Railway Signalling and Operations FAQ: Interlocking Signal Types and Placement

Railway Signalling and Operations

Interlocking Signal Types and Placement

[ Note to Contributors | Copyright & Licensing Agreement ]


This page was inspired by question/thread started by Eddie Oliver.

Interlocking signals come in 2 basic varieties, high and dwarf (or low). Every route into an interlocking must have a governing signal. Very general rules for use of these signals are as follows:

High signals are used for:
  1. Main tracks signalled for travel toward the interlocking in question (or bidirectionally signalled).
  2. Signalled sidings signalled for travel toward the interlocking.
Dwarf signals are used for:
  1. Main tracks or signalled sidings signalled for travel away from the interlocking only.
  2. Unsignalled sidings.
  3. Yard tracks, industrial spurs, etc.
  4. On some railroads, at some locations, for any low-speed route into the interlocking, even for main tracks.

Railroads tend to install high signals at the entrance to an interlocking to govern those tracks that allow a high speed route into the interlocking. Obvious candidates are main tracks and signalled sidings. However, not all main tracks qualify - see below.

Yard tracks, unsignalled sidings, and industrial spurs entering an interlocking did not allow a speed to warrant a high signal. Particularly, since such tracks are "dark" (unsignalled), the engineer must, explicitly or implicitly, approach the interlocking signal prepared to stop. Main tracks whose current of traffic (which are signalled) for travel in the direction opposite to the direction a train is actually travelling are effectively, for that train, "dark", and therefore, similar rules apply. In most cases, such unidirectionally signalled tracks warranted only a dwarf signal.

In those cases where no routes into an interlocking were 'high speed routes', practice varied from railroad to railroad. Some railroads installed high signals, presumably to reinforce the fact that a "main track" (regardless the speed) is being signalled. Other railroads installed dwarf signals at such locations, since the low speed did not require longer sighting distances.

Here is one example:

                          +                 OO---|       OO---|          O---|
    West <--+--> East     +          ------------ ------------ -------------- -----------
                          +         /     /   ++++
    OO---|                +        /     /  o-|
--------- ---------------- ------------------- --------------- -------------- -----------
         |---OO           |---OO     /        +               |---O          |---O
                          ++++++    /         +
------------------------\      +   /          +
--------------- YARD ---------- --/           +         |---O high signal
-------------------------/     |-o            +         |-o   dwarf signal
                               ++++++++++++++++         +++   interlocking limits

Note that the upper main track is signalled for westward travel only, the lower main track for eastward travel only. The single track is bidirectionally signalled, of course. A westbound train on the westwardly signalled track would receive the normal high signal aspect. If the upper track were closed or blocked and a westbound train were travelling on the lower, eastwardly signalled track, it is travelling against the current of traffic and must be appropriately protected (e.g., by establishing a Manual Block). That train would not have seen any automatic signals in its direction, governing its track, since the last interlocking. There is not even a "distant" (or equivalent) signal.

Such a train must logically approach the dwarf signal (the interlocking home signal governing its track) prepared to stop. Since the train must be prepared to stop, i.e. at very low speed when nearing the interlocking, no more than a dwarf is needed.

Range of governance of high and dwarf signals

In N.Am. practice, high signals and dwarf signals each governed specific stretches of track, defined similarly: from the signal in question, to the next signal in advance. There is no concept in N.Am. signalling of one signal "superseding" another within that previous signal's block. Once a fixed (wayside) signal aspect is seen (and therefore, its indication received), that signal governs until the train reaches the next signal in advance over that train's specified route, regardless of signal type.

As usual, no rule is absolute. Two modifications to the above that I can think of are:

  1. Cab signalling. If the cab signal aspect changes between two fixed signals, the new cab signal aspect governs from that point on. In some cases, the train may have to run its length.
  2. Signals given with flags by employees along the tracks, or the ever-popular emergency "any object waved violently by any person along the tracks is a signal to stop".

The above is all in contrast to shunting signals, which, for as best as I understand them, govern together with the high signal, yet are considered a separate signal. In N.Am., shunting would be handled by providing a dwarf signal (usually), possibly in a "stick" mode to allow multiple shunting moves without the signal "dropping". If a high signal happened to exist in that location, the high signal would provide the "shunting" aspect. Example:

----- -------------------------------------------------------------- ---------
   6R|-o   1 \                                               / 7
              \ 1                                         7 /     o-|4L
----- ---------------- ---------------------------- ---------------- ---------
   2R|---OO    3 \                                      / 5
                  \ o-|2L                              /
                   --- ---------------------------- ---
                         --------------------    4R|-o
                         |     PLATFORM     |

Let's assume the mainline tracks are unidirectionally signalled. A train bypassing the passenger station eastbound, all switches and crossovers normal, would receive the 2R signal and would be governed by no other signal within interlocking limits. Thus, that train's block extends from 2R to the first automatic east of the interlocking (off the map).

A train coming into the passenger platform would receive something other than a full clear on the 2R signal and diverge on the #3 switch to the platform. That train's block extends from the 2R only to the 4R signal. When the station work is done, the train will be governed only by the 4R signal. The previous or current aspect on the 2R is irrelevant. In fact, in the simple arrangement above, with the #5 switch reversed, 2R would have to remain at stop. (A call-on would be the best obtainable aspect.) When the train accepts the 4R, its block then extends from that 4R to the first automatic signal (or interlocking signal, but I'm considering the most generic, "cookie-cutter" case) outside the interlocking (off the map).

Distinguishing dwarf from high signals

When encountering a set of dwarf signals, that set of signals is highly unlikely to be in anything but slow speed territory. Recognition of these as dwarves should not be a problem, and clearly, the required sighting distance cannot be great given the speed. Further, the signal in the rear will have warned of an upcoming slow-speed aspect and the engineer should be prepared.

In a location where there is a mixture of high signals and dwarf signals, one has visual cues as to which type is which. The distinction between, say, a dwarf red over green, versus a high signal red over green, is quite easily made by the height of the colors from the ground and proximity of one color to the other (high signal lamps are vertically much farther apart). And of course, the engineer should be very familiar with his territory and know what signal type awaits him.

Special Cases

Pennsylvania Railroad

PRR-style dwarf signals are considerably less visible compared to modern searchlight or colorlight dwarves, because of their lamp/lens arrangement. PRR dwarves have 2 closely spaced white lights, instead of 3 widely spaced amber lights. I estimate the sighting distance at 1/2 mile, maximum.

Pedestal signals are found only in "relatively" slow speed territory. Presumably, a greater range of indications was found necessary in some places, and in some places, an indication for 20- or 25-MPH running was deemed useful. PRR dwarf signals cannot provide such aspects, the best aspect being Slow Clear, allowing for a maximum speed of 15 MPH. Over a single switch or crossover, a distinction between 15 MPH and 20 MPH would be considered unnecessarily complicating. In some longer interlockings in high-traffic area, east of Harrisburg in particular, that 5 or 10 MPH speed difference, considered over a mile or two, can significant increase train density. Further, pedestal signals had a narrow profile, allowing them to be used in areas with little side clearance. They were most commonly used in terminal trackage. However, some branches (Buffalo main line and Bald Eagle Branch in Pa.; Marion Branch in Indiana) saw their use in protecting movements out of passing sidings. See this photo for an example.

Note that there is no specific 20- or 25-MPH indication in the PRR rulebook. This would be handled by giving the engineer a Medium Clear or Medium Approach, normally a 30 MPH indication, but then in the Speed Limits section of the Employee Timetable/Special Instructions, stating that the speed limit at that location is 20 MPH. For example, the Special Instructions may state, in the table of Speed Restrictions,

Between/At Track 1 Track 2 Track 3 Track 4 Other Track
North Philadelphia &
East limits Zoo
70 70 70 70
Zoo Interlocking All Tracks, 25 MPH Track #0
20 MPH

New York, New Haven & Hartford Railroad

The New Haven was unique in the setup of some of its interlocking signals. Two high-style signal heads would be set up on a signal bridge (or mast?) -- so far no different than any other railroad. However, the NH would then put a dwarf signal immediately below (or, if on a mast, directly alongside of) the high signal. The high signal and dwarf signal together were one signal, the dwarf signal acting as the 3rd (lowest) arm of a 3-arm high signal. Thus, where most railroads would have Red/Red/Green on a high signal as the aspect for Slow Clear, the NH would display it as Red/Red on the high signal, Green on the dwarf. I do not know if this high-dwarf arrangement was standard, preferred, common, or uncommon. In later years, and even into today (1997), Amtrak Special Instructions contain a paragraph describing how these signals are to be interpreted:

At an interlocking where a Low home signal and a two unit High home signal in combination govern the movement, the Low home signal must be considered as the bottom unit of a three unit color light aspect.

Mark's Railroad-Related Stuff
A Pennsylvania Railroad Home Page
Railroad Books of Rules and Signalling: A Home Page
Railroad Signalling FAQ
Interlocking Signal Types and Placement

Mark D. Bej, M.D.