An increase in generating facilities by independent power producers (IPP) has translated into increased capacity on local utility transmission systems to accommodate these new energy sources. Utilities must ensure reliable and efficient delivery of power under normal operating conditions and must prepare their systems to withstand the increased fault duty imposed by the increased generation.

PECO-Energy in southeastern Pennsylvania learned that its Whitpain Substation would be subjected to increased fault duty as a result of the proposed siting of a new IPP facility. Upgrading the station to withstand increased fault duties would involve replacing all 21 of the disconnect switches, all the main bus support insulators for more than 800 ft (244 m) of three-phase bus, as well as the support insulators and bus bars on the two tie bus sections.

Planning the Upgrade

The need to minimize outage time led to a requirement to plan the upgrade during three separate outages of three days each. A feasible approach to fit within the outage windows was based on a previous project undertaken a decade earlier when a 230-kV vertical break disconnect switch with ground switches shipped fully assembled except for the leg structures.

Cleaveland/Price Inc. (Trafford, Pennsylvania, U.S.), manufactured the switch and supervised the installation, which took less than one hour to complete. Planning engineers saw this approach for the Whitpain Substation as a way to meet the time constraints for the outages. Accordingly, a specification was prepared for fully assembled switches — or for single-pole assembled switches, which would need to arrive on the job site in time for pre-assembly prior to the installation outage. Cleaveland/Price offered a fully assembled set of switches, and the installation planning was initiated.

Planning the Installation

PECO decided to keep the design and installation work in-house, based on the familiarity of in-house design and maintenance teams with existing equipment, substation configurations and work practices. PECO's project team met with Cleaveland/Price at Whitpain in late July 2003 to walk the job, to provide drawings and to ensure that the switches would properly match the interface points of foundations and hard bus/cable terminations. All the disconnects were to be of vertical break design in one of three distinct configurations: with one ground switch set; with two ground switch sets; and with one ground switch set, but designed to accept the termination of hop-over bus work at the jaw (or clip) end.

All operating mechanisms would be similarly placed on the various switch types. In addition (based on job site discussions of how PECO planned to work the job), Cleaveland/Price had to consider install ability into its final design, since bus work had to be moved and held out of the way during removal of the old and installation of the new switches. Also, it was necessary to replace all of the bus insulators on the three main bus sections and to replace the bus bar and support insulators on the two bus tie sections.

The existing support insulators were of the cap and pin design, having 5-inch (13-cm) bolt hole circles, both top and bottom. The new requirements for bus through-fault withstand required the use of extra high-strength insulators with 5-inch (13-cm) bolt-hole circles on the top and 7-inch (18-cm) bolt-hole circles on the bottom. To address this issue, Transmission and Substation (T&S) maintenance developed a technique to support several spans of three-phase bus at a time using two hydraulically jacked spanning beams that could be placed 50 ft to 60 ft (15 m to 18 m) apart. The beams then could be used to jack up and support all three phases of bus work over that run and about 20 ft (6 m) of cantilever on either end. To modify the insulator support structures as required by the new 7-inch (18-cm) bolt-hole circles, T&S maintenance and T&S design developed a patterned fixture and clamp system that could be used with a magnetically clamping drill press to generate the bolt-hole pattern required by the new insulators.

The existing bus bar in the bus tie section, which was 4-inch (10-cm) schedule-40 aluminum, presented a problem with respect to its rigidity for the span between support insulators in the tie section. To increase the through-fault withstand capability in this area, the schedule-40 aluminum was replaced with schedule-80 aluminum to avoid the need for an additional set of bus support insulators. The existing cap-and-pin-style insulators on these sections also were replaced, as on the main bus sections. To support this entire effort, T&S design and maintenance worked out the hardware details, ordered parts and scheduled work and craning operations. All outage details were coordinated between T&S maintenance, T&S work management and system operations.

Meanwhile, at Cleaveland/Price, the decision was made to design the legs of the four-legged support structures identically, providing the flexibility of mounting the operating mechanisms on the leg structures without modification. The legs and box frame under the pole units were made by Associated Substation Engineering (Bremen, Georgia, U.S.).

While the structures were being fabricated, the pole units were being built and sub-assembled with the insulators and bases. Cleaveland/Price set aside a significant assembly area where it was possible to fully assemble three-pole 230-kV switches on the structure frame and legs, with all of the operating mechanisms fully mounted. In this configuration, a complete set of alignments were performed on both disconnect and ground switches. Following switch alignment, the legs were removed and prepared for shipment, along with the fully assembled switch. In early September, the PECO project team visited Trafford to inspect the first switch before shipment. Permission was given to ship the first group of seven switches, which arrived at Whitpain a few days later. During unloading, the material was placed strategically to allow removal of the old and placement of the new with a minimum of lifting.

Making the Change

When the first outage began on Oct. 1, 2003, work was scheduled in back-to-back 12-hour shifts. Hydraulic jacks with spanning bars were moved into position to support the main bus while lifting cranes were moved into position to remove the old disconnect switches, install the new legs and switches, and support the unterminated bus during the process. Once the lifting devices were in place, bus work head clamps were removed and the main bus was raised about 1 ft (0.3 m). The bus-side and breaker-side connections to a switch were unbolted, and the switch foundation bolts were unfastened. With the unterminated bus work out of the way, the old disconnect switch and support structure was removed as one piece and moved to an out-of-the-way area where salvage operations were performed. After removal of the old switch, the new switch legs were craned into place and fastened to the existing foundation bolts. The fully assembled switch was then craned into position and bolted to the legs. The slide-collar connections for the operating mechanisms and the bus-side and breaker-side terminations were made up. Once assembled, the base of the legs was shimmed to plumb and the connections were tightened. Over the next three shifts, this procedure was repeated six more times.

While switch replacement was occurring, old bus insulators were removed from the supported bus sections, the support columns re-drilled, and the new insulators and head clamps installed. By the end of the second day, all switches were in place and connected into the system. Making up all the ground connections, operating mechanism ground loops, switch inspections and ductoring occurred during the nighttime periods of the second shifts and the first shift of the third day. No switch adjustments were required, and the bus was returned to service on the third day of the outage. In-service testing consisted of acoustic scanning for elevated corona and thermography scanning for joint hot spots. No anomalies were observed.

The only problem encountered during the first outage was the discovery that tie breaker switches' bus-side terminal pads at the jaw end of the switch were about 2.5 inches (6.35 cm) too short because of a lack of information available on the old switches. This prevented proper connection to the bus-expansion joints. A field rep from the supplier, who was on hand to work with PECO during the first outage, notified his factory, which fabricated new parts for next-day delivery. No problems were encountered during the remaining two outages, which were completed as scheduled.

Conclusion

The virtual rebuilding of the entire primary portion of a 230-kV substation had never been attempted by PECO within the time constraints imposed for the Whitpain project. Optimistic estimates of a more conventional approach would have taken at least twice as long, incurring an increase in transmission congestion charges. The success of the project, in large part, was due to diligent planning, creative design, innovative product availability, continuous vendor cooperation, and a knowledgeable and skilled workforce.

David Lemmerman is a senior engineer at PECO-Energy in the Major Substation Equipment Branch of the Technical Standards Group. He has a BSEE degree from Manhattan College, an MSE degree from Union College and is a registered professional engineer in Delaware. Lemmerman was with GE's Switchgear and Relay operations for 17 years before joining PECO-Energy in the late 1980s.
david.lemmerman@peco-energy.com