Search This Blog

Tuesday, January 31, 2017

To Float or Not to Float

I wanted to discuss a topic I have mentioned multiple times in other posts, but never really addressed head on.  This is the distinction between the two flavors of traffic control operation in North America typically known as Centralized Traffic Control (CTC) or Traffic Control System (TCS).  In one flavor, each segment of track has a permanent traffic state that is in one direction or the other with signals against the flow of traffic all dropped to their most restrictive indication (usually Stop and Proceed / Restricted Proceed). 

Traffic set opposed on tks A and 3
 In the other flavor, if no route is set from an adjacent interlocking, the traffic direction "floats" to an undefined state where block signals in both direction display a signal based solely on block occupancy.  This leads to signals displaying favorable indications in both direction at the same time, similar to non-traffic controlled bi-directional ABS lines. 

No following train causes opposing signal to float to Clear 

The first method is most commonly associated with the Union Switch and Signal company, which implemented "traffic levers" on its Left-Right interlocking frame and unit lever interfaces.  The second is most often associated with General Railway Signal, who tended to use its N-X style interface where displaying a signal would automatically set traffic between there and the next interlocking.  However, in modern times these corporate distinctions are not always clear and I have seen exceptions to the "floating" style within a single track segment! 

Hudson Line Intermediate 79 like to float.

While adjacent signal 81 does not!
The "traffic lever" system is perhaps the easiest logically to implement.  Some sort of constant state is transmitted between interlockings and each intermediate (perhaps a + or - voltage on a wire) and a relay at each signal location senses the direction of traffic.  The floating system requires a reliable way to hold the signal state until the train passed and the route is released and then allow opposing signals to clear.  The payoff for this complexity is that while under a traffic lever system the track segment must be completely clear before the direction can be reversed, the float system can handle reverse movements line a train in a non-interlocked siding as soon as the first train passes.  This also reduces problems from track circuit failures that can can cause a "traffic lever" to become locked until the failure is repaired.

In the above video you can see the floating system in action where the far signal immediately upgrades from Stop and Proceed to Advance Approach.  In the absence of any signal at the next interlocking, both signals would flash Advance Approach. Determining which system is in use on a particular line can be tricky.  If you see any of the above behavior with reverse signals automatically clearing or opposing signals in CTC territory both showing favorable indications, then the float system is in effect.  However the absence of said behavior does not indicate the use of traffic lever as the preceding controlled signal could be fleeted for following movements.   However, given sufficient observations, if one never sees signals automatically upgrading behind a train, it is likely that traffic lever logic applies. 

If your opposing signal never clears, you might be in traffic lever country.
I will update this space if/when I learn how these systems are technically implemented and if I can confirm if traffic lever is still associated with US&S and float with GRS.  If anyone has this information, please leave a comment.  


  1. Out here on the BNSF Chicago sub (Racetrack) the signals are continuously lit. The intermediates “float” when no route is lined up. They don’t however clear up directly behind a train for the reverse direction the way block signals do. Instead, they clear up once the train completely passes the next signal.

    I’m guessing it may be a tweak to avoid implying authority for a reverse move. Unfortunately, I do not have any information on the signaling technology installed.


    1. Might have something to do with the cab signaling and block state being transmitted via coded track circuits. I assume that setup probably saved a few relays or logic checks.