Home View from Underground How signal timers after a 1995 subway crash led to NYC’s current transit crisis

How signal timers after a 1995 subway crash led to NYC’s current transit crisis

by Benjamin Kabak

A fatal J train crash on the Williamsburg Bridge consumed the front page of The Daily News on June 6, 1995 .

Early in the morning on June 5, 1995, Layton Gibson blew a red light on the Williamsburg Bridge and crashed his J train into a stopped M train. Gibson was going only about 20 miles per hour when his train collided with the one in front of him, and he died instantly as 54 passengers were hurt in the crash. A subsequent investigation determined that both human and mechanical errors led to the crash, and while Gibson did try to stop when he saw the M train ahead of him, the then-77-year-old signal system did not trip an emergency brake in enough time to avoid the collision.

Now, 23 years later, the MTA still hasn’t gotten around to beginning to replace that signal system, and it’s now 100 years old, a dubious achievement indeed. But in the aftermath of the crash, the MTA determined that many of the then-septuagenarian signals were too close together to adequate stop modern trains faster than their late-1910s counterparts, and thus, the MTA in conjunction with Parsons Brinckerhoff began to implement speed limits throughout the subway system. Richard Perez-Pena, writing for The Times back in 1995, told the story. The emphasis is mine.

About one-third of the system has the antiquated signals that are considered candidates for speed limits, though officials said not all of them will require one. Just how many will is still under study. Joseph R. Hofmann, senior vice president in charge of subways for the Transit Authority, an arm of the M.T.A., said the plan to improve the trains’ emergency brakes would eventually make many, if not all, of the slow-speed orders unnecessary. But that program will not be completed for more than two years, and in the meantime trains will have to slow down in dozens of places where no speed limits were in place before…In any place where there is not adequate stopping distance, the report said, the Transit Authority “must institute speed restrictions or take some alternate immediate action to insure at least 100 percent stopping distance.”

The PB report indicated that more thousands of signals did not have adequate stopping distances, and the MTA instituted immediate speed restrictions on some lines that, The Times said, added around four minutes of travel time. Over the years, those speed restrictions seemed to become a way of life and faded into the background of the subway crisis in which we are mired.

But a few months ago, I noticed something peculiar. Southbound 6 trains heading from 51st St. to Grand Central were inching through the tunnel, and it seemed a timer was to blame. I had noticed a few of these slower trips over the years; for instance, 2 and 3 trains heading north into 96th St. now begin to slow down as early as 86th when they used to run at speed well past the abandoned 91st St. station. I started asking around and heard the tales of new timers identified in the wake of that 1995 crash being freshly implemented and of a group trying to fight back against timers. No one could explain why the timers were necessary now, and nearly everyone involved with decision-making at NYC Transit and the MTA in 1995 has long since left the agency.

Aaron Gordon started asking around too, and he published this piece in The Village Voice last week on slower trains. It is a must-read on the state of the MTA. The gist of it is that much of the MTA’s woes, especially with regards to their default view that “overcrowding” is causing delays, are self-inflicted. The MTA slowed down the trains, and thus service is worse. I’ll excerpt here as Gordon picks up with the post-1995 changes to the speed limits:

NYCT had predicted that the signal modifications would only marginally affect run times. But the 2014 study — the first time the authority had attempted to analyze the impact of any of the revamped signals, using its improved data system — found 2,851 lost total passenger hours per weekday could be attributed to thirteen modified signals alone. That was almost double the predicted impact; for comparison, the modifications of those thirteen signals alone created 5 percent as much lost time as that experienced by riders of the entire London Underground on its average day…

NYCT’s estimates were so off in part because they didn’t account for a human element. The most problematic of the newly modified signals were “one-shot timers,” so called because the operator has only one chance to meet the posted speed limit. One-shot timers are easier and cheaper to install, say MTA sources; the more “shots” the operators have to get under the required speed, the more timing mechanisms have to be installed across a longer portion of track. (An MTA spokesperson disputed this, characterizing the decision to install one- or two-shot timers as safety-related.) But the consequences of going over the speed limit are high — the train is stopped, and the operator gets penalized — so many operators now opt to go well below the posted speed limit just to be on the safe side.

“The grade time signals force us to operate slower, and because they have been installing them gradually, the subways have been slowing down gradually,” the train operator told the Voice. “I used to be able to go from 125th Street to 59th Street on the A line in seven minutes. Now it takes around nine minutes!”

Gordon has talked with some train operators and others within the MTA who are laser-focused on the signal timer issue, and I have spoken with some within the MTA who have largely echoed Gordon’s reporting. A group within the MTA is working like mad to overcome bureaucratic inertia and resistance to rocking the post-1995 boat in an effort to better assess whether timers are necessary both in terms of safety and in terms of cost of worse service. If trains are running at slower speeds, then the MTA simply cannot run as many trains per hour as it needs to in order to meet service demands or as many trains per hour as the system was built to handle. Thus, bad service — slower service, more crowded service — is a choice the MTA has made and is not, as Gordon notes, due to overcrowding but rather a cause of it.

But Andy Byford, the new man in charge of Transit, seems to have his ear to the ground on this issue. In this week’s MTA Board materials, he states that he has “asked my team to look in detail at the effect on line capacity of various signaling changes that were progressively introduced as a result of the Williamsburg Bridge crash of 1995.” He also discusses his hopes to see if these changes can be “safely mitigated.” Gordon received a similar response from the new New York City Transit president for his article:

In an email to the Village Voice, Byford acknowledged that “changes made to the signal system (in response to a crash in 1995) have undoubtedly had an impact on subway capacity,” and added that NYCT leadership has already met to begin reviewing the issue. “We are studying the impact and what was done to see if adjustments can be made while still maintaining the safety benefit these changes (and more onerous flagging arrangements) were brought in to address. This renewed scrutiny is part of my drive to properly understand — then tackle — root cause. As I have already said, the real fix is renewed signal systems and that will be the anchor behind my Corporate Plan that I am currently working on.”

Gordon’s piece ends with a sad story from a TO who feels the MTA is trying to slow down trains at a time when New Yorkers need better subway service. I’ll leave you with the quote I gave Aaron for his story: “If you’re running trains at speeds that are lower than what the system capacity was designed for, then you’re losing capacity, and that’s a choice that you’ve made. If that’s a choice that you’ve made, then you have to prove to people why you’ve made that choice. And if it’s a question of a hypothetical safety situation that doesn’t come into play, then you kind of have to question how that analysis has been reached.” It’s now up to Byford to tell us why these signals are necessary and how the MTA can maintain safety standards while reducing timer-related delays and improving service.

You may also like


BBM4 March 19, 2018 - 12:57 am

Keep the war on timers going. In all honesty, the MTA does NOT need CBTC in the whole system to increase service. Upgrade the current one to modern standards, and watch the need for CBTC disappear.

Also, if all the trains can move like how the 179s move (in terms of acceleration/braking), that should also help trains move faster, as the 179s fly down the elevated like how the 143s/160s move on the L line. Combined with a reduction in timers, the “rapid” will be placed back in rapid transit.

Larry Littlefield March 19, 2018 - 7:35 am

“Now, 23 years later, the MTA still hasn’t gotten around to beginning to replace that signal system, and it’s now 100 years old.”

To be fair, that isn’t quite true. The signals on the bridge, and on the entire BMT Broadway Line, were replaced about the time of the accident — with a new, old-style system that includes plenty of GTs.

The MTA has been replacing signal systems at about a 60-year rate. Except for the 1970s and the past 15 years, when signal replacement was cut — in the latter case because costs exploded, debts exploded, and any money left was sent elsewhere in the state.

The IRT and BMT signals were replaced. The IND signals are the ones over 75 years old. But the MTA is now so far behind, that the original IRT replacements may hit age 75 before they are replaced.

Which could work out, because the MTA is being charged such a ridiculous amount that it really needs to go with an entirely new system and set of contractors.

Larry Littlefield March 19, 2018 - 7:37 am

The situation NYC Transit finds itself in with regard to ongoing normal replacement and expansion is best explained by the I Love Lucy skit of Lucy and Ethel working at the chocolate factory, getting behind early and not being able to catch up.

Stephen Bauman March 19, 2018 - 7:43 am

A subsequent investigation determined that both human and mechanical errors led to the crash, and while Gibson did try to stop when he saw the M train ahead of him, the then-77-year-old signal system did not trip an emergency brake in enough time to avoid the collision.

The NTSB investigated this collision. They ran simulated tests over the same tracks, with the same type of equipment, went past the same red signal at full speed. Two tests revealed that the train traveling 34 mph when the red signal tripped the train. The trains stopped 76.5 and 70.5 feet beyond the collision point.

A simple analysis shows that the emergency braking rate averaged 1.8 mph/sec for the two test runs. Trains are supposed to have an emergency braking rate of 3.0 mph/sec.

NYCT stated out that the maximum attainable speed (MAS) at that signal was supposed to be 27.9 mph. NYCT noted that had the train been traveling at that speed when it hit the tripper, the train would have stopped before impact.

Analysis shows that the train would have stopped 44.6 feet short of impact with a braking rate of 1.8 mph/sec. Analysis also shows that had the train’s brake performed as per the 3.0 mph/sec spec and traveling at 34 mph, it would have stopped 48.5 feet short of the impact point.

NYCT seized upon the train’s higher speed rather than the train’s poor braking performance as the collision’s cause. It’s been living with the consequences of that missed diagnosis ever since.

J Adlai March 21, 2018 - 8:21 am

If I remember correctly, the report stated that the NTSB tested braking performance using full service braking and even at the higher speed the train stopped in a sufficient amount of distance. The lack of dynamic braking in emergency applications seemed to hamper the ability of the train to stop. This seems to be a problem that should have been corrected, in light of the conclusions reached by the FRA after the 1989 San Bernardino train disaster.

Stephen Bauman March 21, 2018 - 9:23 am

Yes they did. The same train that went past the collision point by 73.5 feet with emergency braking, missed the collision by 126 feet with full service braking. One NTSB recommendation the MTA ignored was to apply dynamic braking during emergencies.

N.B. the air brakes alone must be able to stop the train in an emergency within a safe distance. Dynamic brakes require third rail power and working motors. Trains still need to stop within the same distance, in the event of a third rail power failure.

J Adlai April 1, 2018 - 10:32 pm

It would seem that there are basic current technologies that could mitigate that problem.

Michael Finfer March 19, 2018 - 8:38 am

I have been wondering about all the extra timers for a long time. The one approaching 96th and Broadway on the express track is one of the ones that puzzled me. Why was it not arranged so that the timer is only active if there’s a stop signal displayed at 96th St. interlocking? As long as that signal is clear, the need for a timer at 91St is beyond me.

The other place I have to question is the Broadway Line uptown express track. There is a timer at 49th St. and another one entering the station at 57th St. That was fine when 57th St. was a terminal, but why are those timers there when a clear signal is displayed at the north end of 57th St. for the run up Second Avenue? The trains have to slow to a crawl at both of those locations. If the signal is displayed at 57th St to go uptown, those signals should also be clear.

Let’s begin a list of unnecessary timers here so we can compare it to what the NYCTA eventually comes up with.

AMH March 19, 2018 - 10:30 am

I’ve also wondered about those, and others where there’s an interlocking and/or a slight curve (but the northbound Lex express, when running well at least, barrels into Union Square at full speed under those conditions).

My additions:

s/b 2/3 approaching Times Square
s/b 2/3 approaching Chambers Street
s/b 4/5 leaving 125th Street
s/b 4/5 from Wall Street to Bowling Green
4/5 in both directions in the Harlem River Tubes
On another topic, I’ve noticed warped track on some express straightaways that causes the train to jerk to one side. I’m curious whether the CWR upgrades will do anything about that.

J Adlai March 21, 2018 - 8:27 am

How about the A line?

-Timers on straight track between 145th and 125th that make trains slower than the locals in both directions.

-The Southbound express from 125th to 59th is a continuous string of timers from just south of 86th to 59th street. The northbound is a continuous string of timers from 86th to 125th.

-Northbound trains are hit by timers pretty hard between 42nd and 59th, and again between 14th and 34th. Southbound trains used to only be slowed entering Chambers if there was a train crossing in the station, but now they are all slowed. I’ll stop here, but the Brooklyn segment has lots of timers as well.

I did an analysis looking at all of the track geometry of the A line and determined that if trains were a bit faster, and there were no timers, the A line could shave off 12-13 minutes of travel time one-way.

Bgriff March 19, 2018 - 12:43 pm

Agree on the uptown Broadway line coming into 57th.

Another one I’ve never understood is the uptown 8th Avenue express track between 14th and 59th. The local train frequently beats the A throughout that stretch. And meanwhile the A train flies along on the downtown track in the same area (though perhaps we shouldn’t point that out to the MTA).

BruceNY March 19, 2018 - 1:57 pm

The Broadway Line needs to be fixed at 49th Street to be sure.

I was told that uptown 8th Ave. expresses are made to slow down around 23rd St. prior to going over a switch that leads to the now disused lower level. Too bad they just can’t take out the switch now.

Another one is the Uptown 7th Ave. express at Christopher Street. Why are trains slowed down here?

The Lexington Express between 14th and Brooklyn Bridge used to be a really fun roller coaster ride, but now is ho-hum since it goes so much slower now.

There was also that horrific crash by a downtown Lexington Express heading into Union Sq. whose driver was high on drugs. I always thought that crash was the main reason for all the slow-downs, more so than the crash on the Williamsburg Bridge.

TomD March 19, 2018 - 9:03 am

It’s not just the subways. Metro-North has also been slowed down in recent years. Express trips between White Plains and Grand Central that used to be scheduled for 31-34 minutes are now listed at nearly 40 minutes. Furthermore, the older schedules were actually being met with trains were arriving on time. Today it seems that, even with the slower schedules, trains arrive on time less often.

Also telling is the fact that Metro-North no longer publishes on-time performance (at least, I haven’t seen it in years); before Metro-North’s 2013 Sputen Duyvel wreck, Metro-North always touted it’s on-time performance. Now, they seem to think that being on-time means the train was operated unsafely.

SEAN March 20, 2018 - 3:40 pm

I understand where you are coming from, but the time difference is only 2-minutes. The additional time is fat added at Grand Central. I’ve come across schedules in Stamford with minutes deleted on certain express trains.

rob March 19, 2018 - 11:45 am

What happened to ‘the plan to improve the trains’ emergency brakes’? Done?

Michael Thaddeus Doyle March 19, 2018 - 12:47 pm

I said this privately to Aaron on Twitter and I’m happy to say it publicly here. Back in the late-90s/early 2000s when I was transportation planner and then associate director of the Permanent Citizens Advisory Committee to the MTA, of which the New York City Transit Riders Council is a component part, The internal justification for making the signal and propulsion changes across the entire system was different then the public justification. Then as now, the independent PCAC was housed at MTA headquarters, working in extraordinarily close contact with agency and departmental leadership (including then MTA Planning director Bill Wheeler.) And it was privately made clear to us on more than one occasion by MTA leadership that the changes would be rolled out as widely as possible not simply for safety but to avoid legal liability. They told us they were doing it so the MTA would have less of a chance of being sued the next time a rider was injured or killed in a subway accident–and that back-channel justification included not just the Williamsburg Bridge incident, but the earlier and deadly Union Square IRT derailment as well. They would not say that publicly so we could not say that publicly. But we definitively heard it. It was a reason that stank then and it stinks even worse now considering how badly the slowdown damaged the performance.

Shortly after implementing the changes, the MTA came out with “wait assessment” as a replacement performance measure. In response, PCAC released a report which I authored, Timing Is Everything, that complained that wait assessment made it appear that service was getting better, when the performance measures that wait assessment replaced actually showed that performance had gotten worse. We didn’t put it together at the time, but it’s pretty clear to me now that wait assessment was a cover for the immediate degradation in service that the signal and propulsion changes caused. They used safety as a cover to avoid liability by degrading service and then hid the degradation behind a phony performance measure. Go ask Bill Wheeler about his conversations with Beverly Dolinsky on the matter.

Larry Littlefield March 20, 2018 - 9:53 am

The degradation of service is likely to lead to a mass casualty event eventually. Last week the conditions I fear played out I front of me.

A signal failure at Bergen Street took out the F and G trains at rush hour. A, C, and some diverted F trains started depositing more and more passengers at Jay Street. Many headed down the stairs to the R train, and pretty soon the platform was packed. Some smart people started getting out before that became impossible.

But those on the upper level didn’t see this, and were getting on the down escalator, one after the other, nose to tail. What happened if when they got to the bottom there was no where to go, with people piling up behind them? Panic, pushing, and then a bunch of people falling to the tracks, to be hit by an oncoming train.

Perhaps the only thing that prevented this was there was no R train, thanks to a 12-9 at Whitehall.

This will happen sooner or later. It would be very ironic if avoiding litigation is the cause.

Stephen Bauman March 20, 2018 - 12:32 pm

My unverified sources report that it was the legal department that cause the emergency braking performance to be degraded. They were worried that effective emergency brakes could cause passenger injuries.

Larry Littlefield March 20, 2018 - 2:27 pm

I remember that. There were lots of cords being pulled and other “brakes in emergency” situations back in the bad old days. Subway riders were being injured by being thrown down by rapid braking.

al March 23, 2018 - 1:19 pm

Perhaps modify the pull cords so that a single pull would result in service braking application, but a continuous pull or double pull result in emergency braking.

AMH March 26, 2018 - 1:41 pm

They’ve already been modified on the new tech cars so that pulling the handle will hold the train in the next station rather than tripping the brakes.

smotri March 20, 2018 - 2:38 pm

It’s clear the MTA needs to go. It cannot run a railroad with any kind of efficiency or regard to safety and now, it seems, with any honesty – the MTA cooks its books, does not know how to contract for projects such as the Second Avenue Subway without bloated costs and years’-long delays and, as we now see with these timers, figures out ways to cover up the truth in such a simple area as train speeds, chalking it all up to overcrowding. With today’s accident in the Bronx, when several lines were seriously impacted because of the MTA’s incompetence, thousands upon thousands of people had to put up with giant delays. Enough!

johndmuller March 20, 2018 - 7:53 pm

In a way, it’s good to hear that a lot of this poor performance is due to legal CYA behavior. I had been thinking that the reason expresses seemed to make no better time than locals was that the express tracks were in poor condition; to hear that it is overuse of timers meas that trackways do not have to be rebuilt to speed things up, just backbones.

J Adlai March 21, 2018 - 8:37 am

This issue has been a pet peeve of mine for years, and I am glad someone is finally paying attention to this.

There are two types of timing signals on the NYC subway system: Grade Timers and Station Timers. Grade Timers are always in effect. The idea is that they control speed around curves and on downgrades where trains could pick up more speed than intended. Station Timers are designed to slow trains down as they approach a train ahead so that the trains can get closer to one another and capacity is maintained. Therefore Station Timers are only active when there is a train ahead (you can sometimes see them when you see a red signal in a station with an illuminated number at the bottom of the signal).

Guess which one the MTA used for preventing collisions?

“Timer-mania” goes beyond timers installed to prevent collisions. NYCT began installing timers as a solution to any speed related issue. Anecdotally, I was told of an incident where a higher up felt that some trains rounding a curve exceeded the ‘comfort speed’ of the curve. Though there was no safety issue, NYCT was prepared to install multiple timers to restrict speeds to more than 10 mph below that comfort speed. Thankfully, that one got torpedoed. But there are many more timers out there that got implemented anyway, all serving to unnecessarily lengthen everyone’s commute, and in some cases, reduce capacity.

Thomas Schmidt March 22, 2018 - 1:15 pm

A human life of 75 years runs about 650,000 hours long. If 13 timers cost 2851 hours in lost time every workday, then every 230or so workdays the MTA kills off one human lifespan of hours. That’s about 1.2 people per year.

Over 20 years, that’s a lot of wasted time, and that’s only counting 13 timers.


Leave a Comment