DOCUMENTOHENDRIX
DOCUMENTOHENDRIX
DOCUMENTOHENDRIX
for
Southwire
Project Number: 05-148
March, 2006
Principal Investigator:
Raymond C. Hill, PE
Reviewed by:
R. A. Hartlein
Underground Systems Program Manager
National Electric Energy Testing, Research & Applications Center
Southwire, 15kV, 336.4, 18/1 spacer cable was tested with Hendrix polyethylene tie-top and vice-
top, pin type insulators, a Preformed Line Products (PLP) polymeric, tie-top, pin type insulator
(added near the end of the test program), and two models of Hendrix polyethylene three-phase
spacers. The Southwire spacer cable was also tested with a tie-top, pin type porcelain insulator as a
control sample. An equivalent Hendrix spacer cable was later acquired and tested with the Hendrix
vice-top, pin type insulator. Parital discharge measurements were made on all pin type insulators
with both spacer cables. The ac withstand test consisted of the application of accelerated voltage for
a period of six (6) hours. Afterwards, the spacer cables, insulators, and spacers were inspected for
damage.
The results indicate that Southwire, 15kV, 336.4, 18/1 spacer cable is compatible with the Hendrix
tie-top, pin type insulator, Hendrix three-phase spacers, and the PLP tie-top, pin type insulator when
exposed to the high ac voltage stresses utilized in this project. In addition, the Southwire spacer
cable was equal to or more compatible with the Hendrix vice-top, pin type insulator than the
equivalent Hendrix spacer cable when exposed to the high ac voltage stresses utilized in this project.
There are no industry standards for testing spacer cables. The tests were designed to establish, under
dry conditions, the electrical compatibility of spacer cable covering materials and the materials that
are used for spacer cable support devices such as spacer insulators and pin type insulators. In
essence, these tests assess the similarity (or dissimilarity) of the dielectric characteristics of the cable
covering(s) and the material used for the insulated support devices. To make the assessment in a
reasonably short amount of time, the application of voltages that were near the flashover voltage of
the tested components was required. Thus, the test protocol followed is believed to provide some
indication of the compatibility of spacer cables with insulated support devices under dry conditions.
Wet and/or contaminated service conditions were not investigated.
1
Referred to as “spacer cable” in the remainder of this report.
TEST SAMPLES
Covered Conductor (Spacer Cable – Tree Wire)
Southwire provided a reel of Southwire 15 kV, 336.4, 18/1 covered conductor (spacer cable) for the
evaluation. The conductor was 336.4 kcmil aluminum composed of 18 aluminum strands with one
steel strand in the center. The covering was black and composed of an inner layer consisting of 75
mils of HMWPE and an outer layer consisting of 75 mils of TRHDPE. Reference Appendix A.
Southwire also provided a reel of Hendrix 15 kV, 336.4, 19 strand covered conductor (spacer cable)
for the evaluation. The conductor was 336.4 kcmil compact aluminum composed of 19 strands. The
covering was composed of an inner layer consisting of 75 mils of linear low density polyethylene
and an outer layer consisting of 75 mils of high density polyethylene. Although the cable was not
labeled as such, the dimensions and physical characteristics were identical to that of Hendrix
HAC15-IB covered conductor. Reference Appendix A.
1. Hendrix model HPI-35 tie-top, pin type insulator rated 35 kV (Fig. 1),
2. Hendrix model HPI-35 VT (vise-top) pin type insulator with nylon clamping inserts rated
35 kV (Fig. 2),
3. Hendrix model RTL-15 three-phase spacer with ratcheting hanger (all plastic) rated 15
kV (Fig. 3),
4. Hendrix model RTL-15 DM three-phase spacer with metal clamp rated 15 kV (Fig. 4),
5. Preformed Line Products (PLP) tie-top, pin type insulator rated 35 kV (IP-102)(Fig. 5).
700kV Voltage
Divider
Spacer Cable
Insulator
Under
Test
Porcelain
“Control”
Insulator
Spacer Cable
Preliminary test results showed that the insulators would experience a capacitively coupled flashover
in the mid-140 kVac range. This type of insulator flashover did not puncture the spacer cable
covering. The electrical charge was coupled through the spacer cable covering capacitance at the
moment of insulator flashover. Therefore, it was agreed upon with the requestor to proceed with the
dry ac withstand test at a voltage level of 120 kVac. The voltage level of 120 kVac phase-to-ground
represented a test voltage level that was 5.9 times the nominal voltage rating of the insulator and
13.9 times that of the spacer cable. It was further agreed to perform the test for a period of six (6)
hours. No tie-wires were used in the testing so as not to influence the interaction of the spacer cable
and insulators. It was also found, during the preliminary tests, that the spacer cable length had to be
increased to forty (40) feet in order to keep the ends from experiencing tracking down to the
insulator and flashing to ground. Figures 7 through 10 show typical test setups for the pin type
insulators; each utilizing a forty-foot length of spacer cable. Table 1 contains the straight-line
distance from the bottom of the spacer cable insulation, where contact was made with the pin type
insulator, to the ground plane. Table 2 contains the various combinations tested for ac withstand.
Fig. 8 – Hendrix Tie-Top Insulator (left) Fig. 9 – Hendrix Vise-Top Insulator (foreground)
Hendrix HPI-35 Tie-Top, Pin Type Insulator with Southwire Spacer Cable
The Southwire spacer cable tested with a Hendrix model HPI-35 tie-top, pin type, polyethylene
insulator withstood the application of 120 kVac for a period of six (6) hours. The spacer cable
covering and the insulator were inspected for signs of puncture and corona burning; none were
found.
Tested concurrently (in parallel) with the above was a porcelain tie-top, pin type insulator with
Southwire spacer cable. This combination also withstood the application of 120 kVac for the six (6)
hour period. Once again, the spacer cable covering and the insulator were inspected for signs of
puncture and corona burning and none were found.
It was noted, however, that on both combinations, the spacer cable had a surface “fogging” around
the area of the insulator. The spacer cable installed on the porcelain “control” insulator sustained
greater overall surface fogging, which is believed to be oxidation. This indicates that there were
more surface discharges with the porcelain insulator. Note that spacer cable manufacturers do not
recommend the use of spacer cable with porcelain insulators. Figure 11 shows the fogging on the
Southwire spacer cable tested on both insulators.
HPI-35 Tie-Top
Porcelain
Hendrix HPI-35 VT (Vise-Top) Pin Type Insulator with Southwire and Hendrix Spacer Cable
The Southwire spacer cable with a Hendrix model HPI-35 VT (Vise Top) pin type, polyethylene
insulator withstood the application of 120 kVac for a period of six (6) hours without puncture. It
should be noted that, during preliminary test trials, the Southwire spacer cable experienced
capacitive flashover (as discussed earlier) of the insulator at a level of 143 kVac. This was higher
than the Hendrix spacer cable mentioned in the paragraph below, indicating the Southwire spacer
cable had a greater compatibility with this insulator’s dielectric materials.
During the same preliminary test trials, the Hendrix spacer cable experienced capacitive flashover of
the insulator at 114 kVac, and therefore, was tested at the lower voltage of 100 kVac for six (6)
hours. At this lower voltage, the Hendrix spacer cable withstood the voltage application for six (6)
hours without puncture.
There were signs of corona burning on the covering of both spacer cables at locations where the
insulator waffle grips cut into the cable covering. According to the manufacturer, the waffle grip
insert was composed of nylon and was non-conductive. The following figures show these results
and were typical for both spacer cables. The insulator showed no signs of damage.
Corona burns can be seen in Figures 13 through 16 along with some localized melting of the cable
covering material shown in Fig. 13 (left). The markings caused by corona burning (partial
discharges) are identified as “feathery” patterns oriented perpendicular to the waffle grip indentions.
Figure 13 and Figure 14 – Vise-Top, Waffle Grip Indentions with Corona Burns
Figure 15 and Figure 16 – Vise-Top, Waffle Grip Indentions with Corona Burns
Figure 17 and Figure 18 – Vise-Top, Waffle Grip Indentions with Corona Burns
In Figures 17 and 18, the waffle pattern indentions are shown at lower magnification. Figure 17
(left) shows one particularly deep indention. This model of vise-top, pin type insulator was
equipped with two break-off torque control head plastic bolts (Hendrix B-1 Torque Bolt) which
control the tightness of the grip. During installation, for these ac withstand tests, the bolts were not
torqued enough to cause the head to come off. Therefore, during a field installation in which the
bolts would be fully torqued, the indentions could be even deeper.
Both the Southwire and Hendrix spacer cables experienced the same surface fogging, described
earlier, when tested with the Hendrix VT insulator. However, the Southwire spacer cable installed
on the porcelain “control” insulator sustained the greatest overall surface fogging. This indicates
there were more surface discharges with the porcelain insulator, which is to be expected due to the
greater difference in dielectric constants. Note, again, that spacer cable manufacturers do not
recommend the use of spacer cable with porcelain insulators. The following figures show the spacer
cables tested with the Hendrix VT and porcelain insulators.
HPI-35 Vise-Top
Porcelain
Figure 20 – Hendrix Spacer Cable Surface Fogging and Waffle Cuts (Side 1)
Figure 21 – Hendrix Spacer Cable Surface Fogging and Waffle Cuts (Side 2)
Further partial discharge (PD) measurements were performed with both spacer cables on all the pin
type, polymeric insulator samples. This time, however, inception and extinction were included on
the same graph for each combination. The following table contains these results while the apparent
discharge characteristic graphs are provided in Appendix D. The partial discharge level used here
for defining inception and extinction was 5 pC. Values were rounded to the nearest one-half
kilovolt.
Table 4 – PD Inception and Extinction for Spacer Cable with Pin Type Insulators
Insulator Inception Extinction
Type Spacer Cable Type (kV) (kV) Comment
not measured 17.5 hand tight
Southwire
Hendrix not measured 17.5 torqued
Vice-Top 29 27.5 hand tight
Hendrix
30 27 torqued
Hendrix Southwire 21.5 21
Tie-Top Hendrix 22 19.5
PLP Southwire 50.5 48.5
Tie-Top Hendrix 49.5 47
The partial discharge inception and extinction levels between the Southwire and Hendrix spacer
cables shown in Table 4 were quite similar, with the exception of the Hendrix vice-top insulator. All
of the extinction levels were higher than the nominal phase-to-ground voltage for a 15 kV system;
however, not so for a 35 kV system. Keep in mind that the spacer cable was rated for a 15 kV
system and the pin type insulators for a 35 kV system. The PLP tie-top, pin type insulator was more
compatible with both types of spacer cable as the inception and extinction voltages were more than
twice those of any other combination.
Dielectric Test – Properties of the Hendrix Vice-Top, Pin Type Nylon Waffle Grip Insert
During installation, the Hendrix nylon waffle grip inserts cut waffle (criss-cross) patterns into the
covering of both types of spacer cables. This action provides a firm grip, keeping the covered
conductor from slipping through the grips when mechanically loaded. The inserts also caused
corona burning on both cables’ surfaces. Because of this there was a desire to test the insert for its
dielectric properties. To this end, two samples were sent to the Georgia Institute of Technology
School of Materials Science and Engineering to measure the dielectric constant and dissipation
factor. One sample was machined flat on all sides, while the other was left intact (virgin sample).
The full report along with email correspondence is contained in Appendix E.
The effective dielectric constant of the nylon insert ran between 17 and 20. This was quite high
compared to the approximate 2.3 for the polyethylene of the covering material and insulator.
Because the materials have such a wide difference in dielectric constant, electric field between the
spacer cable conductor and the grounded insulator pin would be distorted. This most likely
contributed to the corona burning found on the spacer cable coverings.
The dissipation factor of the machined smooth nylon insert sample was 23.4%. A virgin sample,
with no machining of any surface and using very small electrodes placed within the waffle pattern,
had a dissipation factor of 15.6%. Professor Gerhardt of the Georgia Institute of Technology School
of Materials Science and Engineering indicated that the inserts must have a conductive filler material
of some kind, judging from the readings obtained. A conductive filler material would explain the
high dissipation factor measurement of the nylon material. However, the conductive filler particles
must not have been in physical contact with each other as the resistivity of the material was very
high.
Spacers
Both types of polyethylene spacers submitted for evaluation were tested simultaneously on a
simulated three-phase circuit. This way both spacers would have the same voltage applied during
the test. Figures 23 and 24 show the general test setup for both spacer types.
The two spacers were installed on the messenger wire two feet off center, which located them
located four feet apart. Since it was difficult to maintain a straight, horizontal lay with the spacer
cable, two additional (one of each type) spacers were located outside the test spacers at a distance of
approximately four feet to the outside. These additional spacers were not considered part of the test
since the spacer cables were not perpendicular to the spacers.
The messenger wire was comprised of two, 5/8-inch copper-clad ground rods connected in-line with
a compression connector. The messenger was elevated approximately ten (10) feet above the ground
plane.
Messenger
Phase
Conductors
RTL-15 DM RTL-15
The test voltage was applied to one of the side-phase conductors, while the other two phases and the
messenger were grounded. This configuration simulated both phase-to-phase and phase-to-ground
voltage stresses. The same Biddle 700 kVac Series Resonant Test Set, rated 2A full load
continuous, was used for the voltage source. Previous trial testing showed that a capacitively
coupled flashover (as explained earlier with the pin type insulators) would occur around 85 kVac, on
the spacer with the metal clamp (RTL-15 DM). The flashover occurred between the energized side-
phase conductor and the opposite side-phase conductor (grounded) and, at the same time, between
the energized side-phase conductor and the grounded metal clamp. Therefore, a test voltage level of
75 kVac was selected. This represented a test voltage level that was 8.7 times the nominal phase-to-
ground voltage and 5.0 times the nominal phase-to-phase voltage.
The Southwire spacer cable in combination with both types of spacers withstood the application of
75 kVac, in this configuration, for a period of six hours without puncture or burning. The spacer
cable insulation was inspected under 5X magnification for signs of corona burning and none was
found. The spacers were inspected with the naked eye and no damage was found. In addition, the
spacer cable did not exhibit the overall surface fogging (oxidation) as it did with the pin type
insulators. This was indicative of little to no surface discharges during the test, which was due to
greater distances between electrodes and lower applied voltage than with the pin type insulator tests.
Figures 25 through 28 show the spacer cable and spacer surfaces after the ac withstand test.
CONCLUSION
No damage was evident after the ac withstand test at 75 kV among the Southwire spacer cable and
Hendrix three-phase spacers tested. There was also no damage evident after the ac withstand test at
120 kVac among the Southwire spacer cable and the PLP and Hendrix tie-top, pin type insulators,
with the exception of some surface “fogging” on the spacer cable, which is thought to be surface
oxidation.
At the voltages used in this test program, the Hendrix VT (Vise-Top) pin type insulator caused
corona burning along indentions cut, by the waffle grips, into both the Southwire and Hendrix spacer
cable coverings. This was in addition to fogging of both spacer cables. The Southwire spacer cable
with the Hendrix vice-top, pin type insulator was capable of operating at 120 kVac for the full
duration of the ac withstand test. However, the Hendrix spacer cable with the Hendrix vice-top, pin
type insulator was only capable of operating at 100 kVac for the full duration of the ac withstand
test. With this model insulator, the Southwire spacer cable experienced capacitive coupled flashover
of the insulator at 143 kVac, while the Hendrix spacer cable experienced the same flashover at only
114 kVac. This indicates that the Southwire spacer cable was more compatible with the Hendrix
insulator materials than the Hendrix spacer cable was.
The partial discharge inception and extinction levels for the Southwire and Hendrix spacer cables
were quite similar for all combinations tested, with the exception of the Hendrix vice-top, pin type
insulator, in which the Southwire spacer cable levels were lower than the Hendrix spacer cable.
However, all of the extinction levels were higher than the nominal phase-to-neutral voltage for a 15
kV system. Note that the spacer cable was rated for a 15 kV system and the pin type insulators for a
35 kV system. The PLP tie-top, pin type insulator appeared to be more compatible with both types
of spacer cable as the inception and extinction voltages were more than twice those of any other
combination.
Two Hendrix VT nylon waffle grip inserts were sent to the Georgia Institute of Technology’s
Materials Science and Engineering Department for dielectric evaluation. The effective dielectric
constant at 60 Hz ran between 17 and 20. This was quite high compared to the approximate 2.3 for
the polyethylene spacer cable coverings and insulators. This high difference in dielectric constants
would distort the high voltage electric field between the spacer cable conductor and the grounded
insulator pin. This most likely contributed to the corona burning found on the spacer cables’
coverings. The measured dissipation factor of the waffle grip inserts was 23.4%. It was also noted
that there was conductive filler material within the nylon; although, the resistivity was quite high.
Corona burns on both spacer cables installed on the Hendrix vise-top, pin type insulator were due to
the indentions formed into the covering by, and the high dielectric constant of the waffle grips. It
must be noted that this occurred at a greatly accelerated voltage. Partial discharge tests performed
on both spacer cables with the Hendrix vice-top, pin type insulator showed that the extinction
voltage was higher than the nominal rated phase-to-ground voltage of both spacer cables (8.66 kV).
In addition, the application of accelerated voltage resulted in greater surface fogging on the spacer
cable installed on the porcelain “control” pin type insulator than on any other spacer cable installed
on the polyethylene insulators and spacers. This indicated there were fewer surface discharges when
the spacer cable covering and support had similar dielectric constants. The spacer cable coverings,
pin type insulators, and spacers were composed of polyethylene which are assumed to have a
dielectric constant of 2.3, while porcelain has a dielectric constant between 5.1 and 5.9 2. Because of
this difference between dielectric constants, spacer cable manufacturers do not recommend the use
of porcelain insulators with spacer cable.
The results indicate that Southwire, 15kV, 336.4, 18/1 spacer cable is compatible with the Hendrix
tie-top, pin type insulator, Hendrix three-phase spacers, and the PLP tie-top, pin type insulator when
exposed to the high ac voltage stresses utilized in this project. In addition, the Southwire spacer
cable was equal to or more compatible with the Hendrix vice-top, pin type insulator than the
equivalent Hendrix spacer cable when exposed to the high ac voltage stresses utilized in this project.
EQUIPMENT
• Biddle 700kV Series Resonant Test Set – kilovoltmeter, CQ2102
• Biddle Partial Discharge Detector Model 27000, CQ2163
2
The ARRL Electronics Data Book, Second Edition, published by the American Radio Relay League, 1988
APPENDIX A
Specifications
APPENDIX B
APPENDIX C
Hendrix Spacers
APPENDIX D
Figure 30 – Hendrix Spacer Cable with Hendrix Vice-Top Insulator – Torque Bolts Hand Tight
Figure 31 – Hendrix Spacer Cable with Hendrix Vice-Top Insulator – Torque Bolts Torqued
APPENDIX E
(Blank Page)