In May 2004, Los Angeles Department of Water and Power (LADWP) completed installation of its second 230-kV transmission line, setting several underground high-voltage records. The high-voltage expansion, which runs from the Toluca Lake Receiving Station to the Van Nuys Receiving Station, is 5 miles of 230-kV cross-linked polyethylene (XLPE) cable circuit, similar to a 4.5-mile-long 230-kV XLPE circuit — the Toluca-Hollywood line — that was completed in May 2003. Both high-voltage underground installations marked U.S. distance records and also industry firsts in North America for a municipal utility installing XLPE cable.
The innovative cable is composed of a new XLPE plastic-type insulation, manufactured by VISCAS Corp. (Tokyo), owned by Furukawa Electric Co. Ltd. and Fujikura Ltd. Embedded in the cable jacket is a fiber-optical and temperature-sensing device designed to perform continuous monitoring of the high-voltage cable. Sensing devices send real-time data back to substations pinpointing hot spots, which reduces the risk of outages.
The use of XLPE cable systems at voltages above 69 kV is becoming more prevalent as manufacturing and material improvements are making them more reliable. There are several XLPE installations now in service worldwide for circuits up to 500 kV.
Cable construction consisted of 1000 kcmil compact round copper conductor, 1060 mil of XLPE, 60 copper shielding wires, two stainless-steel tubes for fiber opticals, 140 mil lead sheath and 180 mil MPDE jacket. The cable and accessories were manufactured in Japan.
In the past, LADWP has used oil-filled cable systems at 138 kV and 230 kV. Because of environmental concerns, as a result of spills and leaks and maintenance issues associated with oil-filled cables, LADWP switched to XLPE-insulated cables in the late 1990s.
To test HV oil-filled cables, it is common to use dc-voltage testing and oil testing for both commissioning and maintenance. However, with XLPE-insulated cable systems, it appears that ac-voltage tests and partial discharge (PD) tests are more effective for both commissioning and maintenance testing. Distributed temperature measurements are also being used more frequently with XLPE cables for hot spot location and ampacity calculations.
The majority of utilities, including LADWP, hire consultants to conduct such tests because interpretation of results requires both advanced training and skills generally not available in-house. The necessary equipment for XLPE testing is also generally not available in-house.
LADWP maintains testing divisions consisting of several electrical engineers and approximately 30 testers. These personnel engage in tests for corrosion control, power quality and station tests. For the commissioning and maintenance testing on its high-voltage XLPE cables, LADWP also chose to outsource, due to the lack of training in interpreting results and the lack of equipment for conducting PD tests.
PD testing is an evolving technology for periodic diagnostic testing of XLPE-insulated cables. Partial discharges occur at voids in insulation and at the interface layers between cable and accessory insulation. These discharges emit broadband radiation in the range of 50 kHz to 500 MHz. PD testing is a nondestructive testing method, generally accepted as one of the most effective techniques for locating defects in XLPE cables.
LADWP has conducted PD testing on both 138-kV and 230-kV XLPE cable installation for commissioning and maintenance purposes.
Due to the difficulty of interpreting test data and the lack of PD test equipment available in-house, LADWP has also outsourced testing of its new 230-kV installations to PD testing companies. LADWP utilizes the services of PD testing companies from Europe, North America and Japan.
One or more technical personnel from the testing company conduct the testing with the support of two or more LADWP personnel, depending on the location and type of testing.
LADWP has conducted both on-line and off-line tests. Off-line tests have been conducted with the circuit energized from high-voltage variable frequency (VF) source, while on-line tests are done with the circuit energized from system voltage.
There is considerable discussion and debate in the utility industry and amongst PD testing agencies over these testing methods. Some maintain that on-line testing at rated voltage is sufficient in detecting defects in cables and accessories and causes no damage.
On the other hand, others feel that off-line tests at higher-than-rated voltage are more effective because the partial discharge inception voltage (PDIV) is higher than the voltage at which the discharges extinguish; as a result, manufacturing or installation defects that normally would not ionize at rated voltage can be detected. Additionally, the higher voltage may detect incipient faults. Also, factory testing of cables and accessories is conducted at higher-than-rated voltage.
For the first 230-kV installation on the Toluca-Hollywood Line 3 in Los Angeles, a modular inductor VF source was used for off-line tests and the applied voltage was 190 kV phase to ground. For this case, terminal type tests at terminations and limited localized tests at joints were conducted.
The terminal tests required that a blocking impedance and a coupling capacitor be connected to the cable termination for the circuit phase under test.
The blocking impedance device connected in series with the cable termination. The blocking impedance functions by blocking high-frequency electrical noise from the source or high-voltage connections.
The coupling capacitor is connected in parallel with the cable termination. The coupling capacitor acts as a high pass filter that passes high-frequency signal from the cable and blocks lower frequencies close to that of the excitation voltage. The capacitor, along with the detection impedance, couples PD pulses from the cable system to a microprocessor-based measuring unit for analysis.
A voltage divider measures applied voltage and provides phase-angle information to the measurement equipment.
To reduce electrical noise, a toroid was connected to the top of the termination and high-voltage connections were shielded with flexible metal air duct.
The test was conducted by three field engineers with the support of LADWP personnel for equipment setup and connections.
The test procedure consisted of the following:
A time domain reflectometry test that injects a 20-ns to 500-ns pulse to determine cable profile in terms of discontinuities or joint locations.
Injection of calibrated pulses of increasing pC magnitude in conjunction with noise mitigation measures. This step determines the sensitivity level for the test, which in this case was found to be about 10 pC.
Noise mitigation before high-voltage test. Ambient or background noise was measured, and appropriate filtering was used to reduce noise. There was also noise generated by the power electronics in the frequency converter and noise discharges in air from equipment.
While the HV tests and the off-line PD terminal-type measurement were conducted, localized measurements at joints were also made. These PD measurements were used to compare with PD measurements at the same joints under on-line testing conditions. No internal PD was detected in either off-line or on-line tests.
LADWP has been contracting with several PD testing agencies from Europe, North America and Japan since 2000 to conduct on-line PD measurements of joints and terminations for 138-kV and 230-kV cable installations.
For commissioning of the 230-kV circuit installation, on-line PD measurements were made at joints and terminations. No internal PD was detected.
Testing agencies from Europe and North America employ one-channel testing equipment that uses radio frequency current transformer (RFCT)-type inductive sensors, while Japanese testing agencies employ four-channel equipment using capacitive foil electrodes.
The advantage of a four-channel measurement system is that measurements at the three joints in the manhole can be made concurrently. Three channels are used for the PD measurements, while the fourth channel is used for calibration pulses and also to record noise measurements at the test location.
With single-channel PD equipment, individual measurements have to be made at each of the joints at different times. The RFCT sensor is placed around a jumper connected across the cable joint because the joints have an insulating flange as a shield break.
LADWP also conducted distributed temperature measurements (DTS) on the 230-kV cables. The XLPE circuits contain two multi-mode and two single-mode fibers. The fibers are enclosed in two stainless-steel tubes placed under the lead alloy metallic covering. The DTS measurements were made using the multi-mode fibers. Due to the lack of DTS equipment in-house, this work was outsourced to an engineering consulting company.
The work required one day for setup and for measurements. An engineer from the consulting company conducted the measurements, while LADWP personnel assisted with setting up the equipment in the substation.
The measurements were made using a model DTS-800 instrument for the entire length of the circuits, which were approximately 4.5 miles and 5 miles long each.
The DTS-800 testing instrument measures the temperatures by injecting 10-ns wide light pulses in the fibers. All measurements were single-ended and, thus, made by injecting the light pulses into one end of the fiber. In turn, the backscatter from impurities or “dopants” in the glass fibers is used for the temperature measurements. Using the multi-mode fibers for the measurements, a resolution of 1 m and an accuracy of ± 1°C were obtained.
Because of the length of the circuits double ended measurements were not made. For this case, measurements require connecting two ends of a fiber loop to the DTS-800. Due to the length of the circuits, fiber loops would exceed 10 miles when taking into account the lay of the fibers and thus exceed the measurement capability of the DTS-800 which was limited to 7 miles. Plus the additional fiber-optical splices along the loop would provide for additional optical losses.
The warm and cold spots were related to the location of manholes found through optical measurements. The manholes were identified by their fiber-optical splices and also by the ratio of the length of the fiber to the length of the cable, which was approximately 1.28. The fibers are applied in a helix pattern over bedding tapes placed around the extruded insulation shield.
The hottest spot — 29°C (84°F) — was between MH-9 and MH-10 (see diagram on page 48E), where the cable and ducts are installed in air in a compartment of a concrete bridge that crosses the Tujunga Wash Channel. The width of this warm spot is approximately 165 ft.
The coolest spot (between MH-15 and MH-16) was found where the cables are installed, directly under the Ventura Freeway overpass. This is due to lower ambient soil temperature at increased installation depths and shading from the sun.
The thermally limiting sections were found where the transmission cables cross distribution cables duct banks at the intersection of Burbank Blvd. and Lankershim Blvd. or the intersection of Burbank Blvd. and Satsuma Ave. At these crossings, the mutual heating between the distribution and transmission cables caused an increase in soil ambient temperature of about 4°C (39°F).
LADWP currently operates approximately 9.50-circuit miles of 230-kV and 2.2-circuit miles of 138-kV XLPE-insulated cables. However, there are two circuits planned at 230 kV that are due in service in 2009. Additionally, LADWP has an ongoing program to replace 70-circuit miles of 138-kV oil-filled cable with XLPE-insulated cable, including replacement of a circuit approximately 5 miles, which will be completed in June 2005. For these reasons, LADWP's test engineers will continue to perfect methods for commissioning and maintenance testing of high-voltage XLPE cable.
Vincent Curci has been an electrical engineer at the Los Angeles Department of Water and Power since 1983 and is currently in charge of underground transmission design. Curci, who holds a BSEE and two MSEE degrees, is a member of IEEE and also participates in the Insulated Conductors Committee of IEEE and EPRI Underground Transmission Task Force. Vincent.Curci@ladwp.com
Hassan Motallebi has been an electrical engineer at the Los Angeles Department of water and Power since 1993, working in the areas of control, automation, SCADA, and Underground Transmission. Motallebi who holds a BSEE and a MSEE degree, is a member of IEEE and participates in the Insulated Conductor Committee (ICC) of IEEE. Recently, he managed Toluca-Van Nuys Cable D project for LADWP, which he presented at ICC meeting held in Montreal, Canada, in May 2004. Hassan.Motallebi@ladwp.com
Kishan Kasondra is an electrical engineer associate at the Los Angeles Department of Water & Power, which he joined in May 2002. His responsibilities include designing, installing, inspecting, testing and maintenance of high-voltage underground transmission lines. Kishan is currently managing a cable project to replace a 4.5-mile long SCFF cable circuit with an XLPE cable system. He received a BSEE degree from Loyola Marymount University and is attending the University of Southern California to complete a MSEE degree specializing in power systems. He is a member of IEEE. Kishan.Kasondra@ladwp.com
The temperature measurements for the Toluca-Van Nuys Cable D were taken before the cable was energized and loaded, as a baseline, and afterwards, when carrying a predetermined load for a period of two weeks.
|Concrete Duct bank dimensions||36 × 36||inches|
|Conduit type||Fiberglass (FRE)||—|
|Conduit inside diameter||6.4||inches|
|Spacing between cables||17.5||inches|
|Concrete thermal resistivity||80||°C-cm/W|
|Backfill thermal resistivity||120||°C-cm/W|
|Native soil thermal resistivity||150||°C-cm/W|