georgia Institute of Technology. In his 19 years with The
thumper or D.C. voltage source applies a voltage gradient into VON Corporation, Martin has gained extensive experience
the soil. A zero center volt meter with two probes spaced ~1m with high voltage electrical test and underground cable fault
(3ft) apart will provide a directional indication towards the location equipment. He has been involved in development,
fault with greater deflections indicating closer proximity to the manufacturing, sales and support of equipment used by electric
fault until directly above it (see Fig. 6). Interpreting the utilities around the world to locate underground cable faults,
measurements is complicated by other cables, pipes or types of and to electrically test shielded power cables and systems,
underground conductors in the vicinity. generators, insulated aerial lift devices, and electrical
protective equipment. He is a member of the IEEE, Power &
Energy Society, Dielectrics & Electrical Insulation Society,
and the Standards Association. He currently serves on the
Adcom of the Insulated Conductor's Committee. He can be
contacted at [email protected].
Dr. Tiebin (Tom) Zhao (SM’) holds B.S. and M.S. degrees
both in Electrical Engineering from Tsinghua University,
Figure 6: Earth gradient method and simulated readings China, and a PhD in Power Systems and High Voltage
Engineering from the Ohio State University in Columbus. Dr.
VII. CONCLUSIONS
Zhao is Program Manager of EPRI Underground
Many approaches can be considered to locate failures in Transmission. His current research activities focus on
underground power cable systems. Through the careful underground transmission technology, design, construction,
selection and application of appropriate methods both new and inspection, and assessment for both extruded and laminar
old, fault location can successfully be performed. dielectric cable systems. Before joining EPRI, Dr. Zhao
VIII. REFERENCES worked for FirstEnergy Corporation and Hubbell Power
Systems. Dr. Zhao is a Senior Member of IEEE, a Voting
[1] IEEE 1234-2007, “Guide for Fault-Locating Techniques on Member of IEEE Insulated Conductors Committee, and a
Shielded Power Cable Systems”. Member of CIGRÉ. He can be contacted at [email protected]