To Improve Subway Service, Get Rid of the Unnecessary Signals Slowing Trains Down

If timers are needed to safeguard against a clear and present danger, whether that be a curve, switch, or hill - keep them. If not, remove them.

Headways ain't what they used to be.
Headways ain't what they used to be.

Last week, Uday Schultz, a junior at St. Ann’s School, took the top prize at TransitCenter’s annual TransitSlam with a presentation about the MTA’s excessive use of signals that cap subway speeds, which he produced with his classmate, Ivan Specht. We’re pleased to present a blog-ified version of their show below.

For the first time in a generation, New Yorkers are giving up on the subway as severe delays mount and reliability plummets. The trains have simply ceased to be a predictable way to get around the city.

If we are to stop this mayhem, a massive rethink of how our transit system is operated, funded, and managed is needed. But there are also basic steps we can take in the short run to improve service. One of these steps is the rationalization of subway timers.

These signals, which limit train speeds, have been installed unnecessarily, lengthening trip times and hampering subway reliability.

To understand how these devices came to be such a problem, let’s start back in the 1990s. Two fatal train crashes that decade resulted from subways traveling at excessive speeds. In response, the National Transportation Safety Board urged the MTA to address the issue “by converting more automatic signals to grade time signals.”

Instead of simply regulating the distance between trains, grade time signals place strict speed limits on trains. The idea was to prevent crashes by adding these signals in areas where excessive speed could be deadly — near sharp curves, on steep hills, and at busy junctions.

A timer signal on the Williamsburg Bridge.
A timer signal on the Williamsburg Bridge.

Initially, that’s what the MTA did, and there was a logic to the placement of the new, speed-limiting signals. However, the agency did not stop there. The MTA kept adding timers, even to safe areas of track.

The effects are twofold. Most directly, these speed restrictions have lengthened subway trips. Take the 5 train, which serves the busiest subway corridor in the nation. In 2005, a trip from 180th Street to 149th Street took nine minutes. Today, it takes 11 minutes. Similarly, going from Grand Central to Brooklyn Bridge in 2005 took 10 minutes; today it takes 12.

It may not sound like much, but these small increases in travel time compound across the whole subway system, adding up to countless hours lost for riders.

More relevant to the current crisis, timer signals have made the system significantly less reliable. Before their proliferation, a late train could make up time by going just a bit faster. Today, all trains — late or not — have to constantly slow for these speed checkpoints. Trains that fall behind schedule stay behind, preserving service gaps, adding to crowding, and making commutes miserable.

In a city obsessed with time and speed, this unnecessary slowdown is unacceptable. To get subway service back to where it needs to be, something must be done about it.

If timers are needed to safeguard against a clear and present danger, whether that be a curve, switch, or hill — keep them. If not, remove them.

Removing unneeded timers will not solve every problem with the subway system. No one policy fix can do that. But it’s a clear opportunity to make the system perform significantly better. We would be foolish not to take it.

  • Joe R.

    Same line of reasoning applies to neutering the trains. The resulting poor acceleration adds about 10 seconds per local stop. Coupled with the timers on express portions of a run, you can easily add 5 minutes to many typical trips.

    Note also that you only need timers on a switch if it’s in the diverging position. I’m not really sold on the timers on steep hills, either, unless there is a curve at the bottom of the hill where the excessive speed would be an issue. In the East River tunnels especially you want to build up as much speed as possible going downhill to carry you back uphill. If you don’t, the end result is slogging uphill at something like 17 mph.

  • Anonymous

    Truly, “ex ore parvulorum veritas” (from the mouths of the little ones comes the truth). This subject was picked up in the comments to various articles about the MTA, from time to time, but never has there been a proper story on it.

    The actual truth is, of course, more complicated. Many of those grade timers were installed en masse thanks to Fastrack – more efficient maintenance schedules created the perverse opportunity to get work done that should’ve been done a decade earlier. The timers were installed based on ridership and throughput analyses from that previous decade, without re-evaluating them for current conditions.

    Earlier claims that they won’t reduce throughput no longer held true, in some cases, due to (a) ridership growth and the changes in running times it caused; (b) massive turnover in staff, resulting in less experienced (and more cautious) staff, on average; (c) equipment not always performing with precision – mechanical timers slow down over their lifetime; (d) mistrust and lack of communication between the engineers, maintainers, and operating staff, resulting in excessive caution; (e) a punitive “safety” culture that avoids looking at root causes of incidents, and the resulting fear of being punished when overrunning a red signal even when the problem is with the equipment instead of the employee.

  • Russell.FL

    One egregious timer that I encounter regularly is the A train uptown from 14th street to 34th street. The A train crawls uptown super slowly for no apparent reason. If a local train leaves at the same time, it will often beat the A, that’s how slow it goes. It makes no sense, as the downtown A on the same segment rockets down from 34th to 14th.

  • Robert Hale

    THANK YOU, this piece NEEDED to be written and disseminated!

  • Larry Littlefield

    Does anyone know how much the L was able to speed up, thanks to CBTC? It was supposed to restore speed as well as capacity to the system.

  • Joe R.

    I think it was 4 minutes, which is a huge amount on a line which has only local stops. It just shows how much time is lost when trains aren’t accelerating at their full capability.

  • Sunny

    The problem is not just with the speed restrictions, but the very nature by which timers work, which makes them far more capacity-reducing than just a speed restriction

    Imagine you’re driving down a street, and the speed limit is 20mph. There are a row of traffic lights, each of which is red except for the first. The signals are timed such that if you go at exactly 20mph and everything works as designed, each successive signal will turn green at the very instant you pass it. If you run a red light, you’ll get a ticket. What will you do in this case? Obviously you’ll drive far slower than 20mph, so that you can see the signal turn green in front of you before you pass it, and perhaps give yourself enough room to stop safely if one of them doesn’t work and stays red.

    This is how one-shot timers work, essentially. (For an example, watch this video: Only a train has a much longer stopping distance than a car, the timers in the subway are often calibrated incorrectly, and if one runs a red signal the train will apply its emergency brakes immediately. So train operators will operate far slower than what the sign indicates, because they want to give themselves a buffer if the timer is miscalibrated or malfunctions – if the sign says 20mph, I better go 10-15mph just in case. This not only leads to a terribly rough ride as each signal is encountered, but also reduces throughput far more than the plain 20mph restriction.

  • ortcutt

    With CBTC, they should be able to designate the safe speed for every meter of track. I hope that is one improvement that will come with improved signaling technology.

  • joe lee

    WOW !!!! I like that (R) train countdown clock shown above…..It seem very true, since the
    (R) train for rancid, rarely, or rotten train always runs like crap anyway with very long waits between (R) trains, so a 25 minute wait is typical for this line.

    I remember that the (R) formerly (RR) was one of the BEST lines in the system that you can essentially set your watch by, when it ran between Astoria-Ditmars in Queens, and Bay Ridge-95 Street in Brooklyn with its one hour and four minutes running time, verses the present one hour and thirty six minutes from Forest Hills-71 Avenue in Queens instead. Operating it from Forest Hills has made the (R) a very long and unreliable local line, similar to the old (QJ) line from 1967 to the end of 1973…..

  • Joe R.

    That TO is being a bit overly cautious. I’ve seen some that pass the second signal just as it’s flipping to green. I do something similar riding my bike on roads where I have the signal timing down. Only difference is I don’t get an emergency brake application if I hit the red 1/10th of a second before it turns.

    On another note, the trains run so slowly nowadays it’s like watching molasses dripping off a spoon. It seems like on every section of track once they pick up some decent speed they’re slamming on the brakes for yet another timer. We need to get the “rapid” back in rapid transit.

  • sbauman

    Scheduled running time is still 35 to 40.5 minutes between Canarsie and 8th Ave. The schedule is padded to make the OTP figures look better.

    The real time data shows that trains go faster. This is also true of other lines, so CBTC wasn’t the reason .

  • Isn’t the real issue here the failure to modernize New York’s subway signaling more widely? I moved back last year from New York to London. London now has communications-based train control on the Central Line, Northern Line, Jubilee Line, Victoria Line and Waterloo & City. It’s close to bringing it in on the sub-surface lines – the Metropolitan, District, Circle and Hammersmith & City, which account for 40 per cent of route miles. During my four years in New York, while there was progress in installing the system on the 7 train and there’s work under way on the IND’s Flushing line, it wasn’t introduced on a single new line. It’s still operational only on the L Train. We’ve seen from the horrible Amtrak crash this week and past overspeed crashes that there is a serious problem with human drivers’ ability to reduce speed at the correct places, so I can appreciate why a cautious MTA management might have introduced increasing numbers of these timers. The issue is surely to move away from mechanical signaling with color lights and towards CBTC.

  • sbauman

    Same line of reasoning applies to neutering the trains. The resulting poor acceleration adds about 10 seconds per local stop.

    It’s not that bad. Current rating for acceleration is 2.5 mph/sec. Suppose the train were operating at 2.0 mph/sec because the train were sent out with a few non-functioning motors. The time to reach 30 mph would be 15 sec.

    To reduce that by 10 seconds to 5 sec, would require an acceleration of 6.0 mph/sec. That acceleration never was, even on the Multi’s whose acceleration was rated at 4.0 mph/sec.

    There are ways the MTA manages to waste 10 seconds per local stop. It’s done in two increments.

    The first is the delay in opening the doors. Conductors, by themselves, demonstrated they cannot be trusted to determine which side of the train to open. The MTA’s response was to install “door enabler” circuits. These require both the train operator and the conductor to agree on which side to open. If they disagree, the doors don’t open. Also, also conductors must open their window and point the indication board before opening the doors. There may be inspectors on platforms to monitor this procedure. This adds a few seconds to station dwell times.

    The second delay comes when doors close. Conductors are stationed in the middle of the train and close the rear and front sections in sequence. Most other systems opted to place the conductor at the rear of the train and close all the doors at once. The MTA’s procedure means that door closing time is nearly double than if all doors closed at once.

    The second door closing time waster is how conductors handle doors that are held open by passengers. The re-open all the doors to let the passengers enter. The result is more riders try to enter at doors that had previously closed. This results in another cycle of door openings and closings. The proper solution would be to open only the door that had not fully closed. The new technology trains were supposed to have this feature. It was disabled by the MTA.

    The wasted time in opening and closing train doors at stations accounts for the 10 second delay you have attributed to poor acceleration.

  • sbauman

    Signal systems do not affect maximum service levels (trains per hour). That’s determined by train braking and acceleration rates as well as dwell time within stations. The L train used to operate 24 tph and the 7 train 36 tph. The maximum service levels anticipated for CBTC on these lines is 22 and 30 tph, respectively.

    The Williamsburg Bridge collision, that resulted in the slow downs, was caused by emergency brakes that did not perform as per spec.

    Subsequent tests revealed an emergency braking rate of 1.8 mph/sec vs. the 3.0 mph/sec spec. The tests also revealed that the train speed at the tripper was 34 mph instead of 27.9 mph which was thought to be the maximum attainable speed. The MTA seized upon the latter discrepancy as the collision’s primary cause. They noted that had the train been traveling at the lower speed, it would have missed the collision by 44.6 feet. They did not note that had the emergency brakes performed as per spec and the train been traveling at 34 mph, it would have missed the collision by 48.5 feet.

    Placing speed controls was one of the NTSB’s recommendations. The report also noted that two braking systems are used: dynamic (electrical) and air. Only the air brakes are applied during emergency braking. The NTSB also ran tests using service braking (dynamic + air). The train stopped 126 feet short of the collision. The NTSB also recommended that dynamic brakes also be applied during an emergency. This has not been implemented on previous train models nor on the NTT models purchased after the Williamsburg Bridge collision.

  • Samuelitooooo

    Many times, conductors reopening the doors prevents further delays because otherwise, those passengers would just hold the doors until they get in. That could last seconds and seconds.

  • sbauman

    Perhaps, I wasn’t clear. Only the door that has not closed should be re-opened – not all the doors. Those that have already fully closed should remain closed. This capability was built into the NTT trains and has been disabled.

    The method I described will result in shorter dwell times than when all doors must be able to close simultaneously. In this case, sequential closing on a door-by-door basis provides for shorter dwell times.

  • lockenload

    Does anyone know why the south-bound R/W trains start crawling at slower speeds between Canal St and Whitehall St? The distance between stations remain the same and I wonder whether a speed restriction in that area or the curved tracks (which screech unbearably loud) are the reason?

  • lockenload

    Thanks for the helpful insight on dwell times. I definitely agree constantly re-opening doors, especially near stairs or turnstiles where riders pack in, causes delays. Perhaps future Platform Edge Doors may help alleviate the issue. Why did the MTA disable the feature to re-open a single door and can it be a part of the Subway Action Plan?

    Also, many lines even with the newer R160 cars suffer from rapid acceleration and braking (jolting) action. Sometimes a train starts braking even before it fully leaves a station! Will CBTC help with smoother spacing of trains? It’s dangerous and uncomfortable for any driver to tailgate in traffic, so why can’t a train operator ensure there are ready green and yellow signals before even accelerating? That way trains can coast to a stop and minimize braking, while also reducing the amount of steel dust that constant braking produces.

  • Signaling systems certainly should affect the maximum number of trains per hour, or at least they have in the UK. My understanding is that a lot of the waste with fixed-block signaling systems like the ones on the New York subway and still on some lines of the London underground is that each train needs an entire empty block – usually the length of track a train needs to stop from full service speed, plus a safety margin – behind it. Because those blocks are fixed, a train that’s just leaving one block will have nearly two whole blocks empty behind it, meaning trains can’t run very close together. A CBTC system should work out constantly how much empty space each train needs in front of it to stop from its current speed, plus a safety margin, and maintain only that space. That will obviously often be a shorter space than in a fixed-block system. On Paris’s Line 14, I’ve seen trains get within a couple of metres of each other at low speeds. London’s Victoria Line now operates at 100-second headways at peak times. Obviously, to maximise the potential of the signaling, it’s vital to improve train braking and to ensure excessive braking distances don’t need to be built into the system. But a decent CBTC system should allow trains to run safely at higher speeds and closer together, as far as I know.

  • thoughtfulcitizen89

    Ditto on the sentiments expressed in this post.

    Stations where there are parallel tracks offer the best example of this slowdown. When I was a child the A and F trains would leave Jay Street at the same level of speed and acceleration. You knew this because the trains spent sometime next to each other before the tracks diverged. If your friend was on the adjacent train, and you both departed at the same time, you could count on seeing them for roughly 25 seconds while you were speeding along together in the tunnel.

    Today, the A hasn’t changed, but the F slows down to what feels like 15mph as it leaves Jay Street. The Eight Avenue trains speed by as the F train slogs towards York Street. Your friend, if they were in adjacent train, would be long gone.

    At first I thought it was due to construction, but that was more than two years ago. Now it seems permanent. The result of this change is, perhaps, the addition of another 45 seconds of travel time between Jay and York. Doesn’t sound like much, but if you add that time between every stop, you could end up with a trip that takes ten minutes longer.

  • joe lee

    Yes buddy, its this very short stretch of very sharp curve south of the City Hall (R) line station that makes a very sharp curve away from Broadway, running directly under Saint Paul’s Church, then curving almost immediately under Church Street which turns into, and becomes Trinity Place southwards towards the Cortlandt Street station. You can see the grating for the curving BMT tracks within the church’s back yard areas.

    The main reason is that the IRT Lexington Av Line’s (4) & (5) already occupied Broadway from the Saint Paul’s Church southwards from Vesey Street down to Bowling Green since 1905, so the BMT’s (R) line had to change route paths during this period, and could not continue straight down Broadway.

  • AMH

    That’s insane–the F train even has the straight path!

  • AMH

    Good post–how do we get this done? I wonder if this is why 2 3 express trains slow to a crawl outside of Chambers St and 96 St (on straight track).

  • joe lee

    To answer your question, the (2) & (3) expresses goes and drops as it leaves the Chambers Street Station, and also goes around a 10 MPH very, very sharp turn onto Park Place, and into the Park Place station before going under City Hall Park to line up with Beekman Street, and travels under same, then makes another 10 MPH very sharp turn to get under William Street for the rest of the way to Old Slip, then from there into Brooklyn via the Clark Street tubes.

  • joe lee

    Yes, that’s because the (F) train has to slow down, and switch from the lower level express tracks, before rising up to the upper level local tracks used by the (G) coming from Hoyt Schermerhorn Street so that it can operate local with the (G) line from Bergen Street to Coney Island Stillwell Avenue Terminal. However the (G) terminates at Church Avenue on the same local tracks there.

  • joe lee

    Also, when they sent the (R) line to Forest Hills-71 Avenue in Queens, it became the slowest and most delay-prone route in the system.

  • Anonymity

    That’s Jay-Bergen. He’s talking Jay-York, which is poker straight, and does indeed have a 15mph timer on a straightaway.

  • joe lee

    That’s because you are traveling at a steep downhill slope or grade under Jay Street to let the (A) & (C) lines swerve left, and go above the (F) train tracks before entering York Street in Dumbo before proceeding further downhill into the Rutgers Street Tubes going further downhill under the East River, before going and proceeding uphill again into the East Broadway Station in Manhattan, and to protect the switches before entering the York Street station. They should raise the timer to at least 20MPH at that point !!!!

  • Anonymity

    Those switches are set to go main — they haven’t been used in years and are actually slated to be removed. The downgrade has no curve at its bottom, and given that train operators are, well, trained to control trains, I see no reason that we can’t trust them to take them down the grade/into York at a reasonable speed.

  • jaxbot

    >Signal systems do not affect maximum service levels (trains per hour). That’s determined by train braking and acceleration rates as well as dwell time within stations.

    I think that’s oversimplifying it, though. If you boil down CBTC to a series of timed signals every 20 ft, you can move trains closer together at low speeds and further apart at high speeds. The existing fixed block system, even with timers, does not really support that and will slow and stop trains even when it would be safe to proceed. In effect, you can move trains closer together and increase the TPH. I’m not sure if that affects the theoretical maximum, but in reality it makes a huge difference — if someone holds the doors for 30 seconds on the A train, it causes a ripple down the line, with trains getting held at signals and having to start and stop. If someone does that on the L, the system can compensate for that and trains can keep moving at lower speeds until the distances are restored.

  • jaxbot

    If you’re curious if your train line is affected by timed signals, get in the front car of the train and look out the front window (“the railfan window”). Track sections with timed signals will have a red-and-white aspect, or a white S, indicating speed control is in effect and the signal is only red as a speed restriction. Track sections without this will just have red, green or yellow aspects.

  • AMH

    No, the OP was describing a trip from Jay to York.

  • AMH

    I was describing trains approaching Chambers s/b, and 96 St n/b, which is all straight track with no sharp curves (I edited my original post to clarify this).

  • sbauman

    Signaling systems certainly should affect the maximum number of trains per hour

    That’s a misconception that was invented by signal manufacturers and accepted by gullible transit executives and advocates.

    My understanding is that a lot of the waste with fixed-block signaling systems like the ones on the New York subway

    You were misinformed. The paper detailing how to calculate the length and placement of signal blocks for maximum service levels was written by the BMT’s chief signal engineer back in the 1920’s. There wasn’t any waste in his methodology.

    Let’s assume two trains travel over the same track. The leader leader and follower travel at a constant 30 mph, with the leader departing 90 seconds before the follower. The leader will have traveled 4050′ before the follower starts. The follower’s braking distance is 225′. They are going at the same speed, so they will maintain the same distance between them.

    The only condition by which the follower can reduce this separation is when the leader is traveling slower than the follower. The only time this results in a close separation is when the follower is approaching a station that’s occupied by the leader. That’s the only place where the leader an follower positions need to be measured accurately.

    The “fixed” in a fixed block system refers to the blocks not moving with the train. The blocks are not a constant length. In NYC block lengths vary from 50′ to 1400′, depending on where they are placed. They are shorter at approaches to and within stations because that’s were finer positional accuracy and precision are required.

    There are 4 signals, defining 3 blocks that determine service level capacity. These signals are located at the station entrance (at the rear, signal # 1), a signal within the station (showing only red and yellow aspects in NYC, signal #2), a signal at the station exit (at the front, signal #3) and a signal in the tunnel that’s a nominal 500′ beyond the station exit (signal #4).

    Let’s assume an operation where the follower never sees a yellow or red light and travels at 30 mph on the green. A train enters a station at 30 mph. It will take the train 10 seconds to stop with the service braking rate of 3.0 mph/sec. It will have traveled 225 feet during braking. Let’s assume the platform is 600′ long. This means the train would have traveled 375′ (600-225) within the station at 30 mph before applying the brakes. It will have taken the train 8.3 seconds to travel that distance. The time from when the train enters the station at full speed (30 mph) until it comes to a complete stop is 18.3 seconds.

    The station dwell time is assumed to be a nominal 30 seconds. This means that 48.3 seconds have elapsed before the train is ready to start.

    Signal #1 will turn green, when the train clears the block boundary defined by signal #4. The train will accelerate at 2.5 mph/sec to reach 30 mph in 12 seconds. It will have traveled 270′. It needs to travel an additional 230′ for the train to reach signal #4. However, train length is 600′, so the train needs to travel 830′ at 30 mph to clear signal #4. This will take 18.4 seconds. Thus the departing train will take 30.4 (12.0 + 18.4) seconds from its departure before the signal #1 turns green and the follower can repeat the process. The total elapsed time was 78.7 seconds for a service level of 45.7 tph.

    These figures are in line with the theoretical derivation presented by Lang and Soberman in “Urban Rail Transit Its Economics and Technology.” The nominal figure of 90 second headways and 40 tph is what’s accepted for service level capacity at intermediate stations.

    These numbers were confirmed by the MTA in their SAS DEIS “The current NYCT signal system on the Lexington Avenue line is designed to allow 90-second headways, including a 30-second allowance for station dwell times.” An article in the Sept 1949 issue of Railway Signalling noted regarding the extension of the Fulton St IND line from Bway-ENY to Euclid Ave: “The signal system is laid out for a 90-second headway with 30-second station stops at local platforms and 45-second station stops at express platforms”

    at least they have in the UK…London’s Victoria Line now operates at 100-second headways at peak times.

    I’m not as familiar with the Underground as I am with NYC. Here’s what P. E Garbutt wrote in “How the Underground Works” (c) 1966. This book was published by London Transport. I bought it at the London Transport book shop, when I visited London in 1967.

    “The services in the central area during the peak periods are usually based on regular intervals ranging from 2 minutes down to 1 1/2 minutes, the latter figure being in practice the minimum interval which can be continusouly worked on a single track”

    On Paris’s Line 14, I’ve seen trains get within a couple of metres of each other at low speeds.

    That’s not a strategy that will result in high service levels. Going back to the previous example. Suppose the leader approaches the station at 5 mph. It’s stopping time and distance are 1.7 seconds and 6.26 ft. This means it must travel 593.7 @ 8 mph for 79.2 seconds. This brings its travel time from station entrance to complete stop to 77.5 seconds. Add the 30 second dwell time and the total elapsed time from station entrance to starting to leave becomes 107.5 seconds. It has already taken this train longer than the 78.7 seconds the full speed train required to enter, come to a stop, dwell in the station and clear it for the following train.

  • sbauman

    you can move trains closer together [at reduced speed] and increase the TPH.

    As I noted in my reply to Mr Wright, that strategy results in fewer tph not more.

  • joe lee

    They placed timers approaching Canal Street on the downtown express to slow the express trains down many, many years ago after a derailment that badly damaged R12 car 5784, and R17 car 6580 back in 1967, since they are approaching a curve from Varick Street onto West Broadway, and they don’t want the express trains barrelling into the Chambers Street express station, because there are many switches before entering Chamber Street station, plus that notoriously sharp curve just south of same on the express tracks. Those two cars derailed across the switches, and tracks, and were sliced opened by smashing into the crash walls separating the uptown, and downtown side of the subway. Also next time please think what you are trying to mention before you post, so that you don’t have to edit same.

    Thank You.

    Merry Christmas and Happy New Year.

  • joe lee

    Again yes, there’s a very sharp curve on the express tracks after it descends north of the 96 Street Station, where there are also many switches that needs to be protected, then the two express tracks makes a very tight and sharp right turn under W 104 Street to Central Park West, then curves slightly and goes directly under
    Central Park and takes another curve right onto Lenox Avenue (now Malcolm X Boulevard) into the Central Park North 110 Street Station in Harlem USA…….Anymore questions ????

  • joe lee

    There’s a slight curve when this line downsizes from a four tracked line at Jay Street Metrotech down to a regular two tracked line at York Street, since the (A) & (C) has already curved to the left, and turned onto Cranberry Street for their trip into Lower Manhattan.

    Just picture the (1) line going from 145 Street into 157 Street going from three tracks into two, the (1) line trains also must slow down proceeding northward towards 157 Street….

    Also remember the 1991 collision north of 14 Street Union Square Station on the (4) line which caused five R62 cars to be cut-up on the spot, and some damaged street support columns, when the motorman was drunk, and didn’t slow down at the switches directly north of this station ???

  • Anonymity

    You’re misunderstanding him. He is talking about approaching Chambers from the north, and 96th from the south.

  • Anonymity

    That slight curve can be taken at speed. I’ve seen trains do it before. In fact, I’ve seen trains take curves much sharper than it (Woodhaven on Queens Boulevard Expres) at 40, so I don’t see why this slight one can be done. And again, as to the union square point, the switches at York are quite literally never used. And if they were, timers could be installed that protected diverging trains only.

  • jaxbot

    I don’t think the 8th Ave line has timers, though. I think the T/O was just following a cluster of trains, but I’ll check next time I ride to see if they get white aspects on the way up. I’ve had times where the local is faster than the express, and times when the express is faster than the local. Usually during peak hours it makes more sense for me to ride the local because the express dwells forever as people cram in and out of it.

  • jaxbot

    Yeah, I used to be so confused why conductors using New Tech Trains would open and close the doors when the door system will allow them to close without reopening them fully. But one day I saw a woman grab the door and when they tried local recycling, she held onto the train and wouldn’t let it leave the station until she was let on. Wasted about 45 seconds.

  • jaxbot

    See my reply above — what do you think a good solution would be to mitigate the bad behavior of riders that results in train delays? Obviously recycling should only be done to doors that didn’t close, but can anything be done to get people to back away from doors that are half-closed?

  • Joe R.

    I don’t think you understand exactly how the neutering affected acceleration rates. The initial rate of 2.5 mph/sec remained the same. Older cars hold that rate to about 17 mph. NTTs in theory can hold it to 25 mph or so but the neutering capped the initial rate at around 20 mph. Once past the constant acceleration point the acceleration rates dropped dramatically. As an example 20 to 30 mph now takes about 20 seconds. 30 to 40 takes about 30 to 40 seconds. Prior to neutering the figures were roughly 6-7 seconds and perhaps 10-12 seconds, respectively. The NTTs are even faster, but as I’ve never been on an unneutered one, I don’t know the exact numbers. Based on their power output, you might be looking at 20 to 30 mph in under 5 seconds, and 30 to 40 mph in 7 or 8 seconds.

    When you do a little numerical integration, and assume a 3 mph/sec service deceleration rate, it takes roughly 72 seconds to do a half mile start to stop with the neutered trains, and they reach a peak speed of 37-38 mph before they need to start braking. Unneutered trains require about 62-63 seconds to do the same half mile, start to stop. I’m assuming they reach a peak speed of 40 mph, which only takes ~1000 feet, and then run at that speed until they need to start braking. In theory they can continue accelerating, maybe reaching 45 mph before needing to brake, but the time gains from that are minor, perhaps a second at most.

    Obviously want you described about opening and closing the doors adds yet more time, so with both things we could be looking at adding 15 to 20 seconds per local stop. That’s a lot of wasted time for really no gains whatsoever in safety.

  • sbauman

    what do you think a good solution would be to mitigate the bad behavior of riders that results in train delays?

    That’s easy. Run trains more frequently. The certainty that the next train will really be only 1.5 to 2.0 minutes away will discourage door holding. Moreover, greater service levels, for the same demand, means fewer passengers on trains, platforms and crossing the door threshold. This will result in less of the scheduled dwell time being spent with passengers crossing the door threshold.

    N.B. there are 4 components in my dwell time definition: 1 – train is stopped in station with doors closed; 2 – train doors open with passengers crossing the threshold; 3 – train doors open but no passengers crossing the door threshold; 4 – doors closed but train still stopped in station. This differs slightly from the literature definition. My first component is usually lumped with the braking. I prefer my definition because it isolates train movement from what happens when the train is motionless within the station.

  • sbauman

    The initial rate of 2.5 mph/sec remained the same…

    Two part piece-wise linear approximations served the engineering profession well before computers. The important quantity to calculate is the time required for the train to move 1100′ from its starting position. Why don’t you take your acceleration model and use a Runge-Kutta integration program to calculate this time. If you do, you will probably find the time difference to be less than 0.5 seconds. If you have a deterministic relation for how the acceleration declines as a function of the current velocity, I’ll run the number myself. N.B. I have done that calculation for PCC cars.

    it takes roughly 72 seconds to do a half mile start to stop with the neutered trains, and they reach a peak speed of 37-38 mph before they need to start braking.

    How long it takes a train to go between to stops is not what’s required to determine service level capacity for intermediate stations. The parameters are: the time for a train to stop within a station; the time the train remains stopped within the station; and the time it takes for a train to leave the station to a point where the signals before the station show a green aspect. One other parameter, if you want to get technical is the emergency braking stopping distance. This determines how closely trains may follow one another. It determines how far the train must travel, without a signal system, before it’s safe for the follower to enter the station.

    If instead of measuring the station-to-station time you measured acceleration and braking times within a single station, you would have a good handle on what the current signal system can handle.

  • AMH

    I’m talking about straight track approaching the station, not the curve some distance after the station.

  • joe lee

    North of Chambers there are a lot of switches in the area.

  • joe lee

    They are trying to protect the switches just north of this station.

  • joe lee

    Those switches at York Street, and north of East Broadway are used just in case of single tracking is required in the Rutgers Street tunnel, when needed.


Straphangers: Ancient Train Signals a Prime Culprit of Subway Delays

Has your subway been delayed recently? Blame New York City’s aging transit infrastructure, especially its outdated signal system. Then start fighting to make sure Albany fully funds the MTA’s next capital plan. A new report from the Straphangers Campaign shows just how prevalent signals failures are on the subway system. In 2011, the MTA sent […]