Railroad Signalling: Overview of Operating Appliances

Railroad Rules, Signalling, Operations:

Overview of Operating Appliances

Mark D. Bej

[Copyright & Licensing Agreement]  [Note to Contributors]  [Back to FAQ]


Track, Track Designations

Track, Track Designations

A train and the track to run it on are surely the most basic items needed for our discussion. Yet it must also be clear which track is to be used for that train. For this purpose, various terminology has arisen.

North America

In North America, most larger railroads used terminology similar to the following: main line, branch [line], secondary [line, track], yard track, etc. These terms, or variants, may actually have a formal definition in the book or rules (BoR) or employee timetable (ETT). If a formalized terminology was in place, a track's could provide some degree of information about it, as names such as the above tended to be somewhat hierarchical: main line tracks were generally massively built, heavily travelled, well maintained, high speed, signalled, and important; each of these characteristics diminished as one progressed down the hierarchy.

There was often other terminology, sometimes formalized (in the BoR or ETT), sometimes entirely colloquial: siding, spur, platform track, running track.

Railroads which primarily had 2 tracks or fewer -- that is to say, most railroads outside of the Boston-Washington-Chicago triangle -- could use simple terminology such as 'eastbound' or 'westbound main [track]' if there were 2 tracks, or 'single main [track]' if there was only 1. Other places the two tracks were considered the 'main [track]' and the 'siding'.

Railroads with large amounts of 3- and 4-track territory had to resort to numbering. The most frequently used track numbering scheme was as follows:

Note that in this sheme, eastbound tracks are numbered with even numbers, from inside to out, just like eastbound trains are typically numbered. Westbound tracks receive odd numbers. Tracks can be added to the outside of the already built set (in boom times) without causing a renumbering, and similarly subtracted in lean decades. The New York Central and New Haven (at least) used this scheme.

The one definite exception was the Pennsylvania, which numbered its tracks in consecutive order:

Finally, I've been assuming right-handed operation. While this is generally true in North America, the rule is not universal, as the Chicago & North Western operates left-handed.


The British system has been to name tracks after their function and direction relative to the datum point, London. Tracks leading in some sense toward London are 'up'; those away from London are 'down'. (There are some exceptions.) The terminology derives from the Control Officers' (despatchers') train graph: 'up' trains appear on an upward line on this graph, whereas 'down' trains move in the opposite direction. The former British Rail also has a lot of multiple-track territory. In 4-track territory, one may find an 'up slow', 'up fast', 'down fast', and 'down slow' tracks. The 'slow' tracks are 'slow', of course, because there are passenger platforms alongside them where trains on the slow track generally stop. Other tracks may be designated by the types of trains that run, e.g., 'up goods' [freight].

Switches [Points]

Switches are controlled and connected with the signalling system in various ways and to various degrees. Much of this is detail beyond the scope of this introduction.

I do want to comment on switches in the context of model railroads, however. It should be noted well by model railroaders wishing to signal their layouts that railroads do not scatter their physical plant anywhere they may please. Rather, major junctions are exactly that: junctions, where the switch and signal apparatus has been brought together into a fairly compact entity. Between these compact entities are comparatively long stretches of "nothing"; in the eastern US, 5-10 miles is typical; in the western US or Canada, probably much more.

Unfortunately, few model railroads are planned this way. The compression required of models already reduces the "long stretches of 'nothing'" to effective nonexistence. On top of this, most layout designers do not plan for, or are entirely unfamiliar with, signalled operation. Connections between tracks are placed wherever the designer feels like putting them, rather than in a few compact locations. The result is a layout that is difficult if not impossible to signal realistically.




Bibliography and Related Reading

Mark's Railroad-Related Stuff
Railroad Books of Rules and Signalling: A Home Page
Introduction to Railroad Signalling

Mark D. Bej
+1 216-444-0119


> Where do you put the lights in relationship to the switches? You're now speaking of an "interlocking" or a "controlled point" or whatever the favorite terminology of your favorite railroad is. Outside. I'll leave it at that for now. Let me just make 2 quick comments, regarding model RRs in particular: Model RR compression being what it is, the 'long boring' stretches of nothing tend to be compressed much more than terminals. Well, these long boring stretches of nothing are where the automatic signals usually live. So model RRs have an inordinately large number of controlled (interlocking) signals. Second, most model RR plans out there plan for no signals. Their control method consists of the dispatcher yelling "Hey, Bill, stop your train before you get to that big bend over there" -- or worse, the train jerking to a stop as the head locomotive hits an unpowered block. Since models don't have to worry about physical plant expense, switches are literally _everywhere_. What you find on the _prototype_, in contrast, is that switches (controlled points) are _concentrated_ together. A lot of control circuitry has to be built, and it would be very expensive to stretch this stuff over _miles_. So in general, most model RRs are very difficult to signal, since nearly all signals end up being controlled signals, and in many instances, the model is simply not designed to accept a sensical placement of signals. > What kind of signals were used in classification yards? Very little. Most RRs had some provision for a hump signal (for hump yards) that often had 3 or 4 aspects: stop, back up, hump fast, hump slow. (Probably most RRs had only 3). Many RRs intentionally used a different form of signal for their hump signal than they did for their mainline signals -- obviously to avoid confusion. E.g., the PRR used position-light signals on the mainline (rows of 3 yellow lights), but in Enola Yard the hump signal is a 3-lamp color-light signal (red/yellow/green, I assume, has anyone ever seen it directly?). The Western Maryland used color-light signals on its mainline, but it used a PRR-style position-light signal at its yard in Hagerstown. But. The B&O, which used color-position light signals on its mainline used the same type of signal for its hump signal in Cumberland, Md. (though mounted quite high). > I've a small 4-lead classification yard and was thinking either dwarf > signals or switch stands (no lights but metal flags or whatever). Very few 4-track yards on the prototype would warrant dwarf signals. Only if this was a yard for high-speed freight would it probably be signalled. E.g., a 4-track coal yard for the Pennsylvania Power & Light plant would certainly not be signalled. However, a 4-track yard where hotshot trains set off one block and pick up another, refuel, recrew, and leave within half an hour or so to continue their cross-country trek would be reasonable candidates for signals [though not under Conrail, as the recent effort at Collinwood Yard demonstrates - Ed.]. Enough for now? -- Mark D. Bej bejm@eeg.ccf.org ------------------------------------------------------------- For assistance with the list "PRR-Talk", send the message "help" to "listserv@dsop.com". If problems persist, contact "listmaster@dsop.com".