Amtrak s 138 kv Transmission System that supplies traction power to electric locomotives traveling on the

Transcription

1 Word Count: 3039 Replacement of Amtrak s 138 KV Cable through the Baltimore Tunnels Raymond G. Verrelle, PE Amtrak Engineering Department 30 th St. Station, 4 th Floor South Fourth Floor, South Tower Philadelphia, PA Phone: Fax: Robert J. Verhelle Amtrak Engineering Department 30 th Street Station Fourth Floor, South Tower Philadelphia, PA Phone: Fax: ABSTRACT Amtrak s 138 kv Transmission System that supplies traction power to electric locomotives traveling on the Northeast Corridor runs from Landover, Maryland to the Hudson River Tunnels in New Jersey. The 138 kv circuits are bare, aerial conductors with the exception of the four high pressure oilostatic cable circuits that ran through the Union and B&P tunnels in the Baltimore, Maryland area. The oilostatic system used oil to insulate as well as cool the cables in the conduit system. The cables were installed in steel pipes in an underground conduit system and encased in concrete envelopes along the tunnel walls of both tunnels. They needed to be demolished and re-built in place due to the limited clearances in the tunnels. The cables had been in continuous operation since 1935, had outlived their design life and had become increasingly difficult and costly to maintain. The proposed replacement was 161 kv class solid dielectric cables. In 2003 the prototype design of a replacement for one of the oilostatic circuits through the Union Tunnel was started with the intent to replace all four circuits. The construction was started in 2004 using contractor forces and a design was awarded for remainder of the project. The construction was completed utilizing 55 hour weekend outages with train service on at least one track during the outage. Both tracks were returned to service for revenue operation at the end of each outage. The construction was completed in BACKGROUND Overview of Amtrak s Power System

2 As a part of its Electrified Northeast Corridor, Amtrak operates a single phase, 25 Hz electrification system that supplies power to electric locomotives operating between New York City and Washington, DC. This electrification system consists of a 138kV transmission network that powers the trains via step-down substations that are typically about 10 miles apart. The electrification system was originally installed by the Pennsylvania Railroad between 1915 and Much of this infrastructure has outlived its useful life and is being replaced by Amtrak as part of a state of good repair program. While most of the infrastructure replacement is generally straight forward, the replacement of the 138 kv oilostatic cables in the B&P and Union Tunnels presented a unique challenge. History of the Baltimore Tunnels and the Oilostatic Cable System The tunnels that surround Amtrak s Penn Station Baltimore are comprised of the Union Tunnel and the Baltimore and Potomac Tunnels (See figure 1). Union Tunnel The Union Tunnel was constructed as part of the Union Railroad in the late 1870 s before it became part of the Pennsylvania Railroad, Penn Central and eventually Amtrak. The original two track tunnel is approximately 3,400 feet long. In the early 1930 s a new two track tunnel was constructed next to the existing Union Tunnel to create a four track mainline to the North of Baltimore. The new tunnel was built to accommodate the proposed electrification. During the electrification, the Old Union Tunnel had to be changed to a single track operation due to the additional clearance needed for the electrification. Baltimore and Potomac Tunnels The B&P Tunnels were constructed between 1871 and The B&P Tunnel consists of three, two-track tunnels the John Street Tunnel (1,150 ft. long), the Wilson St. Tunnel (3,650 ft. long and the Gilmore Street Tunnel (2,200 ft. long). The tunnel construction was essentially cut and cover construction with masonry side walls with a four to six foot thick brick arch. Some renovations took place in 1935 when the catenary and oilostatic cables were installed and some additional renovations (Slab and drainage work) took place in the early 1980 s as part of the Northeast Corridor Improvement Project (NECIP). Oilostatic Cable System When the city of Baltimore voiced protest over the PRR installing 138 kv overhead transmission lines, they needed to figure out a way to get 138 kv circuits into a tunnel with limited space. The PRR looked to the Okonite Calendar Company for the solution. A relatively new technology patented by W.C. Bennett employed the use of high

3 pressure oil as both the insulating and cooling medium for a high voltage pipe cable. This installation was one of the first of its kind in the country. The pipe cable system began at Baltimore Substation located near Pennsylvania Station. The #120 and #220 transmission circuits leave Baltimore Substation and run North through the New Unions Tunnel where they transition to aerial conductors at the Bond Street Portal. The #121 and #221 transmission circuits leave Baltimore Substation and head South through the B&P Tunnels where they transition to aerial conductors at Appleton St. (See Figure 2) There was an oil pumping station located in Baltimore Substation that maintained the oil pressure in all four circuits at 200 psi. Each circuit had its own oil shut off valve so an individual transmission circuit could be taken out of service without affecting the other three circuits. The oilostatic system originally housed both phase conductors of one transmission circuit in a seven inch steel pipe that was filled with the pressurized oil allowing for a relatively small construction envelope. The conductors were 500 MCM copper insulated with oil impregnated wood pulp paper wrapped in one layer of cotton tape. The seven inch oilostatic pipe was wrapped in a hot tar and enamel coating for corrosion protection. Splices and terminations were specifically designed for this system. In the Union Tunnel, the duct bank for the oilostatic cable was built on the elevated bench wall in a trough filled with sand and covered with a removable concrete slab. In the B&P Tunnels, the oilostatic pipes were installed one on each side of the tunnel floor and encased in concrete. CABLE AND DUCT BANK DESIGN Cable Selection The first task was to determine a replacement for the existing oilostatic cable. The options that Amtrak considered feasible were high pressure fluid filled (HPFF) cable or solid dielectric cable. While the existing oilostatic cable provided over 70 years of very reliable service, the system demanded a considerable amount of maintenance as well as posed an environmental issue if a leak was to develop in the oil pipeline. The solid dielectric technology has been in the industry for over 20 years, but Amtrak s only experience with the solid dielectric cable was in a power station that was installed in The main driving factors for the choice of solid dielectric cable were: 1. Low maintenance and operating costs. 2. No environmental issues. 3. Simple installation. 4. Lower, overall costs (from previous studies performed by Amtrak).

4 The decision was made to go with solid di-electric and the cable selected was a 161 kv class cable with a 750 MCM copper conductor (See Figure 3). There would be one phase conductor per conduit. One limitation on this project was the reel size due to clearance and handling issues. This limited the cable reels to a maximum of 2,200 feet. Splices and terminations were specified per the cable manufacturer s recommendations. Conduit System Design Duct Bank The proposed duct bank would consist of concrete encased schedule 40 PVC. The duct bank design started with the #1 transmission line heading north to the Union Tunnel (designated as the #120 transmission). The routing was essentially independent of the existing oilostatic duct bank so both transmission lines could remain in service during the bulk of the construction. The new duct bank was designed to contain both the #120 and #220 transmission circuits to a point just before the south portal of the Union Tunnel. The duct bank then split off into individual circuits at the south portal of the tunnel so each circuit could be installed in the bench wall after removal of the existing oilostatic line. The existing sand filled trough would then be filled with concrete to encase the new cable ducts. The next design task was the #1 transmission line heading south to B&P Tunnels (#121 circuit). The unique design issue with this circuit was the location of the duct bank in the existing bench and the limited space in which to work. This design would also be applied to the 221 circuit which had an identical routing on the opposite side of the tunnel. Another issue was the overall length of the tunnel and the need to make a splice inside the tunnel. This issue was solved by designing a long coffin like splice chamber that fit into the operating envelope of the tunnel and still provided enough space to make one splice in each cable. The chamber wound up being 40 feet long by 18 inches wide by 18 inches high (compared to the 16 foot by 6 foot by 6 foot manholes). There are diamond plate doors on the top of the chamber for its entire length to provide access to the cables. The doors are locked down and danger - high voltage signs are placed on the doors to provide appropriate warning for any non-electrical personnel working in the tunnels. This duct bank consisted of a 2 foot by 2 foot buried concrete duct bank from Baltimore Substation #20 to the John St. portal (the north portal of the B&P tunnels). One unique feature of this portion of the routing was a directional bore under six active tracks in which a 20 inch casing pipe was bored under the track to house the #221 circuit. The

5 duct bank in the tunnels had to fit within the existing bench wall envelope due to the limited clearances in the tunnel. There also had to be a relocation of the existing conduits attached to the bench wall that carried C&S cables. Manholes and Splice Chambers Manholes for this conduit system were sized based on the overall length of the cable splice. While the splice itself was only about four and a half feet long, there needed to be at least 8 feet of exposed cable per side of the splice to physically install the splice. That resulted in a manhole with an inside dimension of 16 feet long by 6 feet wide by 6 feet high to contain a single circuit splice (both phases). The manhole was widened to 8 feet where it contained both circuits (all four conductors). The manholes were designed to E-80 loading due to the proximity of some of the manholes to the track influence line. The manholes were cast in a top and bottom half and connected via a standard ship lap joint. The manholes were equipped with all the necessary cable pulling and racking accessories and two entrances. Modifications at Transitions from Cable to Bare Conductors There were some structure modifications required on the risers at Baltimore Substation #20, Bond St., and Appleton St. to accommodate the new cable, cable terminations and cable protection. In the high vandalism areas, protective shielding was added to the cable termination platforms. DUCT BANK AND CABLE CONSTRUCTION Ductbank Construction The initial ductbank construction for the #120 circuit was taken on by Amtrak forces in This portion was essentially the duckbank work for both the #120 and #220 circuits from Baltimore Substation #20 through Pennsylvania Station to the south portal of the New Union Tunnel. Five manholes were also installed. This construction was relatively straightforward less the usual underground encounters on a railroad right of way that is over one hundred years old. The remainder of the ductbank work through the tunnel was performed by the contractor that was awarded the cable installation portion of the contract since the cable had to be removed from service to continue with construction. It was intended to have the cable out of service the least amount of time possible during construction. The #121 duct bank was the next phase of the project. The intent of this phasing was to have the #1 transmission line completely replaced in the same location through both Baltimore Tunnels. The duct bank was constructed with both the #121 and #221 circuits in one ductbank until the #221 circuit was required to cross under the tracks at the

6 directional bore. The directional bore of the 20 inch casing pipe containing the two 6 inch PVC pipes was performed by Amtrak forces with no disruption to Railroad operations. The remainder of the ductbank through the B&P Tunnels was performed utilizing 55 hour weekend outages. This outage allowed absolute use of a single track in the tunnels so demolition and construction activities could be performed with a single track operation and maximize the production. The railroad was returned to two track operation for the weekday service. The first step was to drain the transmission circuit in question of all the oil and remove the cable in accordance with all applicable environmental requirements. Next, the ductbank was demolished using rail mounted backhoes, dump trucks, etc. (See Figure 4) With the old ductbank removed, the new schedule 40 PVC conduits were installed and anchored against the tunnel wall and reinforcement was installed. (See Figure 5) The concrete pours were done via a concrete truck with rail gear that was refilled by concrete trucks waiting on the right of way outside the tunnels. Cable Installation Once the duct bank cured, it rodded and mandreled to make sure there were no obstructions or collapsed ducts. After the ducts were confirmed as clear, a pulling rope was installed. All cables were pulled in specified outage periods. Great care was taken to make sure that the cables were properly lubricated and the pulling tensions were not exceeded. (See Figure 6) After the cables were installed, splicing and terminating began. It took two eight hour shifts to complete one splice. The preparation of the splice was critical in getting the splice correct. In particular there is a click fit connector that is used to splice the copper conductor. In this process the conductor is exposed from the insulation on both ends to be spliced. The copper connector is then placed between the exposed copper conductors and the conductors are jacked together until a click is heard on both sides of the fitting. The click signifies that the conductor is properly seated in the fitting and will ensure the full current carrying capacity of the splice. 161 kv class porcelain potheads were installed on the terminal ends of the cables at Baltimore Substation #20 and the risers at Bond and Appleton Streets. Testing and Commissioning Once a complete transmission cable was installed, the cable needed to be tested. Since the factory tests in accordance with AEIC CS7 were performed prior to getting the cable on site, the only tests deemed necessary for commissioning were the Jacket Integrity Test and the 24 Hour Soak Test. The Jacket Integrity test consisted of

7 removing the ground from the metallic sheath and connecting a voltage of 10 kv DC between the metallic sheath and the cable jacket for one minute. The 24 hour soak test is the actual connection of the cable to transmission voltage without load on the system. This is normally done by energizing the line though a transformer and opening all other switches connected to that line. If the cable has no issues for the 24 hours, then it is ready to go into service and the manufacturer s warranty begins. Hot collar tests were performed on each bushing to determine dielectric losses. This test utilizes a high voltage band attached near the top petticoat of the bushing with the center conductor of the bushing grounded. Each cable system was tested and passed the commissioning tests prior to being put in service. SPLICE FAILURE ON THE #121 TRANSMISSION LINE While the entire project went relatively well, there was one issue that was worth some discussion. After the #120 and #121 transmission lines were constructed and put in service, there was a failure in the #121 transmission cable about 6 months after it was placed into service. The fault was located using a cable thumper and the failure was at a splice in one of the long splice chambers located along the tunnel wall. (See Figure 7) The cable manufacturer came quickly to replace a small piece of cable along with two splices in the chamber and returned the cable to service. It was later discovered after a forensic investigation of the splice that one end of the click fitting did not engage and allowed an arcing conductor to eventually break the splice down. That, combined with a cold spell, made the cable contract and led to the final failure. Through the contractor s investigation of their QA/QC records, a change of shift happened when they were installing the click fitting on that splice. Apparently due to a miss-communication, it was assumed that both sides of the splice were clicked when one was not. The contractor later modified the inspection sheets to indicate and confirm each side clicking into place. There have been no other issues and the cable has performed satisfactorily to date. CONSTRUCTION COST AND SCHEDULE The construction was essentially completed in four phases, one for each circuit. Phase one was the #120 circuit at with a duration of about 8 months. Note that this phase included the duct bank for the #220 circuit from Baltimore Substation #20 to the south portal of the Union Tunnel. The second phase was the #121 circuit with a duration of 12 months. The third phase was the #220 circuit which took 4 months. The fourth and final phase was the #221 circuit. This phase took 12 months. The total cost of all four phases was around $22 Million. All phases were completed

8 within budget and on schedule. In the total duration of the project, there were no Amtrak reportable injuries and only one minor contractor reportable injury. Figure 1 Union and B&P Tunnel Sections

9 Figure 2 One line schematic of underground transmission near Baltimore

10 Figure 3 Solid Dielectric Cable section

11 Figure 4 Construction Equipment

12 Figure 5 New Duct Bank in place

13 Figure 6 Cable Pull

14 Figure 7 Faulted Cable Splice