A quick fix overcomes a bus duct failure and keeps the power flowing until the new cable circuit is installed.
Three Years After Inheriting a 30-Year-Old 345-KV SF6 Bus Duct installed on a critical transmission line, Ameren Corp. (St. Louis, Missouri, U.S.) engineers faced significant repair challenges.
Approximately 1100 ft (335 m) of bus duct has operated in the middle of a 30-mile (48-km) line that runs from the Duck Creek power plant to the Tazewell substation in central Illinois, U.S. Critical to the Ameren system, this line provides stability to the Peoria, Illinois, area and is one of two outlets for Duck Creek. Although the bus duct is located at the Edwards plant, the line does not connect to the plant, it simply passes through the switchyard. Aware this bus duct was nearing the end of its useful life, Ameren was in the process of evaluating replacement options when, on June 22, 2006, the 1100-ft section of 345-kV bus duct failed as a result of a lightning strike on the overhead line.
Maintenance crews worked diligently to locate the fault and repair it, but their attempts were unsuccessful due to inadequate test equipment. When Ameren learned an on-site repair was unlikely, the utility quickly switched gears and began investigating other options to put the line back in service. Stringing a single phase of overhead was feasible, so within a week, the conductor was engineered and installed. The 345-kV line was reenergized with one phase overhead (the bus-bypass phase) and two phases through the surviving bus duct. It was a tight squeeze to maintain clearance over the ground and stay safely below the 15 138-kV and 69-kV circuits exiting the Edwards switchyard.
After energizing the circuit via the temporary overhead bypass, Ameren began investigating permanent replacement options that could be put in place within a year. Primary considerations for all alternatives included reliability, schedule and cost. Safety, constructability, outages and operability were factors as well.
Ameren identified several options, including a new gas-insulated system (GIS) bus duct system, a strain bus, a rigid bus, several overhead line solutions and a solid dielectric cable.
The option of a new bus duct was eliminated early in the process. Not only was this the most expensive option, it was also the most difficult to repair. A rigid bus could be installed below the 15 circuits in the Edwards yard, but the risk of a falling line was deemed too great. A strain-bus alternative would span the entire Edwards switchyard, creating a risk to the equipment below it and requiring multiple outages at Edwards. This would be a difficult installation.
Overhead line alternatives were also complicated. Placing a line east, between the plant and switchyard, exposed the generator step-up unit connections to damage from line mechanical failure. A line route west of the switchyard exposed the Edwards exit feeders to similar risk. This design option had one span crossing over six exit lines. Another option was bypassing the switchyard to the east, routing a new line around the plant — crossing the Illinois River — and rejoining the line to Tazewell. The river crossing would have involved wetlands, the permitting for which would have placed timely completion at a considerable risk. The permitting was estimated to take a minimum of two years, and the maintenance group was reluctant to rely on the existing temporary system for more than a year.
Ameren had never installed 345-kV insulated cables. To overcome concerns and develop some familiarity, many meetings were held with several cable manufacturers. The collective result of these first meetings was that solid dielectric cable and accessories are very reliable, and failures, should they occur, would be easy to identify through visual inspection. Installation of a cable tray system would use existing foundations and require minimal outages. And, some manufacturers could provide cable in 10 months, allowing a fall 2007 in-service date.
The transmission planning department approved a 1200-A-rated cable for current system needs, even with a 2000-A line rating. When an increased line rating is needed, a parallel cable can be installed in the bus duct location. This, and the fact the existing overhead bypass phase could remain as a backup for possible cable failure, confirmed the solid dielectric cable option as Ameren's choice.
DESIGNED AROUND SPACE CONTRAINTS
Ameren chose six manufacturers to bid on this project. Given its lack of experience with cross-linked polyethylene (XLPE) cable above 69 kV, the utility relied on these manufacturers to answer questions that arose regarding the cable specification. Ameren selected Prysmian (Milan, Italy) based on its competitive price and other key factors. The relatively short lengths — 1180 ft (360 m) per phase — for this project were a problem for some manufacturers. Prysmian, however, was already producing cable for another utility, so the Ameren order was simply added to that production. Prysmian also had field personnel in the area.
The cable selected was a 1600-mm2 (3158-kcmil) compact segmental copper with a semiconductive tape covering. The conductor shield was 55 mils of super-smooth semiconducting copolymer. The insulation was 1050 mils of super-clean XLPE, and the insulation shield was 60 mils of super-smooth semiconducting copolymer.
A semiconducting water-swellable bedding tape plus crepe paper was applied over the insulation shield. Forty-six copper wires and a 60-mil copper contact tape were applied. The outer jacket was made of a 6-mil copper-foil-laminate radial water barrier with 175 mils of black high-density polyethylene (HDPE) that was graphite-coated. The termination chosen for the project was a 16-ft (5-m)-tall composite-type insulator with a glass-fiber-reinforced epoxy-resin tube and silicon-rubber sheds. Silicon oil is the insulating medium and is pressurized through a separate oil-reservoir tank with an expansion bladder.
Consideration was given to underground conduits or an aboveground cable tray. Ameren wanted to be able to visually inspect the cable in the case of a line fault, and a railroad-track bridge would make conduit installation difficult. Therefore, the cable tray was selected.
Layout of the cable tray system began immediately to allow time for Prysmian's evaluation of its feasibility.
The cable tray design used two 18-inch (457-mm) steel beams with cross members and incorporated the use of existing pilings that were intended for a second bus duct run. This allowed the pile caps to be poured without an outage on the temporary overhead bypass phase. Also, setting the tray and cable between the beams provided physical protection. The design included shielding over each beam in the areas where Edwards' exit circuits crossed the cable to provide further physical shielding in case a line came down. A 36-inch (914-mm)-wide cable tray accommodated all three 6-inch (152-mm)-diameter cables while allowing room for the cables to zigzag along the length of the run.
It was a challenge to complete the tray installation without taking the existing equipment out of service. This particularly affected the cable tray arrangement at each end as well as at the terminations. The bus duct was installed leaving very little room between its ends and the switchyard edge for cable terminations. The wide clearance needed between the lightning arresters and terminations further hampered equipment location. Ameren also wanted to keep the existing operational bus ducts available as a backup, so energized parts to ground spacing was needed from these bushings to the new cable terminations and arresters.
As a result of these stipulations, the final locations for the terminations led to a tray design in which the 36-inch tray split into single 12-inch (304-mm) trays as they approached the terminations. In some instances, these single trays had dramatic bends around existing equipment or in between steel structures, and pushed the threshold of the cable bending radius.
After reviewing the proposed cable tray design, Prysmian suggested some changes for a successful installation. Extra cross beams in the steel design were required for clamping along the cable run, and expansion supports were added on both sides of the railroad track. Termination stands were made taller so the cable would be vertical as it entered the bushing. Finally, additional support was needed on each stand to alleviate pressure from terminator bases.
A Duck Creek power plant outage was scheduled for spring 2007. This was Ameren's opportunity to take the line out of service to install the tray support beams and again later to install the cable. This would allow completion of the fall 2007 schedule. The outages were carefully coordinated, because several other projects were planned around the plant outage.
This necessitated performing foundation work beforehand without a line outage. The energized overhead bus-bypass phase made it impossible to use digging machinery, so excavations were done using soft dig. This work had to be done during February and early March. The ground was deeply frozen, which made soft digging take longer than expected. The ground had to be thawed with hot water as soil was being washed away. Also, because the cable had to be pulled below the railroad bridge, it was necessary to move a wall to make the bend. In spite of these impediments, the work was completed on schedule.
During a subsequent two-week outage, large steel sections were set and termination foundations were drilled. The two 18-inch steel-beam sections were preassembled and quickly set in place. The cable tray also was laid in the beam assemblies during this outage. The steel and tray were grounded to allow the overhead bypass to be re-energized while assembly was completed.
The next step was to piece together the individual tray sections to route the cable ends to the terminations. This was accomplished without an outage. This part of the work was tedious and went slower than anticipated. Concurrently, manufacturing and transportation delays were experienced with the cable equipment. This squeezed the available outage time, pushing up against the summer peak-load months; therefore, the final outage was pushed to fall 2007.
In September 2007, the line was taken out of service for Prysmian crews to do the cable installation. Approximately two days were needed to set up the roller system to guide the cables during each pull, and it took about one day for each cable pull. The cable pulls went smoothly. Separate Prysmian crews performed cable terminations with different crews simultaneously working on each end. The terminations went well, and the cable was energized for the first time on Nov. 16, 2007, effectively removing the problematic bus duct from operation on the system.
A TEAM ACHIEVEMENT
Ameren believes that the cable system at Edwards is top quality. The cable tray and steel led to extra cost and time, but they were necessary for this unique installation. The space constraints on the installation caused many design issues that normally would not be encountered. This was an innovative project for Ameren that tested its engineering prowess.
The authors wish to acknowledge the involvement of the following fellow Ameren staff: Tracy Dencker, Transmission & Distribution Design; Jerry Shelton, Transmission & Distribution Design; Ralph Simmons, EDTS Construction Services; and Michael Welsh, POS Project Engineering.
Luke N. Wollin (LWollin@ameren.com) received a BSEE degree from Southern Illinois University at Edwardsville and a MBA degree from Washington University in St. Louis, Missouri. He joined Ameren in 2003 in the Transmission Substation Design Group and is currently a career engineer in the group. Wollin is also a member of the IEEE Power & Energy Society and a registered professional engineer in the state of Missouri.
Harry L. Hayes III (email@example.com) received his BSEE degree from Washington University and his master's degrees in finance and business administration from Webster University. He joined Ameren Corp. (formerly Union Electric Co.) in 1979 and has held various positions in distribution engineering. He is currently a consulting engineer in Ameren's Distribution Standard Group. His work has included specialized projects in connectors, transformers and cables. Hayes is a senior member of IEEE, an active member of various IEEE working groups and task groups, as well as an active member of various ANSI C119 subcommittees and the AEIC Cable Engineering Committee.