The Broad Way: Documentation: PRR Catenary

A Pennsylvania Railroad Home Page


[ Note to Contributors | Copyright & Licensing Agreement ]


Chris Coleman
for saving this post of mine to the RAILROAD listserv. It was originally written 1994.10.28. Since setting up my own web site, I have "repatriated" it [a political joke for my Canadian friends].
Gordon Davids
for several corrections


Please see also the excellent 2-article series recently (1997) appearing in the Keystone.

Wire structure

Two basic varieties of catenary wire were used. The first employed 2 wires, the messenger wire and the trolley wire. The messenger wire is the one that is in the shape of a catenary (the name of a mathematical curve based on an exponential - essentially the curve assumed by any object of constant density when suspended from 2 points). The trolley wire is then suspended from the messenger by means of hangers and clips. This type of wire was used in areas of lower and slower traffic; many Pennsy freight lines used this variety.

The second variety adds a third wire, the auxiliary wire, to the first two. The auxiliary wire rides just above the trolley wire. It contains more copper than the others, making it softer, but also allowing it to conduct better. It also provided a little more "give" to the catenary structure as a whole, making it ideal for lines with higher speeds.

Catenary wire is under extreme tension; an Amtrak "ET" man once told me that it is quite possible to walk on the catenary wire). [ET = electric traction: a distinct department from PRR's C&S = communications & signals.]

Section Breaks

Section breaks are gaps in the overhead catenary wires. In this way, each section of catenary can be turned on or off individually. As you proceed along the track, a section break looks like the following:
  1. the catenary shoe is running on wire #1 (the 'previous' wire), on which it's been running for, say, the last 10 miles (16 km).
  2. On catenary pole A [real cat poles would be numbered with milepost values, to hundredths of a mile], wire #2 starts. It's attached to the vertical pole, not the cross-brace, for strength.
  3. Part way between cat poles A and B, wire #2 passes through insulators and continues dropping.
  4. At cat pole B, wire #2 looks like real catenary (before, it was just a group of wires). Wire #2 is now located alongside of wire #1, maybe a foot (1/3 m) away, and the trolley wire is higher than that of wire #1. Thus, the pantograph shoe is still in contact with wire #1.
  5. From cat pole B to cat pole C, wire #2 drops while wire #1 rises. I.e., about halfway out, the "pan" shoe is contacting both wires for a brief second.
  6. From cat pole C to cat pole D, and for the next 10 miles, the pantograph will slide along wire #2. In the meantime, wire #1 rises up and away to a set of insulators.
  7. At cat pole D, wire #1 attaches to the catenary pole.

Here is my best effort at 3-D image creation, depicting a standard section break. The catenary poles, as depicted, are designed for 2 or 3 tracks. I used a 2-wire (messenger and trolley) design so as not to clutter up the diagram with a 3-wire design (messenger, auxiliary, and trolley). Also for clarity, I am depicting the wire for only one of the tracks.

Thus, the wires have no physical and no electrical connection over the track. There is an electrical connection between the 2 wires if they are both connected into the power supply. Thus, one sees no spark as the pantograph passes from one wire to the next.

Each section on the PRR was (max.) about 10 mi. long. About every 20 miles, i.e., at every other section break, there a substation stepped the supply power down to the catenary's 11,000 V and fed the sections in each direction. Section breaks have no relationship to signal locations (track blocks), *except* that interlockings often had their own sections.

Other Section Breaks

There is a "simpler" way 2 catenary sections were separated, used infrequently and in slow-speed territory, because it led to a generally weaker structure. This could be found where space was a premium (e.g., yards) and one did not have 4 catenary poles over which to spread the break.

Section Breaks at Interlockings

Interlockings are particularly complicated and difficult wire. Each crossover was strung as 2 halves, each half attached electrically to its own "straight" section, to allow for electrical separation should the need arise. Where these joined "straight" sections, a physical wire-to-wire connection was made. Interlocking towers contained "plate order" books showing all of the electrical catenary sections and which switch (on a special panel in the tower -- and in the Power Director's office) turned off which section of track.

Moreover, the wires for crossovers had to be attached to catenary poles for tensioning. Consider that, for each crossover, one had to find a pole, lower a wire, attach it to the main track's wire, run it for half the length of the crossover, then raise it out of the way, insulate it, and run it to another pole. Running one of these wires to another pole often meant crossing it through several other sets of wire for adjacent tracks. Because of the proximity to these other wires, an electrical connection was made, and the wire which is "just passing through" was actually energized, though it served no active purpose other than to provide tension. It is no wonder that, once installed, the interlockings were not changed much afterward.

Further, there were section breaks at/near the home signals, so that tracks within the interlocking could be isolated from tracks outside the interlocking. Further still, if several logical routes were possible within the interlocking (i.e., if the interlocking were more than a simple "complete" set of crossovers), each of these was circuited separately. E.g., for a 2-track simple interlocking, section breaks would occur as follows (denoted *): [the long distance from the westerly home signals to the crossovers is only for parallelism with the 2nd diagram below]

               ---                                               ---
                |                                            OO---|
                ||-o                          /     \             |
                |                            *       *            |
                |                           /         \        o-||
                |---OO                                            |
               ---                                               ---

but the following is a bit more complicated because of the parallel diverging routes, which are kept separate for ease of maintenance:

               ---                                               ---
                |                                            OO---|
                ||-o                /         /     \             |
                |                  *         *       *            |
                |                 /         /         \        o-||
                |---OO                   /                        |
                |                       *                        ---
                |                      /

Plate Orders, Interlocking Towers, Model Boards

The model board (track diagram) in the tower was equipped with small red lights to indicate track sections with (temporarily) de-energized catenary. In my examples above, at the locations of each * would be 2 red lights, maybe an inch (2.5 cm) apart. The white line denoting the track would be a bit thicker between the red lights than elsewhere. One red light was for one catenary section, one light for the other, and the thick line represented the section break. E.g., on my diagram immediately above, the uppermost track's 3rd catenary section from the left, the one with 2 crossovers, would have 4 (minimum) red lights denoting its limits: one at each * on the straight track, one each at the * on each crossover. Long sections had additional red lights in between.

In electrified territory, track without catenary was shown on the model board as a dashed line.

These red lights are turned on via a simple switch under operator control on a separate small panel on the desk. They are not interlocked with the interlocking machine (for, obviously, one could still run steam- or diesel-powered trains under de-energized catenary. When the power Director or Dispatcher needed to turn off a section of catenary, a "Plate Order" was issued, and the operator was then authorized to turn the switch for that "plate section", turning on those red lights. Each tower was provided a book entitles "Power Plates" which specified which plates corresponded to which sections of track. These were long and narrow (1 ft. by 2-3 ft.), screw-bound books with hard covers. Each sheet showed the entire interlocking, with the respective plate outlined in a thicker red.

Power Sources, Power Directors

On a system as large as the PRR's, there had to be several power sources. The 4 main ones were New York, Philadelphia, Baltimore/Washington, and Harrisburg. I believe that the stretch from Philadelphia to Paoli, the oldest installation of catenary, had its own power supply. Also, the portion from the North (Hudson) River Tunnels through Sunnyside Yard may have had its own power supply as well. The power for the Newark-to-New York PATH trains is tied into the PRR's power grid, and PRR-style catenary poles still continue into Jersey City.

Each of these regions had a Power Director. The Power Director's office contained a huge model board, 10 or 12 feet (3-4 m) high showing all of the circuits: high-voltage lines, signal supply lines, substations, transformers, switches, circuit breakers, and catenary sections. There was also a rather minimalist track diagram. There are red and green lights scattered about, for closed (connected) and open (disconnected) circuits, respectively. The Power Director could open or close the breakers and throw switches as needed to supply power to a section that lost it or to redistribute the high-voltage power as necessary. These switches are unique as well. They are out in the field, i.e. in the substations. The Power Director, in throwing his switch, controls a low-voltage circuit that fires up a motor in the substation. The motor turns a large copper rod 90 degrees, allowing it to make break one connection and make another. Electrical isolation is provided merely by the air gap.

Power Directors' Offices
City Building Location Status
Baltimore Pennsylvania Station 2nd floor, east side, next to fmr. dispatchers' office closed
Harrisburg PRR Station 2nd floor, south side of bldg., above STATE tower operational
New York 32nd St. in separate building ?south of Penn Station reported to be closed (building condemned)
Philadelphia Pennsylvania (30th St.) Station 4th (?) floor, southeast side of building closed

Power Directors' offices have been closed as their duties have been progresively transferred to the console in either CETC in Philadelphia, on the 8th floor of 30th St. Station, or PSCC in New York. These stations contain a computer with a large screen, on which several "zoom" settings can be used to view a "model board" similar to that which was in the former offices. The disadvantage is that it is impossible to view the entire board as a whole.

Phase Breaks

The respective territories of the power plants, and Power Directors, had to meet, and these junctions are wired with special sections of catenary called "phase breaks". Phase breaks exist in the following locations:

Pwr. Dir. 1 Pwr. Dir. 2 Location
Baltimore Philadelphia East ("north") of Perryville, Md., on the Corridor (PW&B)
Baltimore Harrisburg Northwest ("west") of Perryville on the C&PD Branch
Harrisburg Philadelphia East of Thorndale, Pa. (P&C)
Philadelphia New York ??? at/near MORRIS interlocking;
also, at the western tunnel portal of the North River Tunnels

The PRR identified phase break locations with "PB" signs strung in the catenary, one or 2 catenary poles before the phase break proper. These signs are a square(ish) piece of sheet metal, painted black, with the letters "P" and "B" created with holes drilled out of the metal. If you understand a section break, you'll understand a phase break -- phase breaks are quite simply 2 section breaks in sequence. Here's how it works:

  1. The catenary shoe is on wire #1.
  2. At catenary pole A, a "PB" sign hangs in the catenary and the phase break signal is attached to the upright (see also map of THORN interlocking on this site).
  3. Wire #2 starts on the upright of catenary pole B.
  4. Partway out, wire #2 passes through insulators and continues dropping.
  5. At catenary pole C, wire #2 looks like real catenary and is alongside wire #1. Wire #2's trolley wire is still higher than #1's.
  6. From cat pole C to cat pole D, wire #2 drops while wire #1 rises.
  7. At cat pole D, wire #2 is lower than wire #1; wire #1 starts rising away.
  8. From cat pole D to cat pole D, the pantograph remains on wire #2. Wire #1 rises to a set of insulators.
  9. At cat pole E, wire #1 attaches to the upright. On the other side of cat pole E, wire #3 is attached to the upright.
  10. Between cat poles E and F, wire #3 attaches to insulators and continues to drop.
  11. At cat pole F, wire #3 is just a bit higher than wire #2.
  12. Between cat pole F and G, wire #3 drops a bit and wire #2 rises a bit so that the pantograph now rides on wire #3.
  13. At cat pole G, wire #2 is higher than wire #3.
  14. Between cat poles G and H, wire #2 rises, to insulators.
  15. At cat pole H, wire #2 attaches to the upright.
  16. At cat pole I, if the track is signalled for bidirectional operation, one may find the opposing PB sign and phase break signal.

Wire #1 as described above is attached to one power supply, wire #3 to the other, both reasonably permanently. The center section of wire, the phase break itself, thus extends for about three catenary poles, not far. Under normal operation, with two power companies' supplies being of the same voltage and synchronized in frequency, the center section remains connected electrically to one supply or the other, and the engineer notices nothing. However, if the power supplies should get out of sync, circuitry will detect this and cut out the center section. It is my understanding that equipment then tries to get the 2 power supplies back in sync with each other; I don't know how this happens.

The reason why phase breaks are needed at all is as follows. If the power supplies were to be hooked together while out of phase, one would start to get large shifts of electricity from one power company to the other: first one would be the source of power and the other the load (sink), then they would reverse roles, 50 times per second. These huge loads tend to cause the power supplies to get even further out of sync, worsening the original problem.


If the phase break circuitry is tripped and the phase break is cut out, the special phase break signals will light to inform the engineer. Engineers must then turn off the controller (throttle) and coast through the section. Dropping pantographs is not needed. The phase break signals are a special version of standard PRR signals. They contain 8 lights, along the outer ring, no center lamp. The ones I've seen have no signal face. They normally remain dark -- though the newest phase break employing these signals, on the SEPTA ex-Reading Ninth Street Branch, near Brown St., has 4 PRR-style phase break signals in each direction that are lit continuously because this is a phase break between former-PRR territory and former-RDG territory.

There is also a "drop pantographs" hand signal. On PRR, this is done by lantern swinging the lantern in a circle, at arm's length, perpendicular to the track. (The hand signal for "reverse" is a circle at half arm's length.)

Dead Section

There were permanent dead sections of catenary, i.e., sections that were never hooked to electrical power. These were identified by a sign hung in the catenary, similar to the PB sign, though with the letters "DS". I never really identified where these were used or why.

Track in Electrified Territory

The return path for the electrical current used in locomotives and self-propelled cars is the "wire" most easily available: the rails themselves. The running rails (NOT third rail, where present) are at ground potential, so they are safe (electrically speaking) to walk on and touch.

The huge catenary current must be returned without interfering with the cab-signal coding signal, which does double duty in detecting block occupancy and "telling" the previous signal what to display. The cab signal coding signal/occupancy signal is itself an alternating current signal, but of much lower voltage and of different frequency from the catenary AC signal. To allow this, on PRR lines, you will find "impedance bonds" at locations with insulated rail joints. These are low, flat boxes about 1.5 ft (500 cm) square, set in the middle of the track, with thick cables going to the 2 rail sections ahead and 2 behind. These do have a significant shock hazard (on their interior, that is). The impedance bonds allow the 25 Hz catenary power to pass but prevent the 2 Hz and below cab signal code from passing from one block to the next.

Track circuits in electrified territory are quite different, employing, at the least, a centrifugal relay. This relay has a contact consisting of a metal ball on a flat metal rod (probably spring steel). When de-energized, the ball is in contact with a smaller-diameter circular contact. When the relay is energized, a motor starts to rotate the contact. The centrifugal force eventually forces the ball outward, breaking the original contact. As the motor continues, the ball ultimately contacts an outer, larger-diameter circular contact. Thus, this 'relay' provides 2 contacts, is a break-before-make, and provides an inherent delay.

Pantogaphs; Double Pantograph Orders

Electric locomotives and self-propelled cars (Silverliners, Metroliners, etc.) are provided with 2 pantographs apiece. Both pantographs are electrically connected via a roof wire (on insulators). Moreover, Silverliners (?and Metroliners) come as mated pairs, i.e., semi-permanently coupled. An electrical connector between the pairs connects the 2 pantographs of one car with those of the second. Normal operation for electric locomotives is with only 1 pantograph up; for self-propelled cars, 1 pantograph of every 2 cars. This is to reduce wear on the wire as well as the pantograph shoe.

The area of the country where the PRR electrified is famous for its winter sleet and freezing rain, however. (Snow is less of a problem.) For better contact, reduced arcing, and better clearing of sleet from the wire, a "double pantograph order" would be issued in such weather conditions. Often, an "ET" man would ride the train and observe the wire for problem areas. The sleet (and wet snow, too) is heavy, and not uncommonly would cause tree branches to break off and become lodged in the catenary.

The End of the Line

What happened in Greenwich, Harrisburg, Enola, and Potomac Yard? Generally, the entire receiving yard was under wire. An escape track or two was under wire so that the "motor" (as the Pennsy electric locomotives were sometimes called) could be taken into the engine terminal. A few tracks of the engine terminal were under wire as well. Finally, departure tracks were electrified, too, though not in their entirety. It was up to the hump tower operator to make sure the first car was under wire.

Locations where the wire ended were marked with, what else, a sign strung in the catenary. These signs read "AC MOTOR STOP" and were made with holes in sheet metal, just like the PB and DS signs. [The yellow speed restriction signs strung in the catenary, now visible on the Northeast Corridor, are an Amtrak creation.]

Electrical Safety

This is such an obvious concern that a special book, the Electrical Operating Instructions, C.T. 290, was issued. This book goes into excruciating detail about how things are to be performed and which employeers are qualified to do what. Needless to say, novices must stay away and avoid urinating on the wire. [There is an apocryphal story of an electrocution as a result of same...]

A catenary wire remains extremely dangerous even when turned off, because of a phenomenon known as inductive coupling. Essentially, the wire builds up a static charge, sufficient to kill, from neighboring wires. A wire must be well grounded before anyone approaches it. ET employees have special grouding poles for this purpose: the pole is first attached to the rail; then it is allowed to contact the wire. A minimum of 2 grounding poles are placed, one on each side of the work site.

Road bridges over electrified track are easily identified by the high walls preventing any view of the railroad. There are very few bridges where such barriers are not in place.

Mark's Railroad-Related Stuff
A Pennsylvania Railroad Home Page
PRR Catenary

Mark D. Bej
+1 216-444-0119