Jump to content

Attention: In order to reply to messages, create topics, have access to other features of the community you must sign up for an account.
Nick

headway shortened to 90 seconds? or not and are trains somewhat tailgating?

Recommended Posts

After noticing that the On the Go Travel Stations occasionally announced subways on the same track being due to arrive in the same minute, I wondered if the announcements were wrong or if safety was at risk. I had heard that the normal minimum headway, without supervisory permission for an engineer to key past a red light, was 2 minutes and 36 seconds, so, even allowing for rounding, same-minute arrivals should be the subject of a warning to someone somewhere. Now I'm told by the MTA that headway can be as short as 90 seconds. With rounding (because two arrival systems that I know of, Countdown Clocks and On the Go, give times in whole minutes only), a 90-second headway would allow two same-minute arrivals. I asked an engineer (not on the #7 line), who said that he thought 90 seconds might be true on the #7 line. I'm surprised that the signal system was upgraded that much and citywide without a public congratulation on such a large and costly project; I couldn't find anything on the MTA website about it (although maybe I didn't use the right search terms or dig into results deeply enough). Headway standards could easily differ between the IRT and the BMT/IND divisions, which have different gauges, preventing passenger trains from one from running on the other, so signal systems could be harmlessly incompatible. Is there more information on this subject? Is a 90-second headway normal throughout all lines? Are trains sometimes too close without a supervisor knowing about it in real time?

Share this post


Link to post
Share on other sites

1 hour ago, Nick said:

I asked an engineer (not on the #7 line), who said that he thought 90 seconds might be true on the #7 line. I'm surprised that the signal system was upgraded that much and citywide without a public congratulation on such a large and costly project;

You see, it’s human nature to weigh the bad impressions much more heavily than the good impressions. Even if the (7) could run a train past a station every 90 seconds, what do people actually observe?

Share this post


Link to post
Share on other sites
2 hours ago, Nick said:

IRT and the BMT/IND divisions, which have different gauges

Not true FYI, they both use standard. The difference is the width and length of the cars, and turning radius, (IRT trains are noticeably smaller). Most work trains use IRT spec cars because they can fit everywhere in the system, while BMT/IND cannot. 

  • Like 1
  • Upvote 1

Share this post


Link to post
Share on other sites
3 hours ago, kosciusko said:

Not true FYI, they both use standard. The difference is the width and length of the cars, and turning radius, (IRT trains are noticeably smaller). Most work trains use IRT spec cars because they can fit everywhere in the system, while BMT/IND cannot. 

Nick is likely mistaking the track gauge for the Loading Gauge, which defines the maximum height/width that can use a defined right of way. The A-division loading gauge is smaller that doesn't allow most equipment to use those routes.

  • Thanks 1
  • Upvote 1

Share this post


Link to post
Share on other sites
Posted (edited)

OTG show "scheduled" arrival times. Thats the time from the PDF timetable. Ignore the OTG kiosks.

Edited by bulk88
  • Like 1
  • Upvote 1

Share this post


Link to post
Share on other sites

The fixed block signal system on both divisions was designed to handle minimum train separation of a about 90 seconds. That said, NYCT never operated trains that close together as doing so loses you all flex capacity and recovery margin, making service extremely prone to slowdowns and backups. They normally kept tph to about 30, giving a 30 second margin of error. 

Over time, however, that capacity was degraded. In order to be effective at stopping trains, block signaling is designed around trains operating at their maximum attainable speed for the governed section of track. Consequently, the closer you operate trains to their maximum speed profile, the more capacity you have. Degrading brake rates and the scourge of timers have made such speed impossible, reducing line capacity throughout the system greatly. The reduction largely stays within what was flex capacity, so we’re still able to run approximately the same number of trains, but the second something goes south, everything goes to shit.  That lack of flex capacity and therefore operational flexibility is a main driver of the grotesque service disruptions we see, one could argue moreso than the disruption itself. 

CBTC somewhat ameliorates this. The newer cars operating on the 7 and L have relatively ok emergency braking rates, are uniform, and have the signal system calibrated to them in real-time, allowing those block lengths to change dynamically. Thus you get shorter headways. That said, you don’t necessarily need CBTC to get it’s headways. If you build a fixed block signaling system calibrated perfectly to maximum speeds, braking rates, and expected train separation, you can achieve 40tph in regular service like Moscow does today. It’s just a matter of precision and operational discipline — especially in the realms of things like station dwell and terminal ops. 

  • Upvote 2

Share this post


Link to post
Share on other sites

Tangential question about the kiosks, does anybody know why some kiosks use SubwayTime arrival data while other kiosks use the published timetables?

Share this post


Link to post
Share on other sites

Post #7 (@RR503) is reassuring. From that, I can assume that OTG data on nearness is not something to worry about, and if trains are too close supervisors likely know regardless of OTG.

If some OTG displays use the same data as the PDFs (@bulk88), that's interesting, because it means the NYCTA has decided back when drawing up schedules that trains can arrive in the same minute on the same track without needing supervisory approval each time. But in the past I thought that some paper-and-ink public schedules were overly optimistic, and maybe that's still the case, which would mean that actual schedules, headways, engineer decisions, and supervision are safer, and probably safe enough.

I don't know much about OTG having two different arrival datasets (@Mysterious2train), except that one OTG has a Trip Planner and the other has what some customer service workers think is a Trip Planner but isn't. I reported that misunderstanding to MTA.info and I understand the difference is that different vendors designed them. You can tell them apart by the differences in buttons at the bottoms of the default (resting) screens. Possibly, the two vendors drew from different data sources. I understand one OTG system is to be replaced by the other, but I don't know when. I don't remember which OTG system showed the close-together trains or if both systems did for different occurrences.

I haven't seen the simultaneity on the Countdown Clocks, but they use different data anyway (compared to at least one of the OTG systems), at least the video ones (I don't know about the green-and-amber LED ones and the red LEDs don't show times). I have occasionally seen video Countdown Clocks quickly shorten the stated waiting time, so I don't know what data they use.

The engineer was not saying that any #7 trains run at minute-and-a-half intervals (@CenSin), just that they could, which is all I was asking about.

I got the gauge info long ago (@ kosciusko and @Jsunflyguy) from a transit worker (who added something about 8, 10, and 12 feet and whose job assignment was on the South Brooklyn railway, a freight line) and you're likely right. But mostly not being able to share the rolling stock would still allow different headways by division, if NYCT chose to differentiate that way.

Thank you very much.

  • Upvote 1

Share this post


Link to post
Share on other sites
Posted (edited)

meanwhile in London...
 

 

Edited by HenryB
  • Like 1
  • Upvote 5

Share this post


Link to post
Share on other sites
1 hour ago, HenryB said:

meanwhile in London...
 

<every 90 seconds video>

Thats pretty good, if service was like that therewouldn't be much complaints. But NYC is different because it has local and express service and generally 4 tracks per trunk line. The equivelent would be a train every 180 seconds per track. Especially in NY where two trains can arrive and simultaneously in the same direction.

Share this post


Link to post
Share on other sites
47 minutes ago, N6 Limited said:

But NYC is different because it has local and express service and generally 4 tracks per trunk line. The equivelent would be a train every 180 seconds per track.

Would it though?

Seems to me if everything was coordinated almost perfectly - especially in the bottlenecks where 4 tracks become two, with a great signaling system and high enough speeds we could do one train every 45 seconds in the bottlenecks and 90 seconds outside the bottlenecks - save Rogers, Lenox and Myrtle junctions.

  • Upvote 1

Share this post


Link to post
Share on other sites

I'm not favoring or opposing any particular headway. I was concerned about safety if headway was frequently being violated and no one knew in real time. When safety is present, what headway is better for serving customers given how much rolling stock, how many crews, and so on are available is a separate question with complex answers.

@HenryB, I watched the whole video. @HenryB, @LGA Link N train, and @CenSin, my guess is that running trains very close together (safely) so that a delay of one train (say if a passenger got sick and needed help) could cause a domino effect of delays still need not delay an avarage passenger's arrival at their destination more than if trains were far apart and the same initial delay occurred (this assumes the passenger arrives at the departure point at a random time within a given rush-hour period). Mathematically, it should have no negative effect. On the other hand, in the close-together scenario with a delay, passengers on more trains would experience delays from the domino effect. In that case, the lack of any real delay, even of a fraction of a second, is accompanied by a perception of delay and customers' greater dissatisfaction. And the customers are being reasonable, since they can't make fewer trains run on alternate days so they could conduct and analyze scientifically systematic observations and solicit peer review of the results. So perception may have to rule and the perception debate may lie between shorter intervals with more on-train delays and longer intervals with more on-platform waits. I assume each city would study its own population and proceed accordingly; and even that is costly, since asking passengers would likely be uninformative and experiments on various lines in various stations and in various hours including both rush-hour periods would be needed. So perception is likely to continue to be the measure.

Share this post


Link to post
Share on other sites
On 10/04/2018 at 3:23 AM, N6 Limited said:

Thats pretty good, if service was like that therewouldn't be much complaints. But NYC is different because it has local and express service and generally 4 tracks per trunk line. The equivelent would be a train every 180 seconds per track. Especially in NY where two trains can arrive and simultaneously in the same direction.

The line shown in the video, the Victoria Line, is the equivalent of the L.

Twin tracked, no expresses, no shared running with other routes, CBTC.

Peak on the Victoria is 36 tph.

What’s the peak tph on the L?

  • Like 1

Share this post


Link to post
Share on other sites
12 hours ago, paolo999 said:

The line shown in the video, the Victoria Line, is the equivalent of the L.

Twin tracked, no expresses, no shared running with other routes, CBTC.

Peak on the Victoria is 36 tph.

What’s the peak tph on the L?

IIRC, 21TPH, 26 with a fully capable power system.

You have to keep in mind that both 8th Av and Canarsie are awful terminals for terminating trains.

Share this post


Link to post
Share on other sites
2 hours ago, bobtehpanda said:

IIRC, 21TPH, 26 with a fully capable power system.

You have to keep in mind that both 8th Av and Canarsie are awful terminals for terminating trains.

How much more TPH will be gained if tail tracks are added at 8th?

Share this post


Link to post
Share on other sites

FWIW although they have tail tracks at one Victoria Line terminus, they don’t use them, and instead use stepping back, to get the 36 tph.

I’ve read here before that there’s (maybe) a power constraint on the L. If that’s the only limitation, it seems like a job unfinished after all the work and investment for CBTC.

Share this post


Link to post
Share on other sites
18 hours ago, paolo999 said:

FWIW although they have tail tracks at one Victoria Line terminus, they don’t use them, and instead use stepping back, to get the 36 tph.

I’ve read here before that there’s (maybe) a power constraint on the L. If that’s the only limitation, it seems like a job unfinished after all the work and investment for CBTC.

Tail tracks allow for trains to enter a terminal faster than one with a bumping block since an over-run would just be an operational error and not a collision. I think that most if not all stations that end in a block, require a train to enter around 10mph or less. Even if you have a high speed crossover leaving the station, turnaround time is still limited by the train entering the terminal. 

  • Like 1
  • Thumbs Up 1

Share this post


Link to post
Share on other sites
19 hours ago, paolo999 said:

FWIW although they have tail tracks at one Victoria Line terminus, they don’t use them, and instead use stepping back, to get the 36 tph.

I’ve read here before that there’s (maybe) a power constraint on the L. If that’s the only limitation, it seems like a job unfinished after all the work and investment for CBTC.

Whats Stepping Back?

Share this post


Link to post
Share on other sites
18 hours ago, paolo999 said:

FWIW although they have tail tracks at one Victoria Line terminus, they don’t use them, and instead use stepping back, to get the 36 tph.

I’ve read here before that there’s (maybe) a power constraint on the L. If that’s the only limitation, it seems like a job unfinished after all the work and investment for CBTC.

It's not really an unfinished job. You have to keep in mind that when this CBTC was first conceptualized for the (L) it was not for growing usage, but rather as a test track for the technology. The projections for (L) service requirements never anticipated anything like the rapid gentrification of Williamsburg to ENY occurring. And installing a power system is at best tangentially related to upgrading signalling.

As far as power supply goes, you also have to keep in mind that the Victoria Line was designed from the outset in the 60s to be a very high-capacity line. The (L) was designed in the 1910s.

  • Upvote 1

Share this post


Link to post
Share on other sites
On 4/7/2018 at 7:49 PM, Nick said:

After noticing that the On the Go Travel Stations occasionally announced subways on the same track being due to arrive in the same minute, I wondered if the announcements were wrong or if safety was at risk. I had heard that the normal minimum headway, without supervisory permission for an engineer to key past a red light, was 2 minutes and 36 seconds, so, even allowing for rounding, same-minute arrivals should be the subject of a warning to someone somewhere. Now I'm told by the MTA that headway can be as short as 90 seconds. With rounding (because two arrival systems that I know of, Countdown Clocks and On the Go, give times in whole minutes only), a 90-second headway would allow two same-minute arrivals. I asked an engineer (not on the #7 line), who said that he thought 90 seconds might be true on the #7 line. I'm surprised that the signal system was upgraded that much and citywide without a public congratulation on such a large and costly project; I couldn't find anything on the MTA website about it (although maybe I didn't use the right search terms or dig into results deeply enough). Headway standards could easily differ between the IRT and the BMT/IND divisions, which have different gauges, preventing passenger trains from one from running on the other, so signal systems could be harmlessly incompatible. Is there more information on this subject? Is a 90-second headway normal throughout all lines? Are trains sometimes too close without a supervisor knowing about it in real time?

Trains can't tailgate - even if the engineer (actually, train operator) were keying by multiple red signals, there are two behind a train, and you have to slow to a crawl to key by. The fixed-block signal system is flawed but it's actually pretty good at keeping a train from hitting another. 

I've also seen same - minute arrivals indicated by subway-time, countdown clocks and the kiosks as well. 

Generally, I find these to be the result of merging. 

I see it quite frequently at 57st - 7 av with the R/W and the N. The N is on the express tracks until 42nd street, but then crosses over to the local. An N and a W or an N and an R that are arriving at 42nd-Times at the same time will likely show the same arrival time at 57/7 until one of them actually LEAVES times square. In the absence of ATS and or CBTC - the GTFS data and subsequently the countdown clocks have no way of knowing which train is going to get "the lineup" (The signal to proceed on a particular route at an interlocking) first. 

Share this post


Link to post
Share on other sites
Posted (edited)
On 4/12/2018 at 9:47 PM, N6 Limited said:

Whats Stepping Back?

The crew gets off the train and lets another crew take over. In the meantime, as their (former) train is departing, they make their way down to the correct part of the platform to be in the right position to take the next train out (while that crew steps out). As opposed to having the motorman stay on the same train, but walk to the other end of the platform (which takes about 2-3 minutes) and then get in to start charging the train and whatnot.

Edited by checkmatechamp13
  • Like 1
  • Thanks 1
  • Thumbs Up 1

Share this post


Link to post
Share on other sites

@itmaybeokay, track-switching makes sense and I didn't know enough about routes to consider that. As to keying by, I think they can come close enough to offload passengers from one train door-to-door into another, but it makes sense that they'd approach very slowly. (Larry King, then a radio host, told a story of misunderstanding a request that he use his car to give a push to someone to get their car started; he swung around the block and came from behind at 35 miles per hour. Evidently, they remained friendly despite the injury.)

Share this post


Link to post
Share on other sites
19 hours ago, Nick said:

@itmaybeokay, track-switching makes sense and I didn't know enough about routes to consider that. As to keying by, I think they can come close enough to offload passengers from one train door-to-door into another, but it makes sense that they'd approach very slowly. 

Yes, trains can get close enough to literally touch, but it requires slowing down to a near-stop (slower than walking speed) at every automatic signal. So, it does prevent trains from tailgating, since the following train has to slow to a crawl every 1000 feet or so, and trains neither accelerate nor stop on a dime. Also supervision is going to notice if you start keying by multiple signals, in all likelihood. 

You can read more about Automatic key-by here: There's a wealth of information on the signalling system on that site, though not all of it is current anymore. 

https://www.nycsubway.org/wiki/Subway_Signals:_Train_Stops#ak

Share this post


Link to post
Share on other sites

The headway is typically 2 minutes 24 seconds to 3 minutes as of last September, at least where the system uses fixed-block signaling (almost the entire system still using it), although I don't know if that typicality was a minimum or an average. "CBTC can run 40 trains per hour per line, or a train every 90 seconds. Due to physical limitations in the system, the best the subway can hope to see with a CBTC system is probably around a train every 120 seconds, or 30 trains per hour; still, that’s an improvement over today’s subway, which typically runs only 20 to 25 trains per hour." This is according to Rich Barone, of the Regional Plan Association, according to The Village Voice, September 7, 2017 (https://www.villagevoice.com/2017/09/07/meet-the-century-old-technology-that-is-causing-your-subway-delays/ (as accessed 4-22-18)).

The L line has CBTC and the #7 line is being fitted with it, according to the same Voice source, so I don't know if the #7 has shorter headways yet. I just took the L twice in the pm rush hour and the actual headways were sometimes about a minute.

In answer to my opening post, we mostly do not have 90-second headways, if we have any besides on the L, but trains may be switching tracks so that same-minute arrivals are not likely dangerous.

Share this post


Link to post
Share on other sites
10 minutes ago, Nick said:

The headway is typically 2 minutes 24 seconds to 3 minutes as of last September, at least where the system uses fixed-block signaling (almost the entire system still using it), although I don't know if that typicality was a minimum or an average. "CBTC can run 40 trains per hour per line, or a train every 90 seconds. Due to physical limitations in the system, the best the subway can hope to see with a CBTC system is probably around a train every 120 seconds, or 30 trains per hour; still, that’s an improvement over today’s subway, which typically runs only 20 to 25 trains per hour." This is according to Rich Barone, of the Regional Plan Association, according to The Village Voice, September 7, 2017 (https://www.villagevoice.com/2017/09/07/meet-the-century-old-technology-that-is-causing-your-subway-delays/ (as accessed 4-22-18)).

The L line has CBTC and the #7 line is being fitted with it, according to the same Voice source, so I don't know if the #7 has shorter headways yet. I just took the L twice in the pm rush hour and the actual headways were sometimes about a minute.

In answer to my opening post, we mostly do not have 90-second headways, if we have any besides on the L, but trains may be switching tracks so that same-minute arrivals are not likely dangerous.

So even if (MTA) can achieve 30 TPH on certain lines, where are they gonna get the extra trains necessary, since the orders for new cars now are basically replacement for existing livery?

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×

Important Information

By using this site, you agree to our Terms of Use.