Documente Academic
Documente Profesional
Documente Cultură
(
E
-
3
2
3
4
5
6
8
20
30
40
50
60
80
200
300
400
500
600
800
1
10
100
1000
Tan Data for Ethylene-Propylene Rubber and Butyl Rubber Insulated Cables
C-14
Figure C-13 compares the 60 Hz tan for black Okonite EPR as shown in Figure C-1 to that for
Okonite butyl rubber from Figure C-12. The characteristic shape of the plots is essentially
identical for the two materials such that as the breakdown strength drops below 10 times
operating voltage, the tan value is beginning to increase. For the butyl and black EPR rubber
specimens in the assessments, the 10 times operating voltage value is 24,000 Vac. Due to
insulation thickness differences the voltage stress for the two materials was significantly
different. For the butyl rubber, it was 120 V/mil (4.7 kV/mm) and for the black EPR, it was
141 V/mil (5.5 kV/mm). At a breakdown strength of eight times operating voltage (19,200 Vac),
the tan for both materials has doubled (2X for EPR, 2.4X for butyl) from its lowest value. As
the breakdown strength decreases to four times operating voltage, the tan has increased by a
factor of 16 for the black EPR and by 24 for the butyl rubber. The comparisons show that the
lossier butyl rubber will have characteristically higher 60 Hz tan and have a larger change than
the black EPR as it deteriorates.
Figure C-13
Comparison of 60-Hz Tan Versus Breakdown Strength of Okonite Butyl Rubber to Black
Ethylene-Propylene Rubber
1
10
100
1000
2 4 6 8 10 12 14 16 18 20 22 24
ac Breakdown Voltage/Operating Voltage
T
a
n
D
e
l
t
a
(
x
1
0
-
3
)
Butyl
Black EPR
Tan Data for Ethylene-Propylene Rubber and Butyl Rubber Insulated Cables
C-15
Figure C-14 shows the same data as Figure C-13 but in terms of volts per mil. The figure shows
that tan does not change for Okonite black EPR and butyl until the breakdown strength has
decreased below 200 to 240 V/mil (7.9 to 9.5 kV/mm). For both materials, tan is slightly
greater than 10 10
-3
when the ac breakdown voltage has dropped to 100 V/mil (3.9 kV/mm).
When tan has increased to 30 10
-3
, the breakdown strength has dropped to 75 V/mil
(2.9 kV/mm).
1 V/mil = 0.0394 kV/mm
Figure C-14
AC Breakdown Voltage (in V/mil) for Okonite Butyl and Black Ethylene-Propylene Rubber
Insulations
The EPRI report Advanced Diagnostics and Life Estimation of Extruded Dielectric Cables
(1001727) also describes partial discharge (PD) testing performed on specimens in the laboratory
[88]. These butyl rubber cables had high natural attenuation due to the characteristics of the butyl
rubber and had corroded but intact helically wrapped copper tape shields. When the full length of
one of the cables with severe water-related degradation was tested, PD could not be detected
above background. However, an appreciable PD (50150 pC, depending on test voltage) could
be detected on the full length of the second cable . As both of the cables were cut into shorter
sections following breakdowns, PD was identifiable in each section with elevated tan . The
result of the PD measurements indicates that PD can be present in severely deteriorated butyl
rubber cable that is near failure; however, the combination of high attenuation of the butyl rubber
and shield corrosion that causes the tape shield to act as an inductor and further attenuate the
signal might make detection of PD difficult to impossible.
1
10
100
1000
0 50 100 150 200 250 300 350 400
ac Breakdown (V/mil)
T
a
n
D
e
l
t
a
(
E
-
3
)
Butyl
Black EPR
Tan Data for Ethylene-Propylene Rubber and Butyl Rubber Insulated Cables
C-16
C.6 Conclusions Related to Rubber Insulated Cables and Tan Results
IEEE Std 400-2001 provides acceptance criteria for the use of tan measurements for XLPE
insulated cables [51]. As further field and laboratory data are gathered, that guidance is expected
to be updated. This appendix was developed to provide insights regarding the response of
various EPR and butyl insulations during tan testing and the implications of the results. It is
recognized that the data available are limited, but these data do provide valuable insights.
Figure C-1 for black EPR and Figure C-12 for butyl rubber indicate that tan is inversely
proportional to ac breakdown voltage and that, when the breakdown voltage decreases below
10 times operating voltage, tan increases exponentially with a drop in breakdown voltage. It is
likely that all the rubber insulations behave similarly; however, it should not be assumed that
each manufacturers black or pink EPR has exactly the same characteristics as that of another
manufacturer. It is likely that there will be differences in the initial tan and the rate of change
as the breakdown voltage increases. The field results indicate that a baseline tan can be
determined by testing a few cables that are known to be dry. Even after extended periods, the tan
for such circuits appear to be like new unless there is some non-water-induced flaw in the
insulation or an accessory.
Figures C-1 and C-12 show 60-Hz tan results for black Okonite EPR and black Okonite butyl
rubber. Corresponding VLF figures do not exist. However, a comparison of healthy results for
the materials indicates that a VLF results can be a factor of 1.8 higher than the 60-Hz results. It is
not clear that this factor holds true at all degradation values. If it does, it indicates that severely
aged black Okonite insulation having ~70 V/mil (2.7 kV/mm) or 4V
0
breakdown strength would
have a tan of approximately 72 10
-3,
as opposed to a healthy tan of ~5 10
-3
. A similar
factor for black Okonite butyl rubber would indicate an initial VLF value of ~9 10
-3
. The
specimens evaluated had a thick layer of insulation (190 mils [4.8 mm]); accordingly, 70 V/mil
(2.7 kV/mm) would occur at 5.5V
0
, which corresponds to a VLF of ~68 10
-3
.
Figures C-13 and C-14 show that the 60-Hz tan versus ac breakdown strength behavior of EPR
and butyl rubber are similar. When the data are given in terms of volts/mil, the degradation
graphs overlap as breakdown strength degrades. If the 1.8 factor between 60-Hz and VLF tan
hold true, then when the VLF tan is approximately 20 10
-3
, the ac breakdown strength would
be approximately 100 V/mil (3.9 kV/mm). A VLF tan of 54 10
-3
would indicate that the
breakdown strength is approximately 75 V/mil (2.9 kV/mm). Although these values might seem
to be well within acceptable conditions (for example, on 110-mil [2.8 mm] insulation, the
breakdown strength would be 8250 volts at 75 V/mil (2.95 kV/mm) or a factor of 3.4 times
2400 V), cable experts would consider the cables highly degraded. In fact, cable experts consider
cables with a breakdown strength of 200 V/mil (7.9 kV/mm) highly degraded. When the
breakdown strength has deteriorated to 200 V/mil (7.9 kV/mm) or as little as 75 V/mil
(2.95 kV/mm), the insulation system might not be able to withstand switching and lightning
surges without an electrical tree forming, leading to failure shortly thereafter. Accordingly,
conditions indicating more than doubling of a healthy tan indicate that more frequent testing is
necessary to confirm that the condition is stable. Values indicating 75 V/mil (2.95 kV/mm) or
less indicate that replacement or rejuvenation should be considered.
Tan Data for Ethylene-Propylene Rubber and Butyl Rubber Insulated Cables
C-17
The laboratory VLF tan results for Anaconda pink EPR are consistent with healthy field
results. Unfortunately, no complete tan versus ac breakdown curve is available for the
Anaconda pink EPR. Figure C-8 indicates that increasing tan relates to deterioration, as does
the difference in tan when measured at 1V
0
and 2 V
0
. From Figure C-8, there appears to be a
strong tendency of differential increases of tan with increasing test voltage for the Anaconda
cable. Figure C-9 indicates that, in some cases with a tan of above approximately 30 10
-3
, the
differential increases with increasing test voltage are so large that testing at 1.5V
0
to 2V
0
cannot
be safely accomplished and that investigation of the need to replace the cable will be necessary.
C.7 Effects of Mixed Shielded and Nonshielded Segments on Tan
One plant reports having circuits in which the dry sections have nonshielded cable and
underground (wetted) sections have shielded cable. This is an unusual cable system design, and
the results might have extremely limited value to plants with nonshielded cable. The nonshielded
cables are three-conductor cables with an overall armor. These cables were all black EPR
manufactured by Okonite. Figures C-15 and C-16 show the VLF tan results for the cables with
mixed nonshielded and shielded segments. The test voltages for these two figures were slightly
different, requiring the data to be plotted separately. (Each cluster of test voltages represents one
phase of a cable.)
Figure C-15
Tan Results for Mixed Shielded and Nonshielded Black Okonite Ethylene-Propylene
Rubber Circuits
0
10
20
30
40
50
60
70
80
90
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
T
a
n
D
e
l
t
a
(
E
-
3
)
1.1 kV
2.2 kV
3.4 kV
4 kV
Tan Data for Ethylene-Propylene Rubber and Butyl Rubber Insulated Cables
C-18
Figure C-16
Tan Results for Mixed Shielded and Nonshielded Black Okonite Ethylene-Propylene
Rubber Circuits
In Figure C-15, the cable tests represented by markers 37, 38, and 39 are a special case for two
reasons: 1) only 15% of the cable was shielded and 2) each phase had three conductors in
parallel. The large amount of nonshielded cable significantly elevates the tan . The further
elevation of the tan results could be caused by having the nonshielded cables between the test
set and the shielded sections. The effects of the length of the nonshielded sections on the tan is
described below.
In Figure C-16, the data for a three-phase cable are associated with x scale markers 4, 5, and 6.
The data for the A phase of the cable is shown at scale marker 4. This phase had elevated tan
results and a significant differential between 1V
0
and 2 V
0
. The aged wet section of the cable was
replaced with modern Okonite pink EPR. Also in Figure C-16, the C phase (marker 27) of cable
having the x scale markers 25, 26, and 27 shows similar degradation. The utility plans to replace
the cable based on the tan measurements that have been observed.
Tan Data for Ethylene-Propylene Rubber and Butyl Rubber Insulated Cables
C-19
Figure C-17 shows a plot of the average tan for the mixed circuits versus the percent of
nonshielded cable in the circuits. Although the correlation is not perfect, the tan is
proportionally related to the percent of the nonshielded cable in the circuit. This result is logical,
in that the nonshielded segments would have a lower capacitive current than the shielded
circuits. At the same time, there would still be leakage currents in the nonshielded circuits with
respect to the armor in the three-phase cable segments. The approximate relationship of tan to
the percent of nonshielded cable is interesting but not accurate enough to allow an adjustment so
that an absolute value of tan could be used to assess a cable. However, comparison of results
between phases and significant increases in tan with increasing test voltage obviously can
identify deteriorated insulation, even though the circuits have a mix of shielded and nonshielded
segments.
Figure C-17
Average Tan Value Versus Percent of Nonshielded Cable Section
The testing of the circuits was by means of connecting the test set to the conductor and to the
station ground. It was not necessary to connect directly to the cable shield connection.
0
10
20
30
40
50
60
70
0 10 20 30 40 50 60 70 80 90 100
% Unshielded
A
v
e
r
a
g
e
T
a
n
D
e
l
t
a
(
E
-
3
)
Avg. Tan Delta
Linear (Avg. Tan Delta)
D-1
D
ADDITIONAL POLYMER MATERIALS INFORMATION
D.1 Fundamentals of Elastomers
Any polymer, elastomers included, consists of many (branched) chains having a distribution of
molecular weights, and so it is common to refer to average molecular weights. One can refer to
the number of molecules of each monomer present in the chain or the molecular weight of the
chains themselves. In elastomers, these chains are linked together. These long, intertwined
polymer chains are important because they give the rubber (after cross linking) the ability to
stretch and to return to their original shape. At rest, the long chains have several bends rather
than being straight and parallel. Therefore, when insulation is bent or pulled, the insulation can
stretch because the long chains can straighten. When the stress is removed, the insulation will
return to its original shape
Uncured (unvulcanized) elastomers are inherently soft, flexible, and often tacky. If the chains are
not cross linked, they will not return to the original shape after relaxation occurs. Therefore, to
render them useful, the following two steps must be taken:
The polymer chains must be extremely long, and different chains must be linked together.
This is referred to as cross linking (also referred to as vulcanization or curing). Cross linking
is induced after the cable is extruded; it increases the toughness and improves the physical
and mechanical properties of the elastomer. Different elastomers use different curing
techniques. A cross-linked elastomer is depicted in Figure D-1. The curved lines represent
the chains, and the short (almost horizontal and vertical) lines between chains represent the
cross links. (See Section D.2, Cross Linking, for details.)
Cross linking alone is not adequate to achieve usefulness. A number of other additives, called
fills, must be introduced to the elastomer to attain usefulness. Many of these are inorganic,
and one common required ingredient is clay, which imparts toughness. Other ingredients
improve mechanical and electrical properties and also facilitate processing and stability.
The use of numerous additives requires that sophisticated mixing technology be used to
convert the mixture of ingredients into a blend suitable for conversion into cable insulation
by extrusion.
Additional Polymer Materials Information
D-2
Figure D-1
A Cross-Linked Elastomer
D.2 Cross Linking
If an elastomer possessed two chains of equal length, the cross linking would lead to an increase
in molecular weight; in this case, doubling. Chain length is not a factor in influencing the cross-
linking process. As the degree of cross linking increases, more and more chains are incorporated
into the cross-linked network. In a practical insulation, a significant majority of the polymer
chains are no longer spaghetti-like but are linked together, forming a three-dimensional network.
The elastomers mechanical properties are dominated by this cross-linked fraction. The cross-
linked network is referred to as the gel fraction and the fraction that is not cross linked is referred
to as the sol. Practical cable insulations contain a small sol fraction. Insulated Cable Engineers
Association and Association of Edison Illuminating Companies specifications require the sol
fraction in cross-linked polyethylene (XLPE) to be less than 30% by weight.
Because the gel fraction is so important, another key point is noted. In considering the properties
that are influenced by cross linking, the length of the chain between the cross links is significant.
The closer together the cross links (the small lines in Figure D-1), the lower the length of the
chains between the cross links, giving improved mechanical properties to the elastomer.
Additional Polymer Materials Information
D-3
D.3 Butyl Rubber Compositions
The formulations listed in Table D-1 are from the early technical literature; they describe the
components present in butyl rubber compounds for wire and cable applications [35, 36].
Table D-1
Butyl Rubber Insulation Components
Component
Amount (Relative
Weight)
Butyl 035 (Enjay) 100
Zinc Oxide 5
Calcined Clay 90
Stearic Acid 0.5-1.5
Catalyst: See below
Quinone Dioxime Dibenzoate (also referred
to as DibenzoGMF)
6
Lead Oxide (Pb
3
O
4
) 9
Sulfur 1.5
This formulation had the following properties:
Tensile strength, 840 psi (0.6 kgf/mm
2
)
Elongation, 330 %
Moisture absorption, 7 days at 185F (85C), 11.1 mg/in
2
(1.7 mg/mm
2
)
Dielectric strength, 550 V/mil (21.7 kV/mm)
A second common compound for wire insulation used a similar polymer-filler composition but a
different catalyst system, as shown in Table D-2.
Table D-2
Butyl Rubber Wire Insulation Components
Component Amount (Relative Weight)
Butyl 035 (Enjay) 100
Zinc Oxide 5
Calcined Clay 90
Stearic Acid 0.5-1.5
Catalyst See below
Quinone Dioxime (also referred to as GMF) 1.5
Mercaptobenzothiazole (MBTS) 4
Lead Oxide (Pb
3
O
4
) 5
Additional Polymer Materials Information
D-4
The tensile strength for this composition was 930 psi (6.4 MPa); elongation, 290%; moisture
sorption, 10.9%; and dielectric strength, 695 V/mil (27.4 kV/mm). The curing agent system plays
a significant role in resistance to degradation under thermal stress. The sulfur cure can soften
with thermal aging, whereas the lead oxide cure tends to remain stable.
Polymers such as phenol-formaldehyde resins can be used as the cross-linking agent for butyl
rubber, although no specific notation for wire applications is included in the referenced literature.
What is relevant is that each method of inducing cross linking will impart a different nature to
the cross-linked regions, different additives in the final rubber, and potentially different
responses with aging.
D.4 Ethylene-Propylene Rubber and Ethylene-Propylene-Diene Monomer
Ethylene-propylene rubber (EPR) is a copolymer of ethylene and propylene. Sometimes a third
monomer is added to the composition to assist in the cross-linking process; it has no influence in
plant operations. This termonomer is polymerized with the ethylene and propylene and is present
in rather small amounts (~3-6%). When this third monomer is used, the insulation is referred to
as ethylene-propylene-diene monomer (EPDM). The termonomer provides unsaturation (double
bonds) and facilitates cross linking by use of sulfur. The term EPDM refers to ethylene-
propylene-diene monomer or terpolymer, the latter two referring to the repeating CH
2
units in the
polymer backbone. Figure D-2 shows the EPDM terpolymer, with ethylene-propylene and
5-ethylidene-2-norbornene as the termonomer. EPDM can be cross linked using sulfur, but this
method is not used for modern wire and cable systems.
Figure D-2
Ethylene-Propylene-Diene Monomer Terpolymer
Cable accessories (such as joints and terminations) are made of EPDM, and manufacturers
literature designates them as EPDM.
D.5 Dicumyl Peroxide Cross-Linking Agent Byproducts
The most common byproducts of the cross-linking process that uses dicumyl peroxide are
acetophenone and dimethyl benzyl alcohol (also referred to as cumyl alcohol). Other known
byproducts are alpha-methylstyrene and methane gas. All these byproducts are present in small
quantities, but they can influence aging and reliability. They diffuse from the cable system over
time, with methane diffusing most rapidly. Although the changes due to cross linking do not
improve the electrical properties, peroxide cross linking agent byproducts provide superior tree-
resistant properties to XLPE.
Additional Polymer Materials Information
D-5
D.6 Ethylene-Propylene Rubber Formulations
Exxon [25] has described the process of developing wire and cable insulation formulations based
on black EPR. A typical carbon black formulation from the 1970s using silane-coated clay is
shown in Table D-3.
Table D-3
Typical Black Wire and Cable Insulation Ethylene-Propylene Rubber Compound
Components from the 1970s
Component
Amount (Relative
Weight)
EPR Vistalon 404 80
Low-Density Polyethylene 20
Zinc Oxide 5
Coated Clay 110
Carbon Black 10
Flectol H 1.5
Lead Oxide 5
Wax 5
Silane 2
SR-350 1.5
Peroxide 7
Significant properties for this black, filled EPR are shown in Table D-4.
Table D-4
Properties of Black Ethylene-Propylene Rubber Compound Shown in Table D-3
Property Value
Tensile Strength 1210 psi (0.85 kgf/mm
2
)
Elongation 200%
Elongation Percent Retention After Seven Days at 300F (149C) 95%
Moisture Absorption, 194F (90C), Seven Days, Percent Weight
Increase
0.1% (0.9 mg/in
2
[0.14 mg/cm
2
])
Dielectric Strength 1370 V/mil (53.9 kV/mm)
Dielectric Constant 3.4
Power Factor, Original and One Week, 194F (90C) Water 0.005, 0.0044
Additional Polymer Materials Information
D-6
The moisture sorption is greater than an order of magnitude less than that for butyl rubber. Also,
the dielectric strength is higher, and the physical properties are excellent. A similar (but not
exactly the same) formulation for black power cable insulation from the 1970s was provided by
Exxon [26] using Vistalon 2504, an EPDM (see Table D-5).
Table D-5
Medium-Voltage Black Ethylene-Propylene Rubber Insulation Using Ethylene-Propylene-
Diene Monomer
Component
Amount (Relative
Weight)
EPDM Vistalon 2504 100
Carbon Black 10
Treated Clay 110
Antioxidant 1.5
Lead Oxide 5
Zinc Oxide 5
Process Oil 15
Paraffin 5
Silane 1
Peroxide 3.5
Formulation components can differ between suppliers. In addition, although the function of the
components in any EPR formulation is well known, specific individual components used by
individual suppliers can differ from one another.
Additional Polymer Materials Information
D-7
D.7 Influence of Clay on Properties of Ethylene-Propylene Rubber
Table D-6 shows the compounding of two EPRs: one using calcined clay and the other using
silane-treated calcined clay, which is more representative of modern cables. Other than the clay
types, the polymers are essentially the same.
Table D-6
Compounds of EPR Using Calcined and Coated Clay
Component Calcined Clay (%) Treated Calcined Clay (%)
Vistalon 404 32.7 32.4
Low-Density Polyethylene 8.2 8.1
Zinc Oxide 2 2
Calcined Clay 44.9 44.5
Silane A-172 0 0.8
Flectol H 0.6 0.6
Lead Oxide 2 2
Wax 2 2
SR-350 0.6 0.6
Peroxide 2.8 2.8
SRF Black 4 4
Table D-7 shows the effect of treatment of the clay on EPR properties and the limitations of
calcined clay with respect to electrical properties in EPR. The power factor, dielectric constant,
and water sorption characteristics all improve in the treated calcined clay system. This indicates
that the use of treated calcined clay has been a significant improvement in compounding EPR.
Table D-7
Influence of Clay Nature on Ethylene-Propylene Rubber Cable Properties: Calcined Versus
Coated Clay
Properties
Ethylene-Propylene Rubber
with Calcined Clay
Ethylene-Propylene Rubber with
Treated Calcined Clay
Power Factor After 194F (90C)
Water, 2 weeks
0.062 0.0037
Dielectric Constant After 194F
(90C) Water, 2.5 Weeks
4.09 3.52
Water Sorption After 194F
(90C), 2 Weeks
0.69 % 0.12 %
Another factor in influencing elastomer properties is clay particle sizethe smaller the average
size, the greater the polymerinorganic component interfacial contact and the better the
anticipated properties. More even size distributions with smaller-sized particles have been
implemented in cable manufacture, as well, to improve electrical and mechanical properties.
E-1
E
OFF-LINE TESTS THAT ARE UNDER DEVELOPMENT
E.1 Introduction
Several additional diagnostic tools are being developed but are not commonly used. They are
summarized in this section. Additional information is available in the EPRI reports Advanced
Diagnostics and Life Estimation of Extruded Dielectric Cables (1013085) and Guide for Non-
Destructive Diagnosis of Extruded Cable Systems (1001731) [104, 90].
E.2 Isothermal Return Current
The isothermal return current (IRC) test method applies low-level dc to the cable system for a
defined time. The treated cable system is then discharged by shorting the cable for a brief period,
and then the current from the insulation (also referred to as the relaxation current) is followed
and measured. When the discharge current is plotted against time (in seconds), different cable
systems provide different relaxation curves and peaks as a function of time elapsed after shorting
the system. From the nature of the curves and peaks, the claim is that the state of the cable can be
determined and a risk factor provided.
IRC plots for several cables subjected to various degrees of aging are shown in Figure E-1
[90, 105]. The shapes of the curves and locations of the peaks change as aging progresses.
Figure E-1
Isothermal Return Current Plots for Cross-Linked Polyethylene Cable Having Undergone
Different Degrees of Aging
Off-Line Tests That Are Under Development
E-2
An advantage of this diagnostic method is that the important data are collected after the dc stress
has been removed; a disadvantage is that the return currents are extremely small.
A typical procedure applies 1000 Vdc for 30 minutes (to allow the current to saturate due to the
dc conductivity), grounding occurs for ~20 seconds, and then measurements are made for the
next 30 minutes while discharging is taking place. The use of relatively low dc voltage is
claimed to be unable to inject charges into the system. This process induces a polarization within
the insulation system. The depolarization current is considered to be due to trapped charge
release as well as polymer chain motion. However, the currents that are measured are likely not
only due to the insulation response to the applied dc, but also to the entire cable construction; for
example, shield-insulation interfaces, polymer-filler interaction, impurities such as water (in
addition to likely polar water treed regions), and cable length (joints further increase the
complexity). The analysis of the data generated requires the use of computer-based network
evaluation tools to estimate the aging status of the cable. Information about the cable
construction and length can be fed into the database, but not the obviously more subtle
information. Regardless, the interpretation is quite complex, and it has been noted that IRC can
be considered to be a superposition of different components exhibiting different relaxation times.
E.3 Return Voltage
The return voltage (RV) test applies dc repetitively; after application of the first dc stress, the
cable is discharged and then repeatedly recharged and discharged a total of four times. Key
information sought includes 1) the peak value of the recovery voltage, 2) the time to reach peak
voltage, and 4) the gradient of the recovery voltage curves. The information is fed into a
computer program.
The RV procedure involves charging for ~15 minutes followed by a two-second discharge; this
is again followed by the charge and discharge sequence. The charging time can vary, and the
discharge time can vary and be modified by relating it to the charging time. The dc voltage
application starts low and is gently increased from 0.5V
0
to 1.0V
0
, 1.5V
0,
and finally 2.0V
0
. The
test time for one sample can take 1.5 hours. This charge and discharge process induces various
polarization processes within the insulation. For example, oxidized regions of aged cables (which
result from aging) would possess more dipoles and, therefore, greater potential for changes and
then for relaxation after short circuiting.
By measuring the return voltage as a function of applied dc voltage stress, it is claimed that a
linear response time indicates less aging and a nonlinear response time indicates more aging.
Clearly, one can expect different results depending on the degree of aging, charging time,
discharging time, and local environment.
The procedure measures the open circuit voltage that appears after poling, rather than the short-
circuit current as does IRC. The charging process is, therefore, longer and more repetitive than
for the IRC measurement.
Because the procedure causes charge to be stored and released, it is to be anticipated that, as
cables age, longer times will be required for charges to be released (aged cables are more polar)
and easier trapping of charges. Moisture, ionic contaminants, filler-polymer interfaces, and
shield adhesion can all influence test results. It is desirable to perform a test concurrent with an
unaged cable specimen of the same construction, vintage, and storage history.
Off-Line Tests That Are Under Development
E-3
E.4 Oscillating Wave
The oscillating wave (OW) test provides a PD test methodology for locating discharge sites,
using a damped oscillating wave as the voltage source. This represents an alternative test voltage
waveform that does not require as large a power source as does a power frequency test set. The
OW test set consists of a remotely controlled, high-voltage dc supply; a specially designed, solid-
state switch; an air-core inductor/voltage divider/PD coupler; and an industrial computer as a
control unit. Digital data acquisition and signal storage, as well as analysis and evaluation of the
PD signals, take place in the control unit.
The OW method [106] charges the test cable using the dc supply; the charging time is of the
order of 20 seconds or less (so that potentially harmful space charges do not occur) [E.4]. After
this time, service voltage is reached. At this point, the solid-state switch with fast closure time
creates a series-resonant circuit between the cable and an air-core inductor. The circuit begins to
oscillate, the latter being the resonant frequency with a delay time of 0.1 to 1 second. This
produces a few of tens of cycles and, if PD is initiated, the circuit detects pulses that occur during
the oscillating wave lifetime. The inductor and capacitance of the cable form a resonant circuit
that produces a damped oscillatory sine wave. The frequency range is 50 Hz to 1 kHz.
The rate of decay of the oscillation will depend on the resistance of the circuit (that is,, the
dissipation factor of the insulation). Typically up to 50 shots are applied at a specified voltage
level (it can be varied). From one trace, it is possible to determine the PDIV and PDEV, the
pulse magnitude, and the dissipation factor. Although the test set is primarily designed to detect,
measure, and locate PDs, it can also measure the dissipation factor; however, its sensitivity is not
as high as dissipation factorspecific measuring equipment.
A factor to be considered is that the information developed is related to cable length [107].
Details of the technology are described in the EPRI reports Advanced Diagnostics: Life
Estimation of Extruded Dielectric Cables (1011499), Guide for Non-Destructive Diagnosis of
Extruded Cable Systems (1001731), and especially Advanced Diagnostics and Life Estimation of
Extruded Dielectric Cables (1013085) for nuclear issues [49, 90, 104].
F-1
F
INSULATION RESISTANCE TEST MEASUREMENTS:
THEIR VALUE AND LIMITATIONS
F.1 Introduction
Very-low-frequency or ac withstand, tan , dielectric spectroscopy, and partial discharge (PD)
testing are the preferred test methods for extruded polymer testing, depending on the potential
problem being evaluated and the nature of the cable design. However, insulation resistance
testing is often used to assess cables due to its ease of application and ready availability.
Insulation resistance measurement is a valuable troubleshooting tool if used properly, and in
some cases, it will be a good indicator of severe degradation of a cables insulating system.
However, insulation resistance measurements must be used cautiously as a means of
understanding age-related degradation. Some plants have adopted inappropriately low values
of insulation resistance for acceptance criteria for return-to-service for medium-voltage cable.
In addition, this acceptance criterion does not take circuit length into consideration. This section
describes why higher values of insulation resistance are appropriate for cable insulation
assessment and provides cautions regarding sole dependence on insulation resistance for
assessment of medium-voltage cable. For nonshielded cable, insulation resistance is likely to be
the only test that is practical to use. Consideration of the cautions provided in this section will
improve the value of the test until an improved test is available.
F.2 Determining Minimum Insulation Resistance Value and Improving
Interpretation of Results
Minimum insulation resistance values exist for cable and motor circuits for the operation and
application of elevated voltage tests. Often, insulation resistance testing acceptance criteria for
cables have been based on acceptance criteria for rotating machinery contained in IEEE Std. 43-
2000, IEEE Recommended Practice for Testing Insulation Resistance of Rotating Machinery
[108]. When the original version of this standard was written in 1974, it contained the following
formula for minimum acceptance value of insulation resistance for a winding:
1
min 1
kV IR Eq. F-1
Where:
IR
1min
is the recommended minimum insulation resistance, in M, at 104F (40C), of
the entire three-phase winding.
kV is the rated machine terminal-to-terminal voltage, in rms kV.
Insulation Resistance Test Measurements: Their Value and Limitations
F-2
If individual phases were to be tested separately and guard circuits were used for the phases not
under test, the minimum insulation resistance is three times this value.
IEEE Std. 43-2000 contains an additional minimum insulation resistance value of 100 M for
form-wound windings built after 1970. For separately tested phases, the requirement is 300 M.
Although the minimum insulation resistance of Equation F-1 was meant to apply only to wet or
deteriorated windings, the industry adopted it for general use and applied it to cables and the
combination of motors tested with their field cable. Insulation resistance testing of the motor
through the extruded polymer-insulated cable generally will not be a problem with respect to
understanding the condition of the motor winding insulation system. As long as the cable has not
degraded, the test will be dominated by the motor. However, a number of problems occur when
the condition of the cable is evaluated either with the motor or by itself using the criteria of
Equation F-1. The minimum insulation resistance of Equation F-1 is applicable to a varnished or
asphalt winding that was wet to determine if it was safe to re-energize or to expose to an elevated
voltage test. The leakage currents were not necessarily associated with a localized deterioration
but rather a generalized leakage over the entire surface of insulation the winding. The 100 M
requirement for form-wound windings built after 1970 applies to insulation systems that have
epoxy and/or mica tape insulation systems that have better barriers to surface moisture and,
accordingly, much higher insulation resistance values, even when damp. Extruded medium-
voltage cable insulation systems do not behave like random-wound varnished or asphalt
windings. Wetting the jacket of a shielded cable will have no immediate effect on insulation
resistance. Although wetting the jacket of nonshielded cable might provide a return path for the
test, it will not cause a low insulation resistance. The criteria of Equation F-1 from IEEE Std. 43
definitely do not apply to extruded polymer insulation. The 100300 M value for post-1970
form-wound motors comes closer to being a valid minimum. However, the following are
also true:
The insulation resistance of a cable is indirectly proportional to its length. Accordingly,
minimum values must be corrected to a specific length, such as 1000 ft or 1 km. If the
minimum insulation resistance is 100 M-1000 ft (30.5 M-km), and the circuit length is
250 ft (76 m), the minimum insulation resistance is 400 M. Similarly, if the length is
2000 ft (61 m), the minimum insulation resistance is 50 M.
The introduction to IEEE Std. 43-2000 states Whereas individual insulation resistance
measurements may be of questionable value, the carefully maintained record of periodic
measurements, accumulated over months and years of service, is of unquestioned value as a
measure of some aspects of the condition of the electrical insulation. There are many
variables associated with insulation resistance testing of motors, including temperature and
humidity. The same is true for extruded polymer insulation of cables. Gross trending of the
results over time is useful; however, trying to evaluate small changes (factors of 2 to 10 for
high values) is inappropriate. General decreasing trends of decades are important.
Insulation Resistance Test Measurements: Their Value and Limitations
F-3
Insulation resistance testing is a global or bulk test. It cannot differentiate between an overall
decrease in the insulation resistance in the wall of the polymer and one or two severe,
localized, low insulation resistances. In addition, if a near-through-wall defect has a thin
layer of good insulation between it and the shield or conductor, the insulation resistance test
will not identify it.
Even when evaluating a failed cable, the insulation resistance can be tens of megohms or
more because the fault is likely to have blown out the metal shield, causing long surface
resistance paths between the exposed conductor and the remainder of metallic shield.
Insulation resistance measurement is a useful troubleshooting tool for evaluation of medium-
voltage cable. For extruded cable, values less than 100 M-1000 ft (30.5 M-km) are a strong
indication that a problem exists with the circuit and investigation is necessary. However, the
converse is not necessarily true. Reasonable confidence that the circuit is sound exists when
values of gigohms or higher are present when adjusted for length, but the test will not necessarily
detect voids and partial through-wall damage that could result in ultimate failure under ac
operating conditions. Some improvement in value of insulation resistance results are possible for
cable testing if the following are considered:
The insulation resistances of each phase of a cable should be similar to those of the other
phases. Decades of difference between phases indicates that a problem might exist with a
phase or phases.
Periodic measurements taken over the course of years should be similar. Variations are
expected due to differing conditions at the time of test, but the overall value should remain
high and relatively consistent.
Gross decreasing trends or major decreasing step changes between periodic measurements
are an indication of a problem and should be investigated further. A more sophisticated test is
recommended for shielded cable, such as ac withstand, tan , dielectric spectroscopy, or PD,
as applicable.
The insulation resistance measurement should smoothly increase or remain stable during the
period of the test. Instability in the measurement, such as oscillations or continuing decreases
in values, are indications of a breakdown phenomenon. Such oscillations might be present in
a single-phase cable that has failed and the shield continuity has been lost in the vicinity of
the fault. Returning such a cable to service has a high probability of resulting in a phase-to-
phase fault after a short period of operation.
Insulation Resistance Test Measurements: Their Value and Limitations
F-4
F.3 Insulation Resistance of Good Insulation
Having an insulation resistance of 100 to 300 M-1000 ft (30 to 100 M-km) does not indicate
that the cable insulation is in good condition if it just exceeds these values; rather, it indicates
that the cables can function and likely withstand operating or test voltage without immediate
failure. Cable manufacturing standards, such as Insulated Cable Engineers Association
specification S-68-516 [71], require minimum insulation resistances for new cable based on the
following formula:
IR = 10,000 log
10
(D/d) M-1000 ft Eq. F-2
Where:
D is the diameter over the insulation.
d is the diameter under the insulation.
To allow a comparison of the new cable requirement to the minimum acceptance criteria, assume
a 500-ft (152-m) long cable with a 1/0 AWG (53.5 mm
2
) conductor and an 80-mil (2-mm) thick
600-V rated insulation. A representative cable of this type would have a 0.49 in. (12.4 mm)
diameter. The field acceptance limit would be 1.6 M. The new cable value would equal
the following:
IR = 10,000 log
10
(0.49 in./0.33 in.) M * 1000 ft/500 ft =3440 M
If the circuit was 2000 ft (610 m) long, the new cable value would be 860 M. The length of
the circuit is a significant consideration related to expected insulation resistance values.
A newly installed cable should have a near-factory-value insulation resistance per 1000 ft
(305 m), which, in this example, would be 6.8 G. Using 100 to 300 M-1000 ft (30 to
100 M-km) as an acceptance criteria for the continued use of this cable example results in
a value that is 23 to 68 times the allowable reduction from the original minimum insulation
resistance. Most cables, when manufactured, have insulation resistance values that greatly
exceed the minimum manufacturing requirements, making the difference between the as-
manufactured to the degraded condition even larger.
F.4 Conclusions
Insulation resistance is a valuable troubleshooting tool. However, it must be used cautiously as a
condition monitoring tool. The acceptance criterion of Equation F-1 (1 M plus 1 M per kV of
rating) is not adequate for extruded polymer insulation used in medium-voltage cable. A
minimum value of 100 M-1000 ft (30 M-km) is recommended to preclude severely degraded
cables from being returned to service. Even when this value is used, higher values of insulation
resistance might not be indicative of a sound insulation system. The cautions and techniques for
improving the value of the results described in Section F.2, Determining Minimum Insulation
Resistance Value and Improving Interpretation of Results, should be considered.
Electric Power Research Institute
3420 Hillview Avenue, Palo Alto, California 94304-1338 PO Box 10412, Palo Alto, California 94303-0813 USA
800.313.3774 650.855.2121 askepri@epri.com www.epri.com
Electric Power Research Institute
3420 Hillview Avenue, Palo Alto, California 94304-1338 PO Box 10412, Palo Alto, California 94303-0813 USA
800.313.3774 650.855.2121 askepri@epri.com www.epri.com
The Electric Power Research Institute, Inc. (EPRI, www.epri.com)
conducts research and development relating to the generation, delivery
and use of electricity for the benefit of the public. An independent,
nonprofit organization, EPRI brings together its scientists and engineers
as well as experts from academia and industry to help address
challenges in electricity, including reliability, efficiency, health, safety and
the environment. EPRI also provides technology, policy and economic
analyses to drive long-range research and development planning, and
supports research in emerging technologies. EPRIs members represent
approximately 90 percent of the electricity generated and delivered in
the United States, and international participation extends to more than
30 countries. EPRIs principal offices and laboratories are located in
Palo Alto, Calif.; Charlotte, N.C.; Knoxville, Tenn.; and Lenox, Mass.
Together...Shaping the Future of Electricity
2012 Electric Power Research Institute (EPRI), Inc. All rights reserved. Electric Power
Research Institute, EPRI, and TOGETHER...SHAPING THE FUTURE OF ELECTRICITY are
registered service marks of the Electric Power Research Institute, Inc.
Export Control Restrictions
Access to and use of EPRI Intellectual Property is granted with the spe-
cic understanding and requirement that responsibility for ensuring full
compliance with all applicable U.S. and foreign export laws and regu-
lations is being undertaken by you and your company. This includes
an obligation to ensure that any individual receiving access hereunder
who is not a U.S. citizen or permanent U.S. resident is permitted access
under applicable U.S. and foreign export laws and regulations. In the
event you are uncertain whether you or your company may lawfully
obtain access to this EPRI Intellectual Property, you acknowledge that it
is your obligation to consult with your companys legal counsel to deter-
mine whether this access is lawful. Although EPRI may make available
on a case-by-case basis an informal assessment of the applicable U.S.
export classication for specic EPRI Intellectual Property, you and your
company acknowledge that this assessment is solely for informational
purposes and not for reliance purposes. You and your company ac-
knowledge that it is still the obligation of you and your company to make
your own assessment of the applicable U.S. export classication and
ensure compliance accordingly. You and your company understand and
acknowledge your obligations to make a prompt report to EPRI and the
appropriate authorities regarding any access to or use of EPRI Intellec-
tual Property hereunder that may be in violation of applicable U.S. or
foreign export laws or regulations.
Program:
Generation Maintenance Applications Center
1024044