Timed Signals

This past weekend I rode the Museum Christmas train up to 145th St. and back. I was surprised to see that the express tracks from 59th St/Columbus Circle to 125th St. are protected by timed signals, which cleared as we approached at normal speed for an express train.

My question is: why would this stretch be so controlled when it's mostly straight, level, and normally high speed? Timed signals are usually found on steep grades or approaching station complexes, and maybe sharp curves. But what need is there for this kind of signaling on this long express run?

A large number of timers were installed after the Union Square and Williamsburg Bridge crashes in the early 90s, often limiting the speed to lower than the previously approved maximum speed in the interest of "safety." The idea is that overspeed can cause a crash even on straightaways, I guess? Worse yet, these signals have been poorly maintained, and often clear at a significantly slower speed than they are signed. Recently, the "Save Safe Seconds" campaign has been trying to get it through the heads of the ROW department that this is a stupid idea and try to alleviate some of these issues.

Head-end View wrote: ↑Mon Dec 23, 2019 8:10 pm This past weekend I rode the Museum Christmas train up to 145th St. and back. I was surprised to see that the express tracks from 59th St/Columbus Circle to 125th St. are protected by timed signals, which cleared as we approached at normal speed for an express train.

My question is: why would this stretch be so controlled when it's mostly straight, level, and normally high speed? Timed signals are usually found on steep grades or approaching station complexes, and maybe sharp curves. But what need is there for this kind of signaling on this long express run?

Time signals are frequently installed on long, capacity critical express runs or on downgrades to increase throughput. If unrestricted, the maximum speed of trains through such sections of track would be extremely high, forcing signal engineers to design long control lines to mitigate safety risks. Those long control lines, in turn, force trains to slow/interact with ST signals comparatively earlier than if control lines were shorter -- especially around stations. Given this, it's pretty common (especially on >1980s installations) for GTs to be baked into resignallings, which is what happened on CPW when they redid that portion in '89. Southbound, you have (IINM) a combination of GTs installed during the resignalling and ones modded in to address safety issues after the Williamsburg Bridge crash, but nb it's all original.

Okay guys; I get most of what you're saying about what was done in the wake of the Union Square incident. But how does the Williamsburg Bridge figure into it? Didn't that incident happen on the upgrade onto the bridge? I wouldn't think excessive speed would be a concern on the upgrade.

And as I recall that collision was caused by non-comnpliance with a stop signal combined with antiquated short signal spacing from a bygone era and the train's brakes failed to stop the train in the distance specified in the brake design. So again, how would timed signals have made any difference on that upgrade?

1990s signals rebuilds often lengthened signals spacing, exacerbating the problem. If I'm not mistaken, this decision was in response to the Williamsburg Bridge crash. But the combination of longer wait distances between trains + lower speed limits dramatically reduced capacity on some lines.

I think the greatly increased use of timed signals resulted more from the Union Square crash in 1991.

The Williamsburg Bridge crash was in 1995. News accounts from the incident said the signal blocks on the bridge dated back to 1918 and were shorter spaced for the slower, shorter, lighter weight wooden trains of that era. I believe those signals and blocks were due to be replaced and lengthened even before the 1995 crash.

Head-end View wrote: ↑Tue Dec 24, 2019 8:39 pm Okay guys; I get most of what you're saying about what was done in the wake of the Union Square incident. But how does the Williamsburg Bridge figure into it? Didn't that incident happen on the upgrade onto the bridge? I wouldn't think excessive speed would be a concern on the upgrade.

And as I recall that collision was caused by non-comnpliance with a stop signal combined with antiquated short signal spacing from a bygone era and the train's brakes failed to stop the train in the distance specified in the brake design. So again, how would timed signals have made any difference on that upgrade?

The signals on the Williamsburg Bridge were not capable of protecting a train operating at the maximum attainable speed for the area from colliding with the one ahead of it, which is exactly what happened on that morning in 1995. This was in part because of changing equipment performance since the signals' installation in 1918 and in part because train brakes in the mid-1990s were absolute garbage, but both issues prompted agency action. The former was resolved by slowing trains through areas of the systems with deficient control lines or lengthening said control lines (all the while reducing train acceleration to reduce the number of mods that had to be done), while the latter was addressed through increases in brake cylinder pressure.

Union Square certainly does factor into this story, but the direct consequence of that accident was the installation of WD signals to protect switches (though ofc those have now by and large been superseded by DGTs); it got the ball rolling on "oh $h!t our signals/brakes have serious safety issues" as it was in the years between 91 and 95 in which many of the studies that informed agency actions post-WillyB were contracted, but action came mostly following the latter accident.

Some good reading on this subject, courtesy of @Union Turnpike:

https://drive.google.com/file/d/1ICyGZN ... sp=sharing

Thanks for all the info guys. One other question: You mentioned control lines. Is that the distance between signals, or length of the blocks or something else?

Head-end View wrote: ↑Fri Dec 27, 2019 8:53 pm Thanks for all the info guys. One other question: You mentioned control lines. Is that the distance between signals, or length of the blocks or something else?

Control line is the length of track ahead of a given signal whose occupancy makes the signal red. For a better explanation, highly recommend this page (control lines are explained in the section on single line diagrams):
https://www.nycsubway.org/wiki/Subway_S ... lete_Guide

Everyone: Fascinating subject here with good feedback...

I noticed that this discussion brought up the 1995 Williamsburg Bridge accident - this occurred on June 5,
1995 at 6:18 AM in which two trains rear-ended one another on the bridge approach - for some reason I
mistakenly thought that this accident occurred between two trains going down grade and not up grade
as I would later find out. The cause was the motorman of the second train running drowsy missing signals
or other indications that he was getting too close to the train ahead. Since he was the only fatality in this
wreck the actual cause may never be exactly known such as if there was a sleep apnea or a related issue.

https://en.wikipedia.org/wiki/1995_Will ... bway_crash

I remember from my subway rides over time across both the Manhattan and Williamsburg Bridges that there
was a routine that train operators would use in crossing the bridges which is to pull power on the upgrade to
get the train to the first bridge tower and then use just enough speed to get to the crest in the middle of the
main span and then use the downgrade momentum to coast down towards the tunnels on either end (or the
elevated structure eastbound off the Williamsburg Bridge) going along with the times signals depending on
how close the train ahead was.

I also learned that both the Manhattan and Williamsburg Bridges had problems with their main suspension
support cables that support the center bridge span and that they were supposed to "flex" somewhat with
the weight of a train going across - these had become too rigid over the years of low maintenance and
without there being "give" or otherwise slack in these cables that the bridge itself could fail. This problem supposedly was addressed when both bridges were rehabilitated.

From what I have seen the same protocol is used by train operators crossing both of
these bridges today - is to pull power on the upgrade; run slow and easy across the
main span and then finish up with a good downgrade coast after passing the second
tower off the main span. I took note to the signals and how they kept trains spaced
apart with a train running up ahead - they worked fine from what I remember.

With the Williamsburg Bridge crash how did the train operator of the second train not
get the brakes applied by the signal trippers? Was it that he was going too fast on the
upgrade pulling power for the tripper(s) to work correctly when the collision happened?

MACTRAXX

According to news accounts after the Williamsburg Bridge crash, when the second train passed the stop signal, the trip arm did cause the train to dump. However, the train failed to stop in the distance that it was designed to. In other words the brakes didn't work as well as they should have. An additional factor was that the signal blocks on the bridge which dated back to 1918 were too short for the length and weight of modern trains, so even had the brakes functioned as designed it's possible the collision might still have happened.

Head-end View wrote: ↑Sat Dec 28, 2019 8:41 pm According to news accounts after the Williamsburg Bridge crash, when the second train passed the stop signal, the trip arm did cause the train to dump. However, the train failed to stop in the distance that it was designed to. In other words the brakes didn't work as well as they should have. An additional factor was that the signal blocks on the bridge which dated back to 1918 were too short for the length and weight of modern trains, so even had the brakes functioned as designed it's possible the collision might still have happened.

The probably would still have happened even if the brakes were performing to spec. I believe that the document I linked states that the accident signal had a sub-100% margin of safety even when assuming normal brake function.

As for the operating procedure over bridges, that is indeed what they do, but it's equally a factor of signal design as it is best practice: you cannot accelerate beyond the first tower on either span, as that's where the timers begin.

For a view of the Williamsburg Bridge as it existed at the time of the accident, see this video:
https://youtu.be/JLyDJTk4tIo?t=918
The signal implicated in the accident is at right at the linked time.