The predominant paper-insulated, lead-covered (PILC) cable, at sizes 350 kcm and 800 kcm copper, has served Ameren well throughout the years. Ameren is headquartered in the downtown area of St. Louis, Missouri, U.S., where it has about 3300 circuit km (2050 miles) of PILC cable installed. About 80% of the PILC cables are rated 15 kV or less; the remaining 20% are rated 35 kV. Regardless of the rated voltage, the PILC cables are generally sector-type so that three individual cables can be contained within one lead sheath.

However, changes such as deregulation and environmental constraints have prompted the evaluation of current construction practices. Labor costs, material costs and the availability of skilled workers also affect the utility. All of these issues played a part in the decision to re-evaluate the way the network system was expanded and maintained.

A review of purchased materials at Ameren revealed that changing the network cable design from PILC to an insulated rubber would have the greatest overall financial impact of any material being used on the network system. Ameren concluded that primary PILC cable, while reliable, is expensive to replace or repair. Because of the small-diameter ducts now installed in Ameren's service territory, the company realized it would have to consider reduced-wall cable.

Reduced-Diameter Cables

Reduced-diameter cables, along with higher operating stresses, have steadily gained popularity among major electrical utilities. Initial studies have demonstrated the suitability of ethylene propylene rubber (EPR)-insulated cables rated between 15 kV and 35 kV to operate at as much as 4-kV/mm (101.6 V/mil) stress at the conductor shield/insulation interface. EPR insulation has been found to be particularly suitable due to its ability to operate at higher temperature and in harsh conditions.

When investigating possible cable designs, Ameren took into account duct size, operating voltage, conductor size and neutral size. The majority of the installations in existing Ameren ducts necessitated the use of compact copper conductors with flat strap neutrals. Individual EPR cables are triplexed to facilitate installation packaging and handling.

In 1991, Ameren began using a reduced-diameter 35-kV cable. Initially, the cable purchased was a triplexed 750-kcm compact copper with 7.6 mm (300 mil) of EPR insulation, 3.4-kV/mm design stress (86 V/mil), tape shield and 1/0 AWG copper neutral. This cable was used as a direct replacement for the 35-kV, 800-kcm PILC sector cable.

As usage of the reduced-diameter 35-kV cable increased, Ameren experienced problems pulling the cable into the small ducts. It seems the 1/0 AWG neutral would sometimes cause the cable to jam. Ameren also experienced problems with the tape shield under certain fault conditions. To eliminate future problems, the tape shield and 1/0 AWG neutral were replaced with flat strap neutrals in 1997. After these changes were made, no installation problems or cable failures occurred. Through 2004, Ameren has installed approximately 273 linear km (170 miles) of the reduced-diameter 35-kV cable.

During the late 1990s, Ameren was under pressure to reduce costs and improve reliability. The utility was also confronted with health and environmental issues related to the handling of PILC cables. Additionally, the number of trained lead cable splicers was decreasing; therefore, a project to further limit the use of PILC cables was undertaken.

The success Ameren encountered replacing 35-kV PILC cable with reduced-diameter EPR cable led the company to consider the same strategy with 15-kV cables. This would reap great financial benefit because 80% of the PILC cables are rated 15 kV and below. In 1998, Ameren launched an effort with Pirelli Power Cables and Systems N.A. to develop reduced-diameter 15-kV EPR cables.

For this plan to be economically successful, cables were needed that could be used to replace the 350-kcm and 800-kcm PILC cables. This was going to be a challenge because the replacement cables would have to be installed in 3- and 3.5-inch (7.6- and 8.9-cm) ducts, respectively. Some major reductions in the overall diameters of the individual cables were required to obtain triplexed cable assemblies small enough to fit into existing ducts. Ameren also worked with 3M, Raychem and Elastimold to develop manufactured accessories to do away with the need for lead wipes.

Two reduced-diameter EPR cables were designed for Ameren: a 350-kcm triplexed cable with a maximum assembly diameter of 63.5 mm (2.5 inches) and a 750-kcm triplexed cable with a maximum assembly diameter of 82.6 mm (3.25 inches). The only deviations from industry standards were the thicknesses of the various layers. The first orders for 2742 linear m (9000 ft) each were placed in 1999.

Just having a reduced-diameter replacement cable was not sufficient. Reliable accessories that could be used with the newly designed cable also were needed. To this end, Ameren arranged for testing of the cable and various premolded accessories. Testing the premolded accessories on the reduced-diameter cable revealed several significant results. Performing the standard industry accessory tests revealed that the higher electrical stresses associated with the reduced-diameter cable caused slip-on premolded accessories to fail corona tests. The testing also revealed that both cold-shrinkable and heat-shrinkable accessories would pass all electrical tests. When the assemblies were tested to failure, the standard test levels could be exceeded significantly and all failures occurred in the accessories. Mechanical testing on the reduced-diameter designs indicated that they could be handled in a manner similar to the standard 15-kV EPR cables.

PILC Replacement Program

Armed with the electrical and mechanical test results, Ameren decided to implement a program to replace PILC cables whenever possible. The overall program involved the use of standard- and reduced-diameter EPR cables. If the ducts were large enough for standard-diameter cables, these would be used. If the ducts were too small for the standard-diameter cables, the reduced-diameter designs would be used. Whenever possible, a failed PILC cable would be replaced with an EPR design. Whenever a pothead fails, the failed pothead and PILC cable would be removed and replaced with an EPR design and cold-shrinkable terminator. Heat-shrinkable splices would be used for joining reduced-diameter cables in straight splices and for joining reduced-diameter and standard-diameter cables to PILC cables in trifurcating transition splices.

To improve overall efficiency and reduce inventory, Ameren worked with accessory manufacturers to develop cold- and heat-shrinkable terminators and splices that could be used with both the standard-diameter and reduced-diameter cables.

As experience was gained with the reduced-diameter cables and associated accessories, some changes were made. Specifically, the cable design was modified. In 2001, the encapsulating jacket material on both reduced diameter cables was changed to polypropylene and the thickness of the jackets was reduced to a 0.64-mm (25-mil) minimum point. This change was made to take advantage of the toughness and lower coefficient of friction associated with the polypropylene material. With the reduced thickness of the encapsulating jackets, it was possible to increase the insulation thickness on both reduced-diameter cables to 3.68 mm (145 mil) minimum average. These changes were not made because of any problems associated with the initial cables. The changes were to enhance the initial cable designs and to allow Ameren to maintain the overall cable diameters, reduce pulling tensions, reduce electrical stress and use existing premolded accessories.

Accessory limitations must be considered when designing high-stress EPR cable systems. The limitation with existing accessories has been found to be 2.4 kV/mm (61 V/mil) at the insulation/insulation-shield interface. This affects only the 15-kV class cables. Further insulation reduction for 15 kV class could be achieved with improvements to the present limitations of accessories.

More recently, premolded splices have been successfully tested on the reduced-diameter cables. The key is to use a premolded accessory that does not wipe the silicone grease away from the insulation shield cutback during installation. This update is significant because it will allow more flexibility in the replacement of PILC cables. The use of premolded “Y” splices and premolded “H” splices will now be possible.

Through 2004, Ameren installed approximately 61 linear km (38 miles) of reduced-diameter cable as replacement for 15-kV class PILC cables. Although the quantity of replacement cable may seem small, its economic impact has been significant. Not only have nearly 22 linear km (14 miles) of existing duct been reused, but also the time required to make splices and terminations has been reduced by almost half. No problems have been experienced with the reduced-diameter cables, and no problems are expected. Ameren's plan is to continue using EPR cable to replace PILC cable whenever possible.

Because there is a delicate balance that must be maintained between mechanical protection for installation and electrical integrity, consideration must be given to both insulation and jacket thicknesses. Protection is important as existing ducts may be partially collapsed, have misaligned sections, and/or have broken areas that can damage the replacement cable.

Editor's Note

Ameren and Pirelli Power Cables and Systems N.A. worked closely to develop a comprehensive cable testing program.

Harry L. Hayes III is a consulting engineer in the distribution standards group of Ameren Corp. (formerly Union Electric Co.), which he joined in 1979. Hayes is a senior member of IEEE/PES and serves as vice chairman of IEEE-ICC discussion group A16D (characteristics of EPR cables). Hayes also serves on the ANSI C119 Connector Committee, is the vice chairman of IEEE 386 Standard for Separable Insulated Connector Systems for Power Distribution Systems Above 600 V, and he is the secretary/vice chairman of the Association of Edison Illuminating Companies Cable Engineering Committee. Hayes received his BSEE degree from Washington University in 1978, and his MBA and MA finance degrees from Webster University in 1983. hhayes@ameren.com

Accelerated Cable Life Testing

Accelerated Cable Life Testing (ACLT) assesses the ability of cables to operate in a wet environment.

In this test program, both standard-wall and high-stress design EPR-insulated cables were aged simultaneously under identical conditions in the same tank.

For ACLT tests:

• Conductors were kept dry by using filled strand.

• Cables were tested with overall polyethylene jackets.

• Cable constructions and compounds (same lots) were identical except for insulation wall thickness.

• Cable samples were 15-kV rated.

• Aging temperature was 90°C (194°F) conductor temperature.

• Load cycle was 8 hours on, 16 hours off.

• Aging voltage was 3 Vg (26 kV) continuous.

The 90°C conductor temperature was chosen because it represents the relatively normal ampacity loading conditions expected for cable installed on such a system. An electrical aging voltage of three times the normal operating voltage provides an average stress of 10.2 kV/mm (260 V/mil) on the high-stress design cable and 5.9 kV/mm (150 V/mil) on the standard-wall cables tested. This voltage level, which corresponds to the AWTT level, provides reasonable accelerated aging without excessive stresses applied to the cable.

The test was carried out to six years of actual aging time, although the test program spanned seven years' calendar time. At the completion of the actual aging period, no failures had occurred on the standard or high-stress cables. All cables were then subjected to high-voltage time testing (HVTT) to breakdown. The procedure was 17.5-kV initial voltage held for 5 min, and then the voltage was raised in 7-kV steps and held for 5 min at each value, continuing to breakdown. The results of the standard wall cables are shown in the tables.

In addition to ac breakdown and dissipation factor results shown, samples from each design were tested for partial discharge (PD) and were found to be PD-free (less than 5 picocoulombs) up to four times rated voltage to ground.

A tree count (which provides an indication of the degradation of the insulation) was conducted on the surface of 50 wafers from each design after the six-year ACLT test. Ten wafers were taken from five cable samples from each design. No significant differences were noted between the standard wall and high-stress design cables. The tree counts represent only bow tie trees as no vented trees were found.

Standard Wall Cable Results

	Breakdown Level			Power Factor @ Room Temperature
Sample No.	kV	kV/mm	V/mil	Percent
1	73.5	16.5	420	N/A
2	73.5	16.5	420	N/A
3	80.5	18.1	460	0.263
4	87.5	19.7	500	0.245
5	66.5	15.0	380	0.274
6	80.5	18.1	460	0.275
7	73.5	16.5	420	0.263
8	87.5	19.7	500	0.261
9	52.5	11.8	300	0.302
10	73.5	16.5	420	0.306
Average	74.9

High-Stress Design Cable Results

	Breakdown Level			Power Factor at Room Temperature
Sample No.	kV	kV/mm	V/mil	Percent
1	94.5	37.2	945	0.237
2	73.5	28.9	735	0.239
3	87.5	34.4	875	0.246
4	80.5	31.7	805	0.247
5	52.5	20.7	525	0.247
6	87.5	34.4	875	0.268
7	87.5	34.4	875	0.236
8	73.5	28.9	735	0.219
9	66.5	26.2	665	0.217
10	80.5	31.7	805	0.216
Average	78.4

Standard Wall Cable Tree Count
Total Number of Trees Sized in Mils Per Size Range for 50 Wafers

	1 to 5	6 to 10	11 to 15	16 to 20	21 to 25
50 wafers	16	8	1	1	1
1 sq inch	0.51	0.25	0.03	0.03	0.03
50 wafers = 31.6 sq inches

High-Stress Design Cable Tree Count
Total Number of Trees Sized in Mils Per Size Range for 50 Wafers

	1 to 5	6 to 10	11 to 15	16 to 20	21 to 25
50 wafers	5	7	0	0	0
1 sq inch	0.32	0.45	0	0	0
50 wafers = 15.7 sq inches

Accelerated Water Treeing Tests

Accelerated Water Treeing Tests (AWTT) data were obtained on high-stress EPR design cable, as shown in the photo. High-stress design cables were subjected to three sets of conditions, as shown in the table. The Standard AWTT conditions with water in conductor and outside the cable represent a means for qualification and comparative measure of performance. The AWTT/dry condition with dry conductor and water outside the cable is representative of using a solid or filled-strand conductor to restrict the entrance of water. The AWTT/dry/hi temp condition with a dry conductor and water placed outside the cable and tested at 110°C (230°F) represents the application of solid or filled-strand conductor operating at a normal temperature up to 105°C (221°F). Throughout AWTT conditioning, a continuous test voltage of 26 kV (3 Vg) was applied under all conditions. For the #1/0 AWG -2.54-mm (100 mil) wall EPR cables, this represents an average electrical stress of 10.2 kV/mm (260 V/mil) as compared to 5.9 kV/mm (150 V/mil) for standard 4.45-mm (175 mil) wall cables.

Conditions for High-Stress AWTT Tests

	Condition 1	Condition 2	Condition 3
	Standard AWTT	AWTT/Dry	AWTT/Dry Hi Temp
Cyclic aging	130°C	130°C	140°C
Load cycling	Yes	Yes	Yes
Tc (stress cone)	90°C	90°C	110°C
Ts (tube center)	45 ±3°C	45 ±3°C	55 ±3°C
Test voltage	26 kV	26 kV	26 kV
Water in conductor	Yes	No	No
Water outside	Yes	Yes	Yes